Research Bulletins

^M^^^^ AmericanFoundationfortheBlind ^^ Research Bulletins

Contents

Latif Morkos Cozman On the Possibility of an Entirely Extracranial Visual Prosthesis

p. W. Davall and J. M. Gill A Method for the Comparative Evaluation of Visual and Auditory Displays

G. A. James and J. M. Gill A Pilot Study on the Discriminability of Tactile Areal and Line Symbols for the Blind

Loren T. Schoof II An Analysis of Optacon Usage

Ching Y. Suen and Michael P. Beddoes Spelled Speech as an Output for the Lexiphone Reading Machine and the Spellex Talking Typewriter

N. V. Clarke, G. F. Pick, and J. P. Wilson Obstacle Detection With and Without the Aid of a Directional Noise Generator

Robert B. Welch and Charles Hallenbeck , Editors Recombination and Visual Perception --Ten Papers--

Johan Weckroth , Pentti Miettinen , and Eila Weckroth The Light Permeability of the Retina: Experiments With Blind, Color-Blind, and Seeing Subjects American Foundation ForTheBlind

15 WEST W\ CT-^^^n . C.I

American Foundation for the Blind Itesearch Bulletin

June 1975

Leslie L. Clark, Editor-in-Chief Zofja S. Jastrzembska. Associate Editor Audrey C. Gilmour, Copy Editor Michael E. Monbeck, Production Editor

The Research Bulletin is a piiblication of the International Research Information Service (Leslie L. Clark, Director) , American Foundation for the Blind, 15 West 16th Street, New York, New York 10011 USA.

Published irregularly twice a year, the Research Bulletin may be purchased for $2.50 per copy. Subscriptions for four issues are $9.00.

Articles in the Research Bulletin are abstracted in Excerpta Medica, Psychological Institute for the Han- Abstracts , and Rehabilitation Literature and by the Swedish dicapped and the Institute for Scientific Information of the U.S.S.R. Academy of Sciences

Manuscripts and other material to be considered for publication in the Research Bulletin should be siibmitted in duplicate to the editor-in-chief. All correspondence concerning manuscripts, editorial policy, and further information about research reported in the journal should also be directed to the editor-in-chief.

All correspondence concerning subscriptions and orders for back issues should be sent to the Publications Division of the American Foundation for the Blind. Printed in the United States of America. Digitized by the Internet Archive

in 2010 with funding from Lyrasis IVIembers and Sloan Foundation

http://www.archive.org/details/researchbulletin29lesl , ,

CONTENTS

Page Page

1 On the Possibility of an Entirely 111 Proposals for the Study of Extracranial Visual Prosthesis Adaptation to Anomalous Causal Latif Morkos Cozman Schemata Ian P. Howard 9 A Method for the Comparative Evalu- ation of Visual and Auditory Dis- 114 The Effect of s's Interpreta- plays tion of "Straight Ahead" Upon p. W. Davall and J. M. Gill Measures of Prism Adaptation Charles S. Harris 23 A Pilot Study on the Discriminabil- ity of Tactile Areal and Line Sym- 116 Speculations on a Model of bols for the Blind Prism Adaptation G. A. James and J. M. Gill Robert B. Welch

33 An Analysis of Optacon Usage 118 Past, Present, and Future of Loren T. Schoof II the Recombination Procedure Ivo Kohler 51 Spelled Speech as an Output for the Lexiphone Reading Machine and the 123 The Light Permeability of the Ret- Spellex Talking Typewriter ina: Experiments With Blind, Color- Ching Y. Suen and Michael P. Beddoes Blind, and Seeing Subjects

Johan Weckroth , Pentti Miettinen , 67 Obstacle Detection With and Without and Eila Weckroth the Aid of a Directional Noise Gen- erator CURRENT RESEARCH NOTES N. V. Clarke, G. F. Pick, and J. P. Wilson 145 Blind Mobility Current Research Program 87 Recombination and Visual Perception John Armstrong Robert B. Welch and Charles

Hallenbeck , Editors 14 9 An Electromechanical Numeric Braille Display: A Familiar Tactile 90 Prism Adaptation: Learning As- Representation of Electrically En- pects and Proprioceptive Con- coded Digital Signals tributions George F, Dalrgmple Edward Taub and Israel A. Goldberg 153 The Auto-Com at Kennedy Memorial Hospital: Rapid and Accurate Com- 93 Processes Involved in Behavior munication by a Non-Verbal Multi- Observed in Recombination Handicapped Student Studies George F. Dalrymple , Ann Bullock George Singer, Meredith Wallace and Janice M. Danca and Margaret Austin 157 Report of the Perceptual Alterna- 97 Some Effects of Exposure to Op- tives Laboratory for the Period tical Disarrangement on Adapta- July 1973 - June 1974 tion to Optical Rearrangement Emerson Foulke John J. Uhlarik 187 Auditory and Tactual Displays for 101 The Subject: A Neglected Factor Sensory Aids for the Visually Im- in Recombination Research paired David H. Warren and Bruce B. J. H. Gill Piatt 197 Non-Visual Computer Peripherals 107 The Possible Role of Eye Muscle J. H. Gill Potentiation in Several Forms of Prism Adaptation 213 Computer Braille System Sheldon M. Ebenholtz Don Keeping

109 Motor Transformation Learning John C. Hay (Continued on the next page.) Page Page

217 Aids for the Speech-Impaired: Inter- MISCELLANEOUS nationally Coordinated Development Work 229 Abstracts of Literature on Current Birger Roos Research on Education for the Visually Handicapped 221 Aspects of Some Problems Relating Michael J. Tobin to the Transliteration of the Hausa Language Into Braille 279 Abbreviations of Organizations James I. Murnane

227 Blind Programmers at Rolls Royce - 283 RESEARCH BULLETIN SUPPLEMENT Newsletter 1, July 1974 Terry Tate 291 PUBLICATIONS OF NOTE ,

ON THE POSSIBILITY OF AN ENTIRELY EXTRACRANIAL VISUAL PROSTHESIS*

Latif Morkos Cozman

Following the discovery by V^Jhen a path taken by research Forster in 1929 and Krause and Shum is seen to be leading toward an unin 1931 that the electrical stimula- desired application, it is useful to tion of a point on the surface of the explore the reasons for our disap- visual cortex caused a blind subject pointment, and to seek other means to see a point of white light in a of overcoming the difficulties. particular place in the visual field; Despite the disfavor with which the that the stimulation of a different idea has met, the stimulation of the point of this cortex caused the per- visual cortex via extracranial elec- ception of another point of light in trodes to an intaat brain is a possi- a different place in the visual field; bility that should be thoroughly in- the possibility of an electronic vi- vestigated before being ruled out. sual prosthesis emerged based upon a miniaturized TV camera capable of We deal here with some possi- transforming light and dark picture bilities of stimulating the visual elements into discrete electric cur- cortex from extracranial electrodes, rents corresponding to them. and also with an arrangement for eliminating the need for a scanning A television system involving procedure; these eliminate the intra- the use of the human brain itself cranial implant, with its multiple should not be attempted until it is wires or channels, and the elimina- sure that no possible harm would be tion of the scanner itself. caused. Present research involves the introduction of thousands of microelectrodes in contact with the ELIMINATION OF SCANNING visual cortex. The scanning procedure was de- Not surprisingly, prudent neur- veloped out of its role in transmit- ologists and intelligent patients ting and receiving facsimile and TV might well have serious reservations images. However, it is not necessary about the procedure. It might have to imagine that scanning should be surfaced regarding hearing aids if necessary in cases where radio or an extracranial deaf aid did not ex- wire links are not used (as for exist and so much as a single electrode ample in transmitting pictures in was proposed to be introduced in con- photography, microscopy, and ordinary tact with the auditory cortex. Vir- telescopy). In these cases, all the tually no one would have accepted picture elements are transmitted op- such a hearing aid, even though its tically and reproduced simultaneously. construction might be very much more In a visual prosthesis each picture simple than a visual aid. element has to be transformed into a minute electric current, and discrete electric currents can be acted upon by suitable electrostatic lines of

force (an ES lens) , and by suitable *Paper presented to the University magnetic lines of force (an EM lens) of Alexandria, Egypt, April 3, 19 74, to maintain their flow (with an ES .

lens) in parallel directions. Thus, will arise in the visual device to be t±ie possibility exists of dropping shown. C and D are two coaxial coils out the scanning procedure with all of wire through which direct current its disadvantages in a visual pros- in the correct direction passes, the thesis. This would very much simplify aim being the generation of an axial the required miniature TV camera. magnetic field. Between the two stop- Mention will be made below of how the pers is a dilute solution of sodi\am minute electric currents resulting chloride. A pulse generator and an from the transformation of the light ac ammeter complete the circuit. and dark of the image's picture elements could be separated from each To begin, one single strip in other and constrained to follow paral- B is connected into the circuit by lel paths until they impinge on dif- inserting a pin in its socket. The ferent points of the visual cortex. rectified current is read on the ammeter after switching on the pulse Figure 1 illustrates that numer- generator. Then the experiment is ous signals can flow simultaneously repeated for 2, 3, 4. . . n strips in parallel paths through a dilute connected together in circuit. It aqueous solution of cerebrospinal will be seen that the ammeter regis- fluid (such as that inside the brain ters in every case a current that is and nerve cells) equal (to a very near approximation) to the current for one strip, multiplied by the number of strips of B in connection with the circuit. This could never be the case unless the currents in the weak electrolyte solution followed parallel paths. For this reason, when a strip is dropped out, the current for it disappears. It should be noted that the whole disc in A and the numerous strips in

S form parallel plate condensers , the lines of force between which would be parallel, particularly when aided by the axial magnetic field. Thus, we are actually using an electrostatic lens combined with a magnetic lens. PULSE GENERATOR ELECTRODES Figure 1. Simultaneous Signal Flow EXTRACRANIAL in Parallel Paths Through a Dilute

' Figure 2 shows the position of ' Aqueous Solution of Cerebro- ' Spinal Fluid. the head, the converging lens, the TV camera 5, 1, and the extracranial electrodes 5 and 7. When 5 is positive, 7 is minus and Glass stopper A closes one end ' vice versa. of a glass. tube, with a platinum disk Thus , the curved lines shown at the firmly attached to the inner surface. bottom of Fig. 2 may represent the Stopper S closes the other end. In- electrostatic field. They can repre- stead of a solid platinum disk, it has sent, however, the magnetic field by on its inner surface a similar disk inserting a magnetic North pole at 5, that has been cut into a large number a magnetic South pole at 7, a magnetic of separate platinum strips, to each of North pole at 8 and a magnetic South at Fig. 3. By which is attached a metal wire, a re- pole 9, see adjusting sistor, and a socket that connects the the positions of 8 and 9, the EM will path to a conducting disk F if the pin coincide with the ES field. is put inside the socket. Either one It be that the or more of the strips on B can be put should noted rea- into circuit by putting the pins in son for adopting the positions shown the respective sockets. The resistors in Fig. 2 is to confine the passage of currents through the visual cor- fi'2 ^1/ / •''3/ ^4 f etc. may differ in their ohmic values, and their inser- tex, the area connected with vision. electrode 5 directly on tion makes the voltages at the various Installing the forehead (in a position parallel strips in S differ, a situation that to electrode 7) would entail the .

injury resulting from a disharmony with the normal electrical activity of these areas.

The latest research in neurophysiology and electroencephalography has paved the way for fulfilling these requirements in the most efficient way.

But first, let us describe a miniature TV camera suitable for a non- scanning extracranial-electrode device. In Fig. 4, a conductive glass disk, 1, in addition to being transparent to light rays, passing them to a photoconductive semiconductor layer, 2, on its back surface, puts layer 2 in electrical contact with the high voltage supply HT

Layer 2 is in touch with a first mosaic, 3, consisting of a very large number of very minute conducting glass slabs or elements, arranged in checkerboard fashion, with horizontal and vertical insulating thin linings separating them in a uniform way. A second mosaic, 5, of conductive glass slabs, similar in all respects to the first mosaic, 3, is transparent to Figure 2. Position of Head, Converg- visible rays falling on it from elec-

ing Lens, TV Camera, and Extra- tric lamp bulbs , 6 , and also serves cranial Electrodes. passage of current through other areas as well, where stimulation of motor or sensorimotor centers is possible, in addition to a much longer electrical path as compared with that in Fig. 2.

It would appear that by employing intracranial electrodes to contact the visual cortex we can avoid the passage of currents through the brain. But the electrical circuit must be completed by using a bus electrode , whether intracranial or extracranial.

Thus , since there can be no escape from currents through the brain, they must be carefully chosen so that:

1. No stimulation of the motor or sensorimotor centers is elicited that might result in convulsions or epileptic effects;

2. Current density is low enough to avoid excessive heat generation in the brain that might cause un- consciousness or sleep; or any Figure 3. The Magnetic Field . . : ,

The ES field referred to above be- + tween 3 and 5 , and the employment of an EM field coinciding with it (Fig. 1 3) is 7/////////////2^ achieved with the aid of appropriate magnetic poles. msmffiSEismmB ELECTROPHYSIOLOGY IN SERVICE OF THE BLIND

Figure 4. Schematic of Miniature TV Electrophysiology , and particu- Camera for Non-scanning Device. larly electroencephalography, have revealed many significant facts about the brain which, unfortunately, have as the return path to the ET of the not been coordinated so as to furnish source of current. Between the first the best understanding of the electrophysiology of an mosaic, 3, and the second mosaic, 5, vision— understanding that must be established the is a very thin layer, 4, of colloidal before graphite crystals suspended in an in- difficulties of building a scanless sulating medium, such as thin mineral extracranial electrode prosthesis can be overcome. How can the most effi- oil. The second mosaic, 5, consists of tiny slabs of conducting glass in cient means of transmitting electric an arrangement which is in exact sym- current through the brain be found if metry with the tiny slabs of the first the true electrical activity of the brain cells is not known? It is no slab, 3, in checkerboard pattern with horizontal and vertical insulating longer sufficient to regard a brain cell as a thin linings, each slab in the mosaic, wet battery generating about 0.1 volt, because this is only 5 , being opposite to a corresponding partially true. More useful is the slab in mosaic, 3. The mosaic, 5, forms the border of the device furth- concept that there are brain cells ermost from the impinging light and like those in the visual cortex that as the conductive disk, and is main- function tuned or tunable electri- 1, oscillate, tained in good electrical contact cal circuits which either giving rise to the or alpha with the scalp in a position as near dominant brain rhythm of about 10-per-second as possible to the occipital lobe of or propagate this frequency most ef- one hemisphere (Figs. 2 and 3). ficiently as a consequence of resonance. These notions yield to further The second mosaic, 5, is not elaboration when compared with some connected directly to ET , for the of the basic references in the field. latter is connected to an electrode, Based on the literature, the follow- in contact with the scalp in a 7, ing comments and elaboration illus- position opposite the space between trate the approach the two hemispheres (Figs. 2 and 3)

1. In 19 34, E. D. Adrian and As in the ordinary vidicon, B. H. C. Matthews showed that the surface of layer 2 furthest from alpha rhythms could indeed be layer 1, will have built on to it an driven by repetitive flash electrical image of potential differ- stimuli. This is an instance ences that corresponds to the light of resonance wherein oscillation and dark areas of the optical image, at a certain frequency induced and because conductive glass, 3, and the same oscillation in a cir- the conductive slabs in 5 form a com- cuit tunable to the same fre- posite parallel plate condenser with quency (Walter, 1965, 97). parallel lines of force, the colloidal p. graphite cirystals in 4 will tend to 2. Visual sensations were elicited orient in the same direction; because from the occipital lobes (Walter, the graphite conducts, parallel and 1965, 56) discrete currents of electricity* p. will reach 5 because of the presence 3. Adrian and Matthews proved that of the insulating linings in 3 and 5. the alpha rhythm arises in the visual association areas of the *UK Patent No. 1172412, British Patent occipitus (Walter, 1965, p. 56). Office, 25 Southampton Buildings, Lon-

don, WC2 , England. . "

The occipital alpha rhythms aris- the membrane has been depolarized ing in the visual area of the to a certain cortical level can cortex are as a rule absent or be observed in these cells. The faint in the blind (Babsky, arrival of nerve impulses at such Khodorof, Kositsky, and Zubkov, an automatically discharging cell 1970, p. 277). This is a most leads to an increase in membrane significant observation, and if depolarization, and consequently considered in conjunction with to a heightening of the frequence the known fact that the stimuli of impulse discharge. (Babsky, transmitted to the brain are in et al., p. 96.) This rise of the form of individual groups or frequency of the oscillation "volleys," the amplitude and produced by nerve cells on the duration of the individual im- start or stopping of impulse pulses identical but differing trains has detracted from their in frequency, the whole visual property of tuning, and that is process must be based on the why this writer refers to them alpha rhythm— its production and as "tunable." However, this fre- transmission from the retina to quency rise is of paramount im- the visual cortex. In other portance in the visual process words, the neurones in the optic Within 0.1 to 0.5 second of il- nerve and in the visual cortex luminating the eye, the arriving are fundamentally tuned to the impulses rise in frequency from frequency of the alpha rhythm. the alpha frequency of approximate- This is in agreement with what ly 10-per-second to a very high physiologists have reiterated, frequency; resonance with the namely that the nerve fibre is brain cells suffers, and "visual a chain of relay stations, many sensation disappears when the stations to the inch, constantly eyes are immobile." (Babsky, the regenerating the signal. Before et al . , p. 201.) "During con- the present instance, however, tinuous action of light on a vi- no one has referred to the tunable sual receptor impulses quickly nature of the brain cell, and its cease in the corresponding fibres amenity to resonance as the re- of the optic nerve and this shows laying principle. the important role of eye movements in the visual process." Realizing that the brain cells (Babsky, et al., p. 201.) "Yarbus act as tuned circuits, it is found that the eye, in looking at easy to understand why "The 19th any object, made incessant imper- Century raised its poetic eye- ceptible jumps. As a result the brows at nature's prodigality. retinal image was continually The 20th Century is no less sur- displaced from one point to an- prised by the economy of struc- other so producing new impulses ture and function discovered in in the ganglion cells and the the mechanics of life." (Walter, nerve fibres leading from them. 1965, p. 113.) In the proposed (Babsky, et al., p. 201.) prosthesis the aim is toward a similar high efficiency, and one Spontaneous activity is closely should not be content with what akin to evoked activity, since it is good enough in the electroly- is not possible to distinguish in sis of water, for example, in the the records between the unit migration of the positive cation spikes or slow wave in either. to the cathode and the negative (Davson and Eggleton, 1962, anion to the anode. In other p. 1020.) Thanks to this princi- of words , we must not ignore the ple, artificial stimulation brain cells. the visual cortex, if undertaken with due consideration to effi- 6. Many cells of the central nervous ciency and other requirements, prosthesis. system possess automatism , that will provide a benign is they generate rhythmic impulses even in the absence of Activity in individual neurones external stimuli. Constant slow seemed to be propagated as a fluctuations of the membrane po- traveling wave from one part of tential, accompanied with iso- the cortex to another. (Davson lated or multiple discharges of and Eggleton, 1962, p. 1021.) A^ action potential occurring when traveling wave through a medium, : : . .

like the transverse waves on the needs energy which the alpha surface of water, implies that rhythms supply. the frequency of the wave is the same as the frequency of the 2 In persons in whose EEG records components various particles or there are no alpha rhythms , these reached successively by the wave; rhythms are consumed in vivid another bit of evidence in favor visual imagination. of the primacy of the alpha-rhythm frequency in the brain cell! 3. In persons with persistent alpha rhythms these rhythms are not From these last two points it consumed because the energy will be seen that if electrical im- needed is supplied by means of pulses having the frequency of the auditory, kinaesthetic, and alpha rhythm are made to act on some tactile imagery. cells of the visual cortex, they would propagate in the direction determined by the electrostatic and The inference is that alpha rhy- magnetic fields arranged as to focus thms, in all three cases are essential the wave for vision and visual imagination; that is why they are not blocked Other evidence relating to the when the subject is dependent on linkage of alpha rhythms and vision perception other than visual.

- exists , . ,.,_. In sum, we have attempted to 1. Although it is sometimes said show, through appropriate correla- that alpha rhythms cease when the tion of the functions of alpha eyes are opened, or a mental pat- rhythms and the characteristics tern is seen, or when one in- of brain cells that it is possible dulges in visual imagery, more to build up a very simple visual accurate observation shows that prosthesis using an intact skull and such waves do not cease upon the brain, without the need for scanning arrival of visual impulses, since devices, by stimulating the visual this occurs after a relatively cortex from extracranial electrodes long period. Alpha rhythms are via impulses at alpha-rhythm fre- reestablished when the eyes are quency, by guiding the path of the opened, provided the retinas are discrete currents produced by the uniformly stimulated. If, how- scanless TV camera employed, through ever, the subject tries to fix shaping provided by electrostatic objects in the visual field, the and electromagnetic fields. alpha rhythms are blocked.

2. Absence of alpha rhythms in Comments on the analysis would adults is normally a sign of be welcomed. vivid visual imagination.

3. Individuals with persistent alpha rhythms which are hard to block with mental effort tend to audi-

tory, kinaesthetic , or tactile perception rather than visual imagery. In these cases the alpha rhythms continue even when the eyes are open and the mind is active or alert.

Strangely enough, nobody has ever attempted to synthesize larger con- structs on these grounds. The author suggests, however, that the alpha rhythms

1. Are blocked when an image is perceived because such a process . . . 8

REFERENCES

Babsky, E. B., Khodorof, B. I., Talaat, M. , Hamdi , H. , Gabrawi , B. G. , Kositsky, & G. I., Zubkov, A. A. Ab del- Rahman , U. Talaat' Human physiology . Volume 2. Physiology Volume 4. Moscow: MIR Publishers, 1970. Cairo: 1971.

Bell, George H. , Davidson, G. Norman, & Scarborough, H. Textbook of Walter, W. G. The living brain. New physiology and bioahemistry York: Penguin Books, Pelican Edi- Edinburgh: E. S. Livingstone, tion, 1965. 1961.

Davson, H. , & Eggleton, G. Princi- ples of human physiology London: J. A. Churchill, 1962.

A METHOD FOR THE COMPARATIVE EVALUATION OF VISUAL AND AUDITORY DISPLAYS

P. W. DavallandJ. M. Gill*

Ahstraot : In reading and mo- states. They found that, in general, bility aids for the blind it is de- multiple stimulus encoding is a sat- sirable to design the display to isfactory procedure for increasing optimize the man-machine interface. the information transmission rate This paper describes a possible associated with such displays. Mudd method for comparing various dis- (1965) studied the relative effective- plays. ness of freauency, intensity, duration and interaural difference dimen- Five subjects used compensatory sions of a pure-tone binary display. tracking with random input signals Frequency proved to be the most ef- (five different cutoff frequencies) fective dimension For the purposes with five one-dimensional displays of cueing; intensity was the least where the error is represented by: effective. 1) Visual: deflection, 2) Auditory: mark-space ratio, 3) Auditory: match- Roffler and Butler (1968a) found ing frequency, 4) Auditory: beat that listeners could locate auditory frequency, 5) Auditory: amplitude stimuli accurately in the vertical matching. plane when the stimulus was complex and included frequencies above 7kHz. The significance of the coher- Roffler and Butler (1968b) then found ency between the input signal and that subjects tended to place the the subjects' output is discussed. audio stimuli on a vertical scale in The measured closed- loop frequency accordance with their respective response is compared to that obtained pitch. Higher-pitched sounds were from a modified form of cross-over perceived as originating above lower- model. Finally, two performance pitched sounds. parameters are proposed for assessing the various displays. Vinje and Pitkin (1971) studied human operator dynamics for aural- compensatory tracking using pitch of INTRODUCTION the tone to represent the magnitude of the tracking error. Error polar- Research on the design of audi- ity was indicated in the two-ear tory displays has usually been moti- display by switching the tone between vated by the desire to provide non- ears as a function of error sign. visual displays for those who cannot For the one-ear display, error polar- use vision or to supplement a visual ity was indicated by using modulated display particularly in aeronautical and unmodulated tones. The describ- applications. ing function and remnant data indicated that humans can control as well Pollack and Ficks (1954) studied with aural cues as with visual cues multi-dimensional auditory displays for input signals with cutoff fre- in which each variable had only two quencies of 0.27 to 0.56 Hz.

A variety of two-dimensional Department of Engineering, University auditory displays have been built. of Warwick, Coventry CV4 7AL, England. For instance. Black (19 68) developed . .

a display where the horizontal co- presentation. The operators' task ordinate was represented by time was to minimize the error. delay and amplitude of the signal and the vertical coordinate by frequency {100-400 Hz). Fish and Error Signal Presentation Beschle (1973) also used frequency (200-7000 Hz) for representing the Visual feedback . A horizontal vertical position of the scan but line on a cathode-ray tube display interaural intensity differences moved vertically about a center zero, (up to 40 dB) for the horizontal. the deflection being a function of the error signal. By pushing on the Phillips and Seligman (19 74) de- joystick the line was moved up the veloped a multi-dimensional auditory screen, while pulling brought the display in which frequency, amplitude line down. Full scale deflection was and timbre are all utilized. Robin- * 10 cm in response to an error signal son (Gill, 1974) also developed a of *10 volts (0.5 kg-per-volt) two-dimensional display, as a non- visual equivalent to the cathode-ray Mark-space audio feedback . The oscilloscope, where the frequency of error signal was presented as short the signal depends on the vertical bursts of a single tone. Pulling the coordinate and time delay for the joystick reduced the duration of the horizontal. However, there has been pulses and increased the silent peri- little systematic comparison of the ods, while pushing reversed the ef- various types of auditory displays, fect. The fundamental clocking peri- even though Kramer (19 62) mentioned od for repeating the pulses was 0.15 the need for this over a decade ago. seconds (Fig. 2) . The zero error state was defined as unity mark- This paper describes a method space ratio between noise and silence. for comparing various displays and applies the technique to one visual Matching- frequency feedback . A and four auditory one-dimensional 1-kHz square was fed to the left ear displays. The technique can be ex- as a reference tone while a variable- tended to multi-dimensional displays. frequency square wave of equal amplitude was fed to the right ear. The frequency of the variable signal was EXPERIMENTAL METHOD ' - --^ controlled by the joystick, and could AND ANALYSIS vary from zero to 3 kHz for full- scale error deflection of -10 volts The human operator was presented to +10 volts. Zero error was record- with an error signal e (t) , which was ed when the two inputs were of match- the difference between the input, ing frequency x(t), and the output, y {t) , (Fig. 1). The output was measured from two Beat- frequency feedback . The strain gauges (making two arms of a signal inputs were the same as those bridge section) strapped across a in "Matching-freauency feedback" stiff joystick. The choice of a (above) , except that both inputs were force (as opposed to a position) joy- presented to both ears simultaneously. stick considerably reduced the re- The effect was different, however, in sponse time delay introduced by the that signals of three frequencies muscle motor action of the arm. The ' could be heard; a beat frequency was input, x{t), was band-limited gaussian present as well as the two funda- white noise. Five bandwidths of 0.3, mental frequencies. The beat fre- 0.5, 0.7, 1.0, and 1.5 Hz were used quency decreases as the two funda- for each form of error-signal mental frequencies come closer

—;0075i O IS

G(iw) xlt) c(t) y(t) %- Human Operator

Figure 1. Closed-loop Configuration Figure 2. Zero Error State for Mark- for Human Operator Tracking Tests. Space Audio Feedback Error- Signal Presentation.

10 — — — . .

together. The advantage of this form estimate of the cross- xy -" of frequency matching is that it does spectrum between x and y not require the operator to be tone at frequency u. sensitive. squared-coherency esti-

Amplitude-matching feedback . A mate between x and y at 1-kHz square wave was presented to frequency to both ears. The amplitude presented to each ear as a function of the er-

ror signal is shown in Fig. 3. The The signals x(t) , y(t) and e (t) loudness of the tone presented in each were sampled f ive-times-per-second. ear was approximately a linear func- Approximately ten-minute records were tion of the error (the ear being a taken for each type of input signal logarithmic device to a first approxi- (0.3, 0.5, 0.7, 1.0 and 1.5 Hz-cut- mation) . Zero error was recorded off frequencies. when the amplitudes were matched. The sample records were divided into blocks of N points (128 in this case) for Fourier analy sis using a ci Left ear 90 L Right tar* '2' Amplitude radix fast Fourier algorithm SO- 'db" (Blackman and Tukey, 19 58; Wellstead, 1971) producing N/2 spe ctral coeffi- TO- cients over the freauen cy range - 2.5 Hz. Approximately 30 short- 60- term estimates of e(jijj) , X j (J ) an d \ / ( y(ja)) were obtained for each input 50. \\ cutoff frequency.

>"30' N-1 1 1 1 1 1 1 1 r — — r — I I X. ( juj) = I X (k) exp{-juk) (1) -10 hlO ^ ERROR (volts) k=0

Figure 3. Amplitude V-Error Func- where k denotes the kth time sample

tion for Left and Right Ears for in the ith block and for to = 0.0 39. the Amplitude-Matching Audio Test. L 0

JQj) . Analysis and 63^ (

x(t) : signal input The overall closed-loop frequency response as estimated between x(t) and

y(t) : system output y(t) was calculated as follows:

e(t) : error signal, x(t)-y(t) H(jto) = (2) 't'xx^J'^) Xj^ ( ju) : ith short-term Fourier estimator of xCju) at frequency to. and the human-operator frequency response (open loop) relating e (t) and H(ja)) : frequency-response esti- y(t) is given by: mate relating x(juj) to y(jco) at frequency u. ''xy (jo)) G(jto) = (3) G{ja)) : open- loop frequency- *xe ( ji^) response estimate relating e(juj) to y(jto) at where frequency to 30 = x^(jto) *x^(ja)) (4) ^xx^^"' ^ i) (u) : estimate of the auto- i=l spectrum of x(t) at frequency 00.

11 . , .

30 The parameters of interest are = ^ X. {ja))*y (joj) ?™(j'-') (5) T (the reaction time delay) , Tj-, (the operator's ability to react not only to the magnitude of the error but 30 also to the manner in which it is (6) changing) and c (a measure of the xe ^ ^ _ i=l operator's ability to follow the mean level of the signal accurately) where * denotes the complex conju- A Powell hill-climb routine (Fletcher gate. and Powell, 196 3) was implemented to model the responses obtained for G(ja)) and H(ju) in each test in terms Modelling the human operator of a human operator model described response . For each_test, mean esti- in equation (7) . The fixed low-pass mates of H(J(d) and G(ja)) were ob- element T^ was chosen to achieve a tained. These represented the aver- cutoff at about 5 Hz while Tg , the age performance over each ten-minute stability factor, was only allowed test. A linear model of the human to operate on high-frequency cutoff operator, simil ar to the more complex tests. This is consistent with ex- cross-over mode Is (Young, 1969; perience in that no stability prob- Krendel , McRuer and Graham, 1966), lems were encountered in low- frequency was used: tests. Thus the Powell hill-climb routine was allowed to optimize the

parameters K, t , T^ and c. The gain K e'^i"^ (l+j(^T^) factor K acts in choosing the magni- Gjj(juj) = tude of the human operator's open- (c+ju)) (1+juT^) (l+juT^) (7) loop characteristic response. The S J_i limits imposed on K are physical where K an arbitrary gain factor limitations and not of direct interest to his mode of response. The which can be varied over a large range by the operator. restrictions on the values assumed by T , Td and c were taken from previ- human operator's response- ous results, and from trial and error. time delay. The following limits were chosen: 0.1 K 20.0 a velocity-lead response ^ ^ 0.1 T reacting to the rate of ^ ^ 0.6 0.0 < change in the error signal. ^ c 1.0 The degree of velocity 0.0 < T,,^ 4.0 control exhibited is found and where applicable; to be dependent on the frequency content of the input. 0.0 < T .s 0.3 ^ s a measure of the integral action of the operator. Response evaluation . The concept c is small when the oper- of a measure of "goodness" for a ator is able to track slowly particular form of varying trends in the mean error signal presentation is a level of the error signal. subjective to degree. How- ever, for specific tasks such as continuous a low-pass element modeling tracking and control tasks two parameters may be the operator's ability to considered to disregard high-frequency give a measure of the performance of content in the error sig- one form of display against another. nal which may be due to his own reaction-time delay The ideal operator response is achieved when (t) follows x(t) ex- acting in a closed-loop y mode actly. Under these conditions the magnitude of the closed- loop frequency a fixed low-pass element response is unity over all frequencies of interest. measuring the inability of Deviations from this the muscles to respond to unity gain represent a loss. Thus a factor given by: frequency changes (T = 0.04, cutoff about 4 Hz).

12 : ,

'MAX

C (juj) (10) xy -^ FACT H(jtu) 1-1.0 (8) 'XX yy MAX

at frequency where v< c n< 1. gives a number by which the gain response is characterized. ^mAX ^^ chosen as some maximum frequency of ''Then the coherency equals one interest for which the display is to the output can be defined exactly as be used. Each operator's response can a linear function of the input. VThen is thus be characterized by a factor the coherencj' is zero the output uncorrelated with the input. GfACT- The greater G^, the greater the overall deviation from the ideal Figures 4 and 5 show the mean- response. coherency sauares results measured for two operators. The mean coherency The amplitude of the overall re- was taken over the frequency range sponse does not completely character- zero cutoff for each test, and the ize the response. Phase shifts in the to results for each type of error-signal response can also be detrimental to presentation and for each input cut- the performance of a particular form off freouency are displayed. Fig- of error signal presentation. The 4 was tvpical of results obtained ideal phase characteristics are again ure ^^or four of the five operators tested. zero-phase shift over all frequencies The Visual Test showed a high level of interest. Thus a phase "goodness" of coherency over the complete range parameter is proposed: of inputs used. The spread being 0.6 ^ c2 ^ 0.95. ,F typically MAX

IO-| < 1 1 1 epj(j(^) ^^li (9) 1 1 1 FACT MAX -J 1 o^^2^ * J o 75- : where 6 (joi) is the phase shift, in 4

radians, of the closed-loop response u c H(j(j) is chosen as before. « MAX 2 05- o Thus each operator's response to u \ a particular form of error-signal two presentation is characterized by 025- "goodness" parameters. The display V.I, ^ 1 \;_ which minimizes these two parameters V' 1 1 ' can be judged to give the best re- III' sponse. However, for individual ap- 1 1 1 1 V Al A2 A3 A4 plications , the weight placed on the TYPE OF TEST amplitude factor as opposed to the phase factor is a matter of choice. Figure 4. Mean-Coherency Level Obtained for a Single Operator for Each Form of Display and RESULTS AND DISCUSSION at Each Input Cutoff Frequency.

In assessing the operator's per- (The Mean Level is Taken Over formance the coherency between the the Frequency- Range O Cutoff) input and the operator's output presents a measure of the extent to which Operator No. 2 his output (at a given frequency) resulted as a direct response (albeit V - Visual perhaps a wrong one) to the input. It Al - Mark Space Matching is defined: A2 - Matching Frequency A3 - Doppler Matching Frequency A4 - Amplitude Matching

13 . .

I Oi 1 explained 1 by the hearing 1 defect shown 1

1 1 1 in the operator's audiogram (Fig.

1 1 i K 6) . However the case is made 1 , 1 1 that an audio display must be tailored 1 075 1 v\ to suit the individual. 1 !\\\ : 1 /f

OS V-X^X. 1 Closed- loop frequency response .

1 i v\\>-- -Yj' Figure 7 shows the closed-loop fre- i lO/ 1 quency response results as measured 1 ' 1 1 1 ' at each input cutoff frequency for 2S - -- -f07 ingut _^ -^ each type of error signal 1 %0 '"P*^ presentation. The results are shown for a 1 nsj-?i'-^ single operator. 4r" OO 1 — 1 1 V Al A2 A3 A4 At low-input cutoff frequencies TYPE OF TEST the closed-loop gain response is approximately flat and approaching unity while the phase response shows the Figure 5. Mean-Coherency Level phase shift slowly increasing with Obtained for a Single Operator frequency. The ideal response is de- for Each Form of Display and fined as unity gain over all frequen- at Each Input Cutoff Frequency. cies with zero-phase shift. As the input cutoff frequency increases the (The Mean Level is Taken Over gain response decreases sharply, the Frequency-Range Cutoff) with unity gain only occurring at very low frequencies. The rate of increase in Operator No. 2: phase shift is greater than expected by V - Visual extrapolation of the low-frequency results. Thus, in Al - Mark Space Matching general, the operator changes his mode of A2 - Matching Frequency operation as the input cutoff frequency in- A3 - Doppler Matching Frequency creases . In some cases , the inherent A4 - Amplitude Matching time delay in the operator's response leads to a resonant peak in the closed-loop response. The results for the Mark Space Test (Al) show a low level of co- In measuring the operator's re- herency on all tests As the input sponse it must be recognized that cutoff frequency inc reases the oper- his response will change as his con- ator is less able to track correctly centration and skill vary. For this and the coherency fa lis. Indeed it reason, untrained operators were used is doubtful whether the very high- to try to nullify any skill discrep- frequency tests (1.0 - and 1.5-Hz ancies in the results. Possibly a cutoff) can be tracked at all using the mark-space ratio feedback

I Ul The Matching Frequency Test (A2) ^» I!— shows an overall improvement on the dB Hh-- 1! 1 1 Mark above Space Test, while Beat Frequency I Threshold (I— 1 Feedback (A3) is probably the best display used over the frequency ranges of int3rest. It must be noted that although Amplitude Matching (A4) performed well for low- frequency inputs

(slightly better even than A3) , the lOO operators found greater difficulty no in matching amplitudes quickly than 250 500 lOOO 2000 <000 8000 in matching frequencies Frequency Hz Figure 6. Results of an Audiometer Figure 5 shows the exception to Test the rule. For this operator the on Operator No. 2. Amplitude Matching Tests results were consistently better than either fre- Left ear X quency test. This is possibly Right ear O

14 Frequency Hz a)0-3 CUT-OFF 2 5

-90- Frequency Hz o 65 1-3 d) 1 CUT-OFF 0-1 Vc >. ^w^^:^^^ C ,1 -I80 o E 25 i Frequency Hz 90-J ^"C-"<"^^ ^^^ ^-C: Frequency Hz T7 kr-^/^^ b) 05 CUT-OFF 35 o

"^ s

fin - 65 13 O Frequency Hz

5- Frequency Hz

1 5 el CUT-OFF 5 B5 17 0-| •».,,_--.

c 1 "^^^ 5 \ X ^~^^ 35 Frequency Hz 5- ^ '- ^—jl^^ r^ -^ Frequency Hz I \ ^ c) 07 CUT-OFF ^ o 5 lO "^ X \ o- o ^-^ ^ ^v ^^^f^^:,—:.^.^ ^ \ '-_^. c " ^^"^^""""^^^^ -ISO Q» ^ "^ 85 D ^ A" 2 ^^^ -^ ^^^^'^"~~~\-- \ Frequency Hz 5- 90- \^^_— — '2~^ '^ «» T7 w ^^

1 N. a.

eo-l

Frequency Hz

Figure 7. Measured Closed-Loop Frequency Response at Each Input Cutoff Frequency for Each Visual and Audio Error Signal Presentation. Visual (V) Matching Frequency (A2) .-.-.- Mark Space (Al) Doppler Matching (A3) Amplitude Matching {A4)

15 . . : ) fairer test on which to base conclu- taking into account the physical sions on the acceptability of any limitations imposed by the human op- particular form of auditory display erator. The error function mini- would be conducted with trained oper- mized was of the integral error ators. However, the authors believe squared form given by: that initial reactions are as true a guide as any. There is no quantitative way to allow for concentration lapses

Figure 8 shows the range of results obtained for one particular COST s test (beat-audio error-signal presentation with 0.7 Hz input cutoff frequency) for all five operators. The result for Operator 2 shows a resonant peak in the gain response at about 0.8 Hz while the phase shift at low r^MAX,„ ,. , „ /• ^^2. frequencies tends to be : — (unlike the other operators). ^ This type of response occurs when the operator \f (11) rMAX,„ ,. .,2, is over-reacting to changes in the input signal and subsequently over- correcting. Again the variations in - the responses measured would tend to where advocate tailoring the. audio display • the individual. to measured ".MEAS^^i"' closed-loop response

Human operator modeling . The H model used (equation 7) has the ad- MOD ( j u model closed- loop vantage that the parameters can be response given a physical interpretation. The Powell hill-climbing technique used measured open-loop ^MEAS^^"' to minimize a cost function between response measured and modeled response values '' found the global minimum in the cost S0D<>^ model open-loop function over the parameter N-vector response space and within the boundary values and: chosen. The choice of boundary values for the parameters was based on previ- HmOD^^-) = + (12) ous values used and on experience 1 G^,Qj^(j.)

Frequency Hz 033 066 ID

^^^~^>- ~^''"'~~-~-_~ ^ . ID \-.-'^ ^ --^~- ^~~^^~~^^ ^'^ \\ Oi ~^^ ^^^-^ ^ i Q .v\ 1 '^-^^r^^^ aS-^t- ^~?^-~^^ - 90 '. ^^-^ :; '^^ OS ' \\ • — —' — ^' e \ \ < c ~--'" ~- a. 5

GO leo

O 33 066 I O Frequency Hz

Figure 8. Results Obtained for the Doppler Matching-Audio Feedback with an Input Cutoff Frequency of 0.7 Hz.

16 33-2 O

1 ISO

o oo Frequency H2 O 5 |o

-ISO X '5'«G>«»(»aaBisi(j®^^^^^^^ o

OO Frequency Hz OS

Figure 9. Measured and Modeled Frequency Response, Open-Loop, for a Human Operator Subjected to Tracking a Band-Limited White-Noise Input with 0.3-Hz Cutoff. (The Error Was Presented Visually.)

This cost function and hill-climb input cutoff frequency increases, the process thus optimizes the model par- gain decreases. This is borne out by ameters in terms of both open-loop Table 1 for all the various forms of and closed-loop measurements simul- error-signal presentation. It is al- taneously. Figures 9 and 10 show so noticeable that the mark-space typical results obtained from this presentation (Al) displays the lowest modeling process for a visual test. overall gain figures which is compatible with predictions made from the Table 1 lists the parameter val- coherency results. The be at- frequency ues obtained from modeling the tests presentation (A4) comes closest to conducted on a single operator. A the figures presented for the visual complete table for all five operators presentation; this again might be ex- is not included for reasons of space. pected from the coherency results. However, several points can be inferred from the results presented The lead factor (T^) is a measure which are borne out in the results of the velocity control exerted by the obtained for the other operators. operator; the larger the value of Tp the greater the velocity control ex- The human-operator gain (K) is erted (the more sensitive the response a function of the sensitivity of the to rate of change in the error func- joystick and of the particular test tion) . environment chosen. However, as the

i22n

nq6 «»<8Ba ®®« >d»6®»»i*S®o®9oooX X

Frequency HZ OS 0-Ofl>i

oo CO Frequency HZ 0-4

Figure 10, Measured and Modeled Frequency Response, Closed-Loop, for Human Operator Subjected to Tracking a Band-Limited White-Noise Input with 0.3-Hz Cutoff. (The Error Was Presented Visually.)

17 TABLE 1

Operator Model Paraitieter Values Plus Error Performance Figures for Single Operator

TEST Input cutoff Type of frequency P Display Hz K T C ^S COST Sact FACT

Visual 0.3 2.1 0.40 0.15 2.59 0.0 0.021 0.08 10.6 0.5 0.94 0.33 0.17 2.26 0.0 0.02 0.20 18.0 0.7 0.78 0.21 0.22 1.96 0.0 0.02 0.30 24.0 1.0 0.43 0.18 0.33 1.77 0.0 0.013 0.46 33.7 1.5 0.40 0.10 0.57 1.45 0.02 0.025 0.57 42.0

Mark, space 0.3 0.41 0.3 0.10 1.30 0.004 0.02 0.54 38.0 0.5 0.27 0.24 0.16 1.41 0.08 0.017 0.69 45.4 0.7 0.27 0.23 0.16 0.87 0.06 0.04 0.72 72.2 1.0 0.26 0.23 0.4 0.72 0.10 0.023 0.77 92.2 1.5 0.33 0.18 0.70 0.35 0.09 0.07 0.86 108.7

Matching Freq. 0.3 1.02 0.27 0.11 1.17 0.001 0.013 0.32 23.1 0.5 0.79 0.17 0.25 1.24 0.11 0.015 0.40 25.5 0.7 0.36 0.16 0.20 1.31 0.07 0.009 0.61 46.5 1.0 0.19 0.20 0.0 1.33 0.0 0.024 0.75 63.6 1.5 0.21 0.19 0.08 0.67 0.004 0.035 0.83 87.8

Beat 0.3 1.52 0.40 0.0 1.46 0.003 0.024 0.14 18.1

- 0.5 1.42 0.21 0.32 1.01 0.16 0.016 0.22 28.7 0.7 0.97 0.12 0.25 1.02 0.20 0.02 0.37 48.4 1.0 0.53 0.12 0.60 1.09 0.16 0.017 0.58 59.8 1.5 0.28 0.15 0.13 1.05 0.154 0.02 0.75 100.6

Amplitude 0.3 1.22 0.39 0.26 1.16 0.002 0.022 0.25 23.0 0.5 1.25 0.15 0.49 0.99 0.2 0.009 0.33 29.0 0.7 0.59 0.13 0.64 1.15 0.189 0.02 0.55 48.5 1.0 0.28 0.23 0.48 1.2 0.0 0.06 0.67 60.0 1.5 0.10 0.18 0.25 1.9 0.05 0.04 0.81 80.2

18 . , .

The operator appears to maintain increase in time delay was found at an approximately constant level of the high frequencies for the audio velocity control until at high fre- tests. This might be explained by quencies he no longer responds to the fact that at high frequencies velocity, but degenerates into a form signal interpretation becomes more of bang-bang control. Under this complicated for the audio tests. last mode of operation it is purely This suggestion is only tentative the sign of the error that decides and further tests are required. the controlling action. For some audio tests, notably the Mark Space The lag constant (Tg) is perhaps Test, this degeneration into a form the most difficult to explain in physi- of bang-bang control occurs at input cal terms. Due to the operator's time- cutoff frequencies as low as 1 Hz and delay response, the closed-loop re- possibly even lower. sponse has a tendency to exhibit resonances at frequencies where the phase The operator's time delay (x) shift in the output leads to positive decreases as the input cutoff fre- feedback. Thus, the operator can find quency increases. Fig. 11. This re- himself correcting high-frequency com- sult is well known. However, a slight ponents in the error signal which are not in fact present in the original input. All the operators tended to counteract this by ignoring, to a greater or lesser extent, high- frequency changes in the error signal. The model, as given by Equation 7, allows for this in the form of a low- pass element with a time constant Tg

The greater Tg , the lower the break frequency for the low-pass element. A.11 the tests display low values of Tg for the 0.3-Hz cutoff with Tg increas-

ing for the . 5-Hz and . 7-Hz cutoff frequencies. Beyond 0.7 Hz it depends on the test as to the degree of damp- ing used. There appears to be a mode of operation for some of the audio tests in which the closed-loop resonance effect goes unnoticed and hence uncontrolled. This may be a function of the information being lost in the audio feedback at high frequencies This is certainly true of the amplitude-matching and be at- frequency presentations

Lastly the integrator factor (c) attempts to measure the ability of the operator to track low- frequency trends in the mean level of the sig- OS 7 10 nal. Surprisingly the Mark Space INPUT CUTOFF FREQUENCY Tests shov; up quite well in this respect, being similar to the visual Figure 11. Variation of Three Model response at low frequencies. However, Parameters with Change in Input the variations in this constant are Cutoff Frequency for Each Form of more a function of the operator's Error Signal Presentation and for long term concentration and hence a Single Operator. little can be said.

Visual It must be emphasized that the discussion of changes in the model Mark Space parameters is only applicable to Matching Frequency -.-.-.-.-.-. these particular tests. Firmer con- require much more intensive Doppler clusions testing. However, the trends listed Amplitude Matching

19 were consistently reproduced in tests Contrary to comments made by Vinje conducted for a second time on a and Pitkin (1971), the human oper- single operator, and l^nd weight to ator response to the audio cues the applicability of the model. The presented in these experiments was model-cost function values listed in consistently inferior to visual Table 1 give further reason for be- cues. However, audio presentations lieving the model to be creditable. used here were different from those used by the above authors.

Display performance estimates . The linear model used to describe Although a detailed study of the the operator's response fitted the measured operator responses for vari- measured response well for low- ous types of error presentation is frequency input tests. However, instructive, it is too cumbersome a when the input signal contained high- method for classifying particular frequency components, operator re- forms of display. There are two ways sponse might be better described by of classifying an operator's response the incorporation of a non-linear for a given input in terms of modulus bang-bang control element in the gain response and phase response. model. For this reason the physi- Equation 8 defines a factor Gp^j-.,], cal interpretation of the model which gives a measure of the mean- parameters is still open to argument. square deviation from unity gain The description presented here is over the frequency range of interest, subjectively consistent with the while Equation 9 describes a similar experience of the subjects tested. factor (PpACT^ assessing the deviation of the phase response from a The presentation of a gain-and- desired zero-phase shift. These phase cost function to m.easure the factors are listed in Table 1 for "goodness" of operator response has each test. The weight placed on the advantage of retaining informa- opposed to depends on tion about phase response in its ^FACT ^^ Pp^f^rp the type of task to be performed. crudest form while retaining overall information on the modulus of the

' response. A simple presentation of Ideal values are: v' error variance to input variance 0-° gives no useful information about Sact = the nature of the response.

^FACT = °-0 Finally, this article serves as an introduction to a more thorough It can be seen in Table 1 that the visual display scores best on analysis of two- and three-dimensional audio displays as an alterna- both counts . The Beat Frequency Test tive, or in has an edge on the Amplitude Matching conjunction with visual Test, followed by the Frequency Match- displays. Extension to two-dimen- ing Test. The Mark Space Test scores sional displays introduces cross- lowest. This ordering agrees sub- coupling effects which, although stantially with other results pre- not too serious for visual tracking sented and confirms that G„^^^ and tasks , could prove detrimental to performance for certain combinations P „ would appear to give a reliable of audio displays. perf orrr.ance index.

CONCLUSIONS

Experiments showed that human operator performance depends on the frequency content of the input. It would appear that for visual-tracking tasks the operator can adapt his response to suit the input. However, it v/as found that his adaptability, and hence performance, was impaired by audio presentation of the signal.

20 , ,

REFERENCES

Black, W. L. An acoustic pattern Mudd, S. A. Experimental evaluation presentation. Research Bulletin of binary pure-tone auditory dis- 16. New York: American Founda- plays. Journal of Applied Psy- tion for the Blind, May 1968, ohology, 1965, 49(2), 112-21. 93-132. Phillips, J. A. & Seligman, P. N. Blackman, R. B. & Tukey , J. W. The Two instruments for the blind measurement of power spe ctra. engineer. Research Bulletin 27, Dover Publications, 1958. New York: American Foundation for the Blind, April 1974, 187-216. Fish, R. M. & Beschle, R. G. An

auditory display capable of Pollack, I. & Ficks , L. Information presenting two dimensional shapes of elementary multi-dimensional to the blind. Research Bulletin auditory displays. Journal of the 26. New York: American Founda- Acoustical Society of America

tion for the Blind, June 1973, 1954, 26 (2) , 155-58. 5-18. Roffler, S. K. & Butler, R. A. Fac- Fletcher, R. & Powell, M. J. D. A tors that influence the localiza- rapidly convergent descent method tion of sound in the vertical plane. for minimisation. British Com- Journal of the Acoustical Society puter Journal , 1963, 6, 163-68. of America, 1968(a), 43(6), 1255-9.

Roffler, S. K. & Butler, R. A. Lo- Gill, J. M. Auditory and tactual displays for sensory aids for calization of tonal stimuli in the the visually impaired. Research vertical plane. Journal of the Acous tical Society America Bulletin 29. Nev/ York: American of 1968(b) 1260-6. Foundation for the Blind, 1975. , 43(6) , Vinje, E. W. & Pitkin, E. T. Human Kramer, H. J. Stimulus variables in operator dynamics for aural com- auditory projective testing. pensatory tracking. 7th Annual Research Bulletin 1. New York: Conference on Manual Control American Foundation for the Blind, NASA SP-281, June 1971, 339-48. January 1962, 33-40. Wellstead, P. E. Aspects of real time spectral analysis. Ph.D. Thesis, Krendel, E. S., McRuer, D. T., & University of Warwick, England, 1971. Graham D. The analysis and synthesis of manual closed loop Young, L. P. On adaptive manual con-

control systems. 196 6 IEEE Inter- trol. IEEE MMS-10 , December 1969,

national Convention Rea. , 193-95. 292-331.

A PILOT STUDY ON THE DISCRIMINABILITY OF TACTILE AREAL AND LINE SYMBOLS FOR THE BLIND

G. A. James* andJ. M. Gill**

Abstract : Eight tactile areal symbols for areas, and point symbols and 17 linear symbols for use on rraps to show specific locations or land- and graphics for the blind were pro- marks. This study is concerned only duced on Brailon and tested for dis- with areal and line symbols. criminability in separate sets by the method of paired comparisons. Sub- Several studies have attempted jects ' response times were recorded to define sets of discriminable tac- as latencies. The results indicated tile areal and line symbols for the that only 5 of the areal symbols but blind. Heath (1958) conducted a 10 of the linear symbols met the pioneer study by examining the dis- stringent criteria for discrimina- criminability of 40 tactile areal bility suggested by Nolan and Morris symbols using the method of constant

(1971) . Errors in discrimination are stimulus differences to compare sym- discussed with reference to the param- bols randomly grouped in sets of 10 eters which contribute to the dis- He also found that areal symbols re- crim.inability of the symbols used in mained legible at a size of 50 x the study, and latencies are discussed 50 mm. Culbert and Stellwagen (1963) in relation to "response set." also examined the discriminability of textural surfaces and found 11 out of 40 different patterns discriminable enough from all the others to be INTRODUCTION useful in the preparation of material such as maps for the blind. Nolan It has been shown that there is and Morris (1971) conducted several a need for tactile maps and diagrams studies which represent the most ex- for blind schoolchildren. Leonard tensive source of information. Their and Newman (196 7 and 19 70) demonstrated findings show that the number of tac- that at least half of the subjects in tile areal or line symbols which are a study were able to complete an un- discriminable in a set may not exceed familiar route with the aid of a tac- 8 or ten. They relate this perceptual tile map to provide the relevant in- limit to the parameters which dis- formation. tinguish tactile symbols. A flexible production system is therefore an es- Tactile maps and diagrams are sential requirement in varying these composed of three categories of sym- parameters as much as possible in an bols: line symbols to designate bound- attempt to increase the number of

aries or lines , areal or texture legible tactile symbols within a set.

*Blind Mobility Research Unit, Uni- PRODUCTION METHOD versity of Nottingham|, England. **Inter-University Institute of En- The study conducted by Heath gineering Control, University of (1958) used the Virkotype or Gestetner Warwick, England. printing method. Wet ink print was

23 , :

dusted with a fine resinous powder METHOD which adhered to the wet ink and appeared as a raised plas'tic symbol Subjects when heated. The disadA^antages of this method were stated by Nolan and Sixty-two blind schoolboys were Morris (19 71) : the degree of relief used as subjects. The age range was is poor (0.11 mm.), control of qual- from 11 years 3 months to 19 years ity is poor, and the medium deteri- 1 month. This sample represented orates in humid conditions. all braille readers who were available and in full-time education at The production method used in Worcester College for the Blind. the Nolan and Morris studies in- I.Q. scores, chronological ages, and volved reproducing the symbols to braille reading speeds were obtained be studied by photoengraving in zinc. from the school. They assessed The master was then pressed into soft braille-reading speed in the follow- plaster which was then allowed to ing way harden. The moulds were then used as masters to produce vacuum-formed 1. Boys read braille out loud to copies in plastic 0.20 mm. thick. the whole class for 3 minutes. Embossed symbols were produced at a relief varying from 0.46 mm. to 2. A score was taken for the num- 0. 62 mm. ber of braille-lines completed.

In this country a variety of 3. The number of lines completed production methods have been investi- was then multiplied by 3/4 to gated by Pickles (1970). Briefly, give an average speed in pages this type of approach involves build- of braille per hour. ing up a master map or diagram on transparent cellulose. Various thick- nesses of string and wire are used Apparatus and Selection for line symbols; sandpapers, linole- of Symbols um, and fabrics are used for textures. The master can then be used to pro- Figures 1 and 2 show the appa- duce copies in Brailon on a Thermo- ratus. A wooden board with a frame

form machine. " was used to hold the stimulus cards. . >i^ .''<-: Some previous studies have used a The production methods briefly blindfold to exclude residual vision described are generally time-consum- of some blind subjects but this may ing and therefore expensive if the introduce psychological stress. cost of labor is taken into account. Recent developments at the University of Warwick are based on computer- aided design principles. The relief and type of line or texture is input to a computer from a keyboard. Sym- bol parameters can be varied accurately to include various heights of solid, dotted, dashed, and dot-dashed lines. Symbol specifications are stored by the computer. Once the symbols have been specified the master is engraved in a sheet of Tufnol by a computer-controlled machine tool. A positive copy of the engraved master is made using silicone rubber. Copies are produced in 0.18 mm. Brail- on on a Thermoform machine.

This study is an initial attempt to define some of the parameters governing the discriminability of areal and line symbols produced by a computer-controlled method. Figure 1. Experimental Apparatus,

24 Figure 2. Experimental Apparatus. Figure 3. Position of Subject's Hands for Examination of Areal Symbols. Therefore, a screen with a curtained opening for the subjects ' arms was used. The stimulus cards were con- The 8 areal symbols gave 36 combina- tained in a filing tray. A stopwatch tions , and the 17 line symbols gave was used to record response times. 153 combinations. Three sample pairs of symbols were used to familiarize Selection of tactile symbols for the subject with the procedure. testing was guided by previous research. Areal symbols were varied The order of presentation of the along the dimensions of continuous paired symbols was determined random- and interrupted, density of the pat- ly. tern size of the figures making up the pattern, and the use of vertical, horizontal, and diagonal lines to PROCEDURE differentiate patterns. Linear symbols were continuous and interrupted, Two examiners tested the sub- thick and thin, single and double, jects using the following Standard and smooth edge and broken edge. The interrupted lines were varied with more than one spacing.

Areal and linear symbols were produced in Brailon. The areal ones were 50 x 50 mm. in size and the linear ones were 100 mm. in length. Areal and linear symbols were tested in separate sets. Figures 3 and 4 show how the former were mounted side by side and the latter one above the other on stiff card 120 x 100 mm. The left/right, or up/down position of the symbols was determined randomly. The relief of the tactile symbols was 0.7 mm.

DESIGN

Symbols within each separate set were compared by means of paired- Figure 4. Position of Subject's comparison: each symbol in a set was Hands for Examination of compared with itself and every other. Linear Symbols.

25 )

Instructions : made the decision. To prevent knowledge of results only one stroke of Textures the pen was made by the examiner in a "right" or "wrong" column on the Please put both of your hands scoring sheet. through the curtain onto the raised symbols in front of you. You will Total time taken to complete the find two syinbols side by side. test was about 40 minutes. This included three 60-second rest periods. Are these symbols the same or different? (E gives knowledge of results . CRITERIA

There is no time limit, but re- Jenkins (19 47) used the method • member that once you have made a de- of paired comparisons to define a dis- cision you cannot change your mind. criminable set of tactile aircraft Give the answer "same" or "differ- control levers. He excluded any ent." shapes confused by more than one percent of the subjects. The effects of There will be two more test sym- making an incorrect decision with bols to be sure you understand. I aircraft controls are evident and will tell you if you are right or justify the extremely stringent cri- wrong, but this time do not touch the teria. symbols until I say "Now." Nolan and Morris (1971) report Are there any questions? (Ques- the following criteria as being the tions are dealt with by repeating the most useful in selecting discriminable relevant parts of the instructions.) tactile symbols for the blind:

We will now begin the experiment. 1. Averaae confusion with other ac- I am not able to tell you if you are ceptable symbols must be five right or wrong from now on. Remember percent or less. not to touch the symbols until I say "Now." 2. Confusion with itself or any other single symbol acceptable by criterion 1. should be Lines 10 percent or less.

This time the two symbols are 3. Any symbols acceptable by cri- lines , and they are one above the terion 1. and 2. must be inde- other. Concentrate on the center of pendent of academic grade dif- the lines and not the ends. First ferences . there will be three test symbols to be sure you know what to do. (E Nolan and Morris' criteria are gives knowledge of results.) not supported by any rationale, but as quite stringent arbitrary criteria Are there any questions? 1. and 2. were adopted for the purpose of this study. Criterion 3. was We will now begin the experiment. not adopted because it was considered

I am not able to tell you if you are that IQ , chronological age, and braille right or wrong from now on. Remember reading speed were more reliable vari- not to touch the symbols until I say ables than "grades." "Now."

This experiment is much longer RESULTS than the first one, so there will be two short breaks of one minute each. For the purpose of analysis, braille reading-speed scores were classed into frequencies as shown in Each subject examined every pair Fig. 5. On the basis of these data of symbols and had to report whether the experimental group was divided they were the "same" or "different." into low-_medium- and high-speed The examiner recorded the time to the braille readers. Subgroups were nearest second from when the subject comprised of 16, 26, and 20 subjects touched the stimulus card to when he respectively. Tables 1 and 2 show

26 . .

and for linear symbols, H was 2.03, values too low to be significant at MEDIUM the 0.05 level.

No correlations were found between chronological age and performance or IQ and performance for areal or linear symbols. n. Tables 3 and 4 show the percentage of errors for areal and linear symbols. (Areal symbols are indicated by upper case letters and linear case.) After exclud- I I I r lower I r- r n symbols by 2B ing B, D, and H the remaining areal BR4ILLE READING SPEED symbols were A, C, E, F, and G; these indicated by an asterisk in Fig. 6. Figure 5. Braille Reading-Speed Frequencies k/vVv\A/l that the mean error was no more than 2 on the areal symbols and no more sAA/Wv" than 5 on the linear ones vA/vWV TABLE 1 \AAAAA/ A* Correct Responses Within Braille Reading-Speed Groups for 36 Combinations of 8 Areal Symbols

Low Medium High Total

Mean 35.12 34.53 34.75 34.80 Range 31-36 28-36 31-36 28-36 ++++++ N. 16 26 20 62

TABLE 2

Correct Responses Within Draille Reading-Speed Groups for 153 Combinations of 17 Line Symbols

F* Low Medium High Total

Mean 148.25 148.07 149.70 148.86

Range 141-152 131-153 143-153 131-153

N. 16 26 20 62

1 LI Krushal-Wallis one-way analyses II 1 1 1 1 1 1 1 1 M of variance for braille reading-speed groups and performance were computed linear sym- separately for areal and Fiaure 6. Outside Dimensions of Eight bols. For areal symbols, H was 1.00 Plastic Areal Symbols, 50 mm. Square. Asterisks Identify a Discriminable Set,

27 TABLE 3

Percentage of Errors on Areal Symbols

A B C D E F G H

A 4.8 1.6 3.2 1.6 1.6 4.8

B 11.2 1.6 20.9 3.2

C 9 .6 1.6 3.2 1.6

D 14.5 1.6

E 1.6 1.6 3.2 1.6

F 4.8 3.2

G 8.0

H 11.2

TABLE 4

Percentage of Errors in Line Symbols

1 a 1.6 4.8 1.6 1.6 6.4 1.6 b 3.2 3.2 1.6 3.2 c 6.4 6.4 1.6 d 3.2 1.6 1.6 1.6 e 8.0 62.9 3.2 17.7 1.6 6.4 1.0 f 6.4 1.6 11.2 1.6 1.6 3.2 4.8 1.6 3.2 g 4.8 1.6 3.2 h 20.9 1.6 1.6 4.8 i 3.2 33.8 3.2 1.6 1.6

J 3.2 1.6 k 11.2 14.5

1 4.8 4.8 30.6 1.6 m 17.7 9.6 n 12.9 o 17.7

P 4.8 q 8.0

28 After excluding m, h, k, n, p, j, and evidence to the theory that there may e the remaining linear symbols were be an inherent limitation in the va- a, b, c, d, f, g, i, 1, o, and q; riety of tactual discriminations a these indicated by an asterisk in person can make on symbols of this Fig. 7. kind. Alternatively, there may be limitations in the experimental de- Mean latencies for areal and sign and it is hoped in future re- linear symbols are shown in Tables 5 search to investigate this problem. and 6 respectively. Latency differ- The distinguishing parameters for ences between like and different pairs linear symbols are evident from of symbols were assessed for areal and Fig. 7. For interrupted lines, spac- linear symbols separately. The standing is a distinguishing parameter ard mean latency for different areal for dotted lines (c, d) but not for symbols was 2.9 4 and for like pairs dashed lines (e, f ) . Lines with was 5.78, and for linear symbols the edges broken by vertical projections corresponding figures were 2.45 and (k, m and n) are easily confused and 5.23. To give the significance of the use of projecting lines of dif- the latency differences for like and fering angles might be useful. different symbols the Mann-Whitney U- test was applied and it was found Areal symbols had a limited that the differences for both areal range, but we confirmed Nolan and and linear symbols were significant Morris ' finding that if the areal at less than the 0.001 level. pattern is basically similar, as in B and D, change of direction on di- agonals is not a good cue for dis- DISCUSSION crimination. This is a cognitive problem and might be solved by in- The study was successful in introducing perceptual training. creasing the number of discriminable tactile linear symbols from the 8 One self-error in the areal sym- found by Nolan and Morris (1971) to bols and four in the linear ones de- 10. However, this does not exceed tracted from the number of legible the upper limit of 10 suggested by areal and line symbols. Had it not

Nolan and Morris , and adds further been for these errors , six out of eight areal symbols would have been discriminable and 12 out of 17 lines. One explanation for self-errors is that subjects may be examining the HHHHHHHHHH symbols too closely for subtle differences which do not exist, alternatively subjects have a response set for saying "different" when in actual fact they mean "same."

d" '- ' Results for latencies oppose the I I I I I "mental set" explanation for like-pair errors. Subjects spent significantly more time discriminatina like-pairs of symbols as compared with different. Although one subject did remark "It 0* I I I I I I I I I I I I I I I I I I II becomes mechanical after a while," the evidence shows that subjects did not continue answering "different" p llllllllllllllllllllllllllllll when the symbols were the same.

A criticism of this study is that i' + + + + + + +i- the symbols were presented in the same random order to each subject. In view of the length of the test (40 minutes) practice and fatigue could have been compensated for by alternating the order of presentation. A further Figure 7. Seventeen Linear Plastic criticism is that time-keeping by Symbols, 100 mm. Long. Asterisks stopwatch was both tiring for the Identify a Discriminable Set. experimenters and inaccurate, and

29 TABLE 5

Mean Latencies for Areal Symbols (Seconds)

A B C D

A 5.89 3.22 3.45 3.09 2.32 2.48 3.08 2.22

B 6.42 2.87 6.14 2.11 2.84 3.08 1.87

C 9.13 3.05 3.08 2.00 3.30 2.88

D 6.44 2.11 2.47 2.72 2.72

E 3.26 2.90 4.19 2.56

F 4.40 3.29 3.45

' G 5.42 2.98

H 5.32

TABLE 6 jfqhiiklmMean Latencies for Line Symbols (Seconds) a 3..01 2.82 1.90 1..87 2 .20 1,.96 2..45 2.00 2.22 3.06 2.31 1.85 2.68 2.52 2.03 1.61 2.05

b 3.62 2.08 1.,98 2,.23 1,.96 2., 83 2.25 2.14 2.56 2.12 1.93 2.22 2.24 1.88 2.12 1.90 c 4.79 3..16 2 .00 2..54 2,.46 2.35 1.67 1.83 2.12 2.09 1.96 2.42 2.04 2.09 2.27 d 6..54 2..09 2,.37 2..13 2.85 1.81 1.61 1.96 1.72 2.09 2.45 2.05 2.29 2.29

',- ' 4 6, 3,.19 2.32 2.17 2.09 2.14 2.62 3.93 2.17 2.08 2.24 e ' .08 .35 3.83 f 6,.59 2,.41 4.09 1.98 2.87 2.66 2.00 3.09 3.82 2.03 1.95 2.30 g 5,.87 2.87 2.45 2.08 1.96 1.93 2.52 2.98 2.16 2.46 2.25 h 6.45 2.13 1.90 2.04 2.61 3.20 3.22 2.48 2.00 2.17 i J - • 3.24 4.30 2.04 2.00 2.45 2.55 2.48 2.19 2.03 1.96 2.37 2.17 2.08 J 4.00 2.16 2.01 2.50 k 6.66 1.95 3.72 2.61 2.16 2.48 4.55

1 4.80 2.35 2.43 3.01 4.87 2.04 m 7.67 4.96 2.66 2.03 3.06 n 4.46 2.25 1.88 3.34 o 6.70 4.62 2.95 2.22 P 5.70 4.79 q

30 . . .

more sophisticated timing would be Immediate research includes useful in future work. the assessment of discriminable tactile point symbols, including Future work on areal and linear upper-case letters of the English symbols should include a more system- alphabet, and an excimination of atic analysis of the parameters which the usefulness of this type of

contribute to discriminability , and a tactile code for school-children consideration of the effect of varia- and adults who are braille and tion in symbol relief to increase in- non-braille readers. formation redundancy.

ACKNOWLEDGEMENTS

The joint authors of this paper Professor J. L. Douce (I .U. I .E.C. ) have carried out this work with the The authors are further indebted for financial assistance of the Medical the advice and suggestions of Dr. Research Council and the Science M. J. Tobin (Director of The Research Research Council. The authors wish Centre for the Visually Handicapped, to thank their supervisors Dr. Birmingham) and to the Headmaster and J. D. Armstrong (B.M.R.U.), and pupils of Worcester College for the Blind, England.

REFERENCES

Culbert, S. S. & Stellwagen, W. T. Leonard, J. A. & Newman, R. C. Three Tactual discrimination of textures. types of "maps" for blind travel.

Perceptual and Motor Skills, 1963, Ergonomics , 1970, 13, 165-79. 16, 545-52. Nolan, C. Y. & Morris, J. E. Improve- Heath, W. R. Maps and graphics for ment of tactual symbols for blind the blind: Some aspects of the dis- children. Final Report, American criminability of textural surfaces Printing House for the Blind, 1971. for use in areal differentiation Ph.D. thesis. University of Wash- Pickles, W. J. Raised diagrams pp. ington, 1958. 118-45 in The teaching of science and mathematics to the blind. Jenkins, W. 0. The tactual discrimina- London: The Royal National Insti- tion of shapes for coding aircraft- tute for the Blind, 1970.

type controls. In Fitts , P. M.

(Ed.) : Psy chologi cal research on

equipment design. Washington, D.C.; Siegel , S. Nonparametria statistics U.S. Govt. Printing Office, 1947. for the behavioural sciences New York: McGraw-Hill Book Co., Leonard, J. A. & Newman, R. C. Spa- 1956. tial orientation in the blind. Nature, 1967, 215, 1413-14.

31 »? L AN ANALYSIS OF OPTACON USAGE

Loren T. Schoof n*

Abstract : Three topics are ex- was the age of the participant at plored which relate to Optacon usage. the time of training. The results The first topic is the categorization of this analysis have been updated of Optacon users according to occupa- with seventeen additional partici- tion. Where relevant, the data pre- pants. It is also shown that, during sented here are compared with those the time period under investigation, of the earlier American Foundation there appears to be no significant for the Blind survey (Goldish and change in the effectiveness of the

Taylor, 1974) . A major change is training courses. that students have become the largest group of Optacon users , replacing The third topic is a description computer professionals. However, a of how people in different occupa- wide range of occupational categories tional groups are using the Optacon is represented and a breakdown is on the job. Although the results are given for those people trained at qualitative, it does appear that those Telesensory Systems, Inc. in entry-level positions are more strongly motivated to take advantage The second topic is a statisti- of the Optacon on the job. People cal analysis of factors which affect past the entry-level position seem, to the performance levels attained by find the Optacon enables them to plan the end of the TSI training course. and organize their work more effi- A total of 41 participants in this ciently. The ability to read confi- course during 1973 and 1974 form the dential information without assistance data base for this analysis. Two of another person also appears to be variables were considered as measures of importance to Optacon users. All of performance : Optacon reading speed of the people interviewed felt that and letter recognition. The latter the Optacon was of significant value proved to be unsuitable for statisti- to them in their employment. cal analysis. The factors which were analyzed for their effect on Optacon reading speed were age, sex, age of INTRODUCTION onset of blindness, and braille reading speed. An analysis of variance Since the Optacon (a portable, showed that the only factor which had direct translation, inknrint reading a statistically significant effect aid for the blind) first became available in 1971, a number of studies and surveys have been conducted on *Sensory Aids Specialist, Telesensory topics related to the process of Systems, Inc., Palo Alto, California learning to use it, as well as its 94304. application. Tobin, James, McVeigh,

and Irving (1973) , and Weisgerber, Editor's Note: Mr. Schoof, who is Everett, Rodabaugh, Shanner, and blind himself, performed all of the Crawford (19 74) have attempted to statistical analyses he describes in identify predictors of performance this paper with an Optacon, HP-35 in an Optacon trainina course. Gold- Calculator, and a Perkins Braille- ish and Taylor (19 74) have surveyed writer as his only aids. Optacon users to determine how important the Optacon has proven to be in daily life.

33 The studies relating to Optacon taken. These programs have made it learning have tended to follow the possible for people who would not same pattern. The researcher devel- otherwise be able to afford an Opta- ops a set of training methods and ma- con to do so. These programs may terials based upon his own theory of have signif icantlv changed the occupa- education and his assumed performance tional distribution of Optacon users. criteria. He then selects a group of Also, a number of questions have been subjects who have different levels of raised about the suitability of the characteristics which he believes may Optacon in some occupational situa- affect Optacon learning. The siibjects tions. Based on brief studies of are briefly trained toward these per- Optacon learning and on interviews formance criteria with the Optacon, with newly trained users, some re- according to the methods he has de- searchers have made claims about the veloped. At the end of trainina, appropriateness of this device which' measurem.ents of variables which re- would seem to be no more than con- flect achievement according to these jecture (Tobin, et al., 1973). performance criteria are obtained. An analysis is performed to ascertain In view of the problems cited which factors have a statistically above with previous Optacon-related significant effect on Optacon per- research, it seemed appropriate that formance. The Optacon students are another paper be prepared which would then dismissed without any assurance take into account some of these prob- of access to an Optacon no matter lems. The survey of occupations of what their performance in the train- Optacon users has been updated to re- ing. flect the rapid increase in the number of Optacon users. A statistical Telesensory Systems training analysis of Optacon learning of TSI methods have been developed through trained students has been made. How- an iterative process. They are in- ever, the variables in this type of tended to help the student integrate analysis do not necessarily reflect the Optacon into his daily life as the utility of the Optacon to a blind quickly as possible. At the end of person. To overcome this problem, a training, a student knows that the series of in-depth interviews are Optacon will be his no matter what presented which give a greater feel- his performance during the course. ing for how people in different oc- Since the student will retain use of cupations are using the Optacon. the Optacon after training, there would seem to be significantly more motivation for a student to learn to PART I-OCCUPATIONS use the Optacon in this situation. OF OPTACON USERS However, the TSI training courses are individualized to each student's Before coming to TSI for train- desires and needs. For example, if ing, each Optacon student fills out a student and the teaching staff de- a personal information form. One of cide it is more appropriate, in re- the questions concerns the student's lation to the student's projected occupation at the time he receives needs, to emphasize instruction on Optacon training. From this and other the format of certain documents sources, the occupations of 250 Opta- rather than on speed building, the con users were determined. Table 1 resulting course will reflect this summarizes this information in terms emphasis. Thus, because these courses of percentage of the sample for each are designed to provide a service to occupational group together with the the blind individual rather than ful- corresponding information from Goldish fill a research purpose, the goals of and Tavlor, 1974. the training vary according to each individual's needs. VThile this is by no m.eans an ex- haustive sample of ail Optacon users, Approximately one year ago, the it does give some idea of the occupa- American Foundation for the Blind con- tional groups which are taking full- ducted a survey of Optacon users est advantage of the Optacon. Since

(Goldish and Taylor, 1974) . Among the AFB survey, there are two changes other topics covered was a breakdown worth noting: 1) students have re- of Optacon users by occupation. Since placed computer professionals as the that survey, a number of Optacon dis- largest single group using the Opta- semination programs have been under- con, and 2) housewives have now

34 TABLE 1

Occupations of Optacon Users (General)

TSI Sample AFB Survey Occupation (Percent) (Percent)

Student 25,,90 20 Computer Professional 22,,31 23 Social VJorker (Counselor or Administrator) 17,.13 14

Housewife 12,,79 8 Business (Clerical and Administrative) 11,,55 13

College Professor or Administrator 3,,98 4

Teacher (Elementary and Secondary) 2,,39

Attorney 3,,59 3

Engineer and Scientist 1,,99 5

Other 0,,80 8 replaced business and clerical work- TABLE 2 ers as the fourth largest occupational group using the Optacon. The in- Occupations of Optacon Users (Detailed) crease in the number of students using the Optacon can probably be explained by the growing interest of Students school systems in providing Optacons Elementary (6-13 years) 5 and Optacon training, and by the pro- High School (14-17 years) gram of the Richard King Mellon Foun- 1 College (18-21 dation. This program has allowed years) 21 Graduate School 11 many people such as students and housewives to obtain what would other- Housewives wise be a prohibitively expensive piece of equipment. Business Within each of these broad oc- Receptionist 2 cupational categories , there is a Dictaphone typist 3 wide diversity of possible job defi- Secretary 2 nitions. For the sample of 250 Opta- Technician 2 con users, a more detailed breakdown Assembler 1 of occupations was not possible. Systems Analyst 3 However, for those students trained Programmer 43 at TSI, Table 2 gives a more specific Information person 1 breakdown of their occupations. In Expediter 1 addition to those listed in the table, X-ray film processor 1 a television producer, a braille Finance 1 proofreader, and a meteorologist with Personnel 1 the National Weather Service have be- Taxpayer Service Rep. 1 come Optacon users. Field Rep. for Company 1

Professional Engineering 4 Physicist 1 Attorney 4 Teacher (Elem. & High School) 5 Librarian 1

35 ,

TABLE 2 (Contd.) end of a nine-day training course. They give no indication of what happened to Optacon reading performance Professional (Contd.) after additional practice. To date, no quantitative follow-up studies Psychologist 2 have been done. Sociologist 1 School Coordinator 1 College Professor 5 Training Procedures College Dean 2 High School Counselor 1 Optacon training at TSI is an Counselor 3 intensive course oriented to the Rehab. Director/Center 1 needs of blind adults. Fifty to Rehab. Administrator 5 sixty hours of instruction are spread Rehab. Teacher 8 over nine class days. A one-to-one Rehab Counselor 8 student/teacher ratio is maintained Newscaster 1 at all times. Teacher assignments Minister 1 are so arranged that the student Foreign Veteran i works with a different teacher each Coffee Shop Owner 1 day. A logging procedure has been developed so that each new teacher is familiar with the student's prior PART n-STATISTICAL progress. The course work centers on

ANALYSIS OF TSI five principal areas : eauipment opera- TRAINING COURSE DATA tion, letter recognition, tracking, speed building, and the format of The data used in this analysis common materials. These are not was gathered from July 19 7 3 through taught as separate topics , but must February 19 74. The people who come frequently overlap with one another. to TSI for Optacon training are not Letter recognition, for example, in- considered to be subjects in an ex- volves some practice in tracking. periment. They have paid a fee to attend the course and have, either The first area covered is the through purchase or other means operation of the Optacon. The stu- guaranteed access to an Optacon at dent learns the location and function the end of training. The objective of each of its controls. He learns of Optacon training is to enable them how to adjust these controls to pro- to use the Optacon on the job and in duce the best possible image on the other daily activities. tactile screen for the materials to be read. The teacher usually helps Optacon trainees were asked to the student with the adjustments take tests which would measure the during the first fev; lessons. variables used in this analysis. They were not required to do so and, Almost from the beginning, the in fact, a number of individuals re- student learns to recognize letter fused. Out of the total number of shapes and to synthesize them into people trained during this time peri- words, phrases, and sentences. The od, 41 data points were obtained. materials used to introduce the stu- After the majority of the analysis dent to print letters are in a simple, had been completed, an additional 17 sans serif type font. Care has been data points from the March, April, taken to insure that this initial and Hay classes of 1974 were used to printing is as clear as possible. further update the resulting model. These measures allow the student to The actual number of students trained concentrate on the primary task of in that period was somewhat larger. letter recognition. Later, as he becomes more skillful, other type As with any statistical analysis, styles and poorer quality print can the results here apply only to the be introduced. Upper case letters data upon which they are based. Any are introduced first since they seem extension of them to other individuals to be more easily learned. The format must be based on the assumption that of the lessons, however, is the same these results are true in general as for both upper and lower case letters. well as for the sample analyzed. Furthermore, these are results at the

36 In general, a test is given at a fixed rate of speed. This relieves the beginning of a lesson to make the student of the tracking task. He certain that the student can identify is forced to spend less time on recog- letters that have previously been nizing individual letters. Instead, learned. If he is successful with he must try to read whole words and this, a few new letters are intro- to use context to anticipate new ma- duced. A maximum of eight new upper terial. case or five new lower case letters appear in a single lesson. After the The student can make more effec- student is satisfied that he can tive use of the Optacon if he can identify the new letters, another test take advantage of the special formats is given. At the discretion of the of some printed materials. Samples teacher, the student can move to the of formats such as dictionary pages, next lesson or work with optional bank statements, computer listings, practice material included in the and memos are included in the train- lesson. After a few letters have ing materials. The teacher can show been learned, lessons are included the student how to take advantage of which combine these letters into the format to quickly find important short words and sentences. As new information. Students who come to letters are mastered, they are used TSI for training are also encouraged in these practice lessons along with to bring samples of the materials those letters previously learned. they are interested in reading.

Tracking, the skill of moving A more detailed description of the camera across printed material, Optacon training at TSI can be found is developed concurrently with the in the Teaching Guidelines manual letter recognition and speed building listed in the bibliography. Methods aspects of Optacon training. The described here are the result of ex- first lesson a student encounters is perimentation and revision over a designed to show him the stimulation period of several years. Some of the from the tactile screen and the ef- papers listed in the bibliography show fect that camera movement has on that how the direction and emphasis in Op- stimulation. At the start of train- tacon teaching have changed over the ing, the student uses a tracking aid years (Weihl, 19 71; Baer and Hill, to make this task somewhat easier. 1972) . No lesson plan or teaching This device allows him free movement technique has been preordained as the of the camera in the horizontal direc- one absolutely correct method. In- tion while restricting its vertical deed, the procedures outlined above motion. As the student's ability to will continue to be revised as addi- relate camera motion to the image on tional experience is acquired. the tactile screen improves, the restriction on the vertical motion of the camera is relaxed. Finally, the Choice of Variables student moves the camera freelv with- for Analysis out the use of a tracking aid. The only truly valid measure of The latter stages of the train- Optacon learning is the extent to ing course are devoted to helping the which an individual trained in its student integrate the Optacon more use becomes dependent on it. This rapidly into his daily activities. is, however, difficult to measure in

This help is in two main areas : in- a quantitative manner. The variable creasing the student's reading speed most commonly used as a measure of and familiarizing him with the format performance in training has been Op- of materials he will frequently be tacon reading speed. This variable using. In the first of these areas, was used in the English Optacon a number of short stories have been evaluation (Tobin, et al. , 1973) and included in the materials used at TSI. in the Office of Education evaluation

These stories allow the student to (Weisgerber, et al. , 1974) . Weis- practice both his reading and track- gerber also used a letter recognition ing skills. test as a measure of reading accuracy. Both of these variables were measured A piece of equipment, the Auto- for use in this analysis. matic Page Scanner, is also used in training. It moves the Optacon camera Since this was not an experi- across a page of printed material at mental program, it would have been

37 difficult to insist that Optacon stu- this point, the student should have dents undergo a battery of physical some idea of the context of the story. and psychological tests. The vari- He then reads the remaining three ables selected for their possible paragraphs silently with the in- effect on Optacon learning perform- structor recording the time taken ance had, therefore, to be measurable for each. These three times are without undue imposition on the stu- later used to compute the average dents. Braille reading speed, for reading speed. As with the braille example, seemed to be a reasonable, test, the teacher checks for compre- if crude, measure of tactile sensi- hension. tivity. Tobin, et al. (19 73) found that this variable was significantly Letter recognition is tested us- correlated with Optacon reading speed, ing exercises from the training ma- V^eisgerber, et al. (1974), however, terial. Specifically, a criterion found that there was no relationship test from Lesson 9 is used for upper between braille reading speed and case letters and one from Lesson 20 Optacon reading speed. They did re- for lower case. The student moves port a relation between this variable the camera himself as he identifies and Optacon reading accuracy. Tobin each letter. The instructor records also found a significant relationship his answer without telling the stu- between age and Optacon reading speed. dent whether or not it is correct. This variable was also included in There is no time constraint and the the present analysis. As early as student may only scan each letter one 1970, there was some interest in the time. The percentage of letters cor- possible influence of past visual rectly identified is used as the let- memories on Optacon reading (Baer and ter recognition score. Hill, 1972). This variable was' used by Tobin in his Optacon evaluation. Using standard estimators, the Based on a subjective feeling that means and standard deviations for it might have some effect, sex was these variables were calculated. The also considered as a factor. value "1" was used for a male and "0" for female. Table 3 summarizes the information about the sample. Measuring Variables

Age, sex, and age of onset of TABLE 3 blindness were determined by a questionnaire. Braille reading speed is Characteristics of TSI Student tested on the first day of class. Population Each student works in a separate room and on a one-to-one basis with his Standard instructor. The student is given Factors Mean Deviation three short stories in standard grade 2 braille. He reads each story si- Age (years) 30.12 9.57 lently. The instructor checks for Sex 0.707 0.461 comprehension and, using a stopwatch, records the time it takes to read Age of Onset of each story. After classes are fin- Blindness (years) 5.63 10.62 ished, the three times are used to Braille Readina compute the average braille reading Speed (words/min) 101.35 48.38 speed for each student. Optacon Reading Letter recognition and Optacon Speed (words/min) 9.61 4.65 reading speed are measured on the Letter Recognition final day of training. Again, the (percent correct) 0.9 31 0.051 instructor works alone with the student. As in training, a Visual Dis- play is connected to the Optacon so the instructor can monitor the stu- Distributions of the dent's performance. Reading speed Measured Variables is tested using a five-paragraph story. The teacher reads the first To perform an analysis of vari- paragraph aloud. The second is read ance on either Optacon reading speed aloud by the student with the instruct- or letter recognition, that variable or helping him when necessary. At must be normally distributed.

38 . .

Chi-squared goodness or fit tests interval 0.500 to 1.000 was divided were used to check this hypothesis into seven unequal intervals so that Optacon reading speeds ranging from most of the intervals could be in zero up to 20.22 words- per-minute the range where most people scored. were observed at the en d of the train- The hypothesis is that letter recog-

ing course . For the ch i-squared test, nition scores are normally distributed the interval from zero to 22 was di- with an estimated mean of 0.9 31 and vided into 11 equal two unit segments, standard deviation of 0.051. Table 5 The hypothesis tested i s that, on this summarizes the observed and predicted interval, Optacon readi ng speeds are number of scores in each interval. normally distributed wi th estimated Again, the predicted values have been means of 9.16 and stand ard deviation adjusted to restrict the normal dis- of 4.65. tribution to the interval being tested. Table 4 shows both the observed and predicted number of reading scores in each interval. The pre- TABLE 5 dicted numbers have been adjusted so that the normal distribution is com- Histogram of Letter Recognition pletely restricted to the interval Scores from zero to 22

Interval Observed Predicted TABLE 4 0.50 - 0. 80 1 0.23 Histogram of Optacon Reading Scores 0.80 - 0.90 3 11.96

0.90 - 0.92 8 6.39 Interval Observed Predicted 0.92 - 0.94 9 7.30

- 2 1 1.58 0.94 - 0.96 7 6.32

2 - 4 4 3.02 0.96 - 0.98 7 5.21

4 - 6 8 4. 84 0.98 - 1.00 6 3.60

6 - 8 6 6.45

8 - 10 7 , 7.17 From these figures , the calculated value of the chi-squared statistic 10 - 12 3 6.65 is 12.380 with 4 degrees of freedom. 12 - 14 6 5.13 At the five percent level, the predicted value of chi-squared is 9.488, - 3.30 14 16 3 so the hypothesis of normal distribu- 16 - 18 1 1.77 tion is rejected. The hypothesis would be accepted at the one percent - 0.79 18 20 1 level where chi-sauared is 13.277. 20 - 22 1 0.30 Another area to be tested is the statistical independence of the let-

Based on these figures , the calcu- ter recognition and Optacon reading lated chi-squared statistic is 6.832 speed variables. A chi-squared test with eight degrees of freedom. At of this gives a chi-sauared value of the five percent level, the predicted 86.394. The predicted value of chi- chi-squared equals 15.507 so the hy- squared at the five percent level is pothesis is accepted at this level. 79.082 with 60 degrees of freedom. In fact, the hypothesis is finally The hvpothesis is rejected at this rejected at the 70 percent level level. It would -Finally be accepted where chi-squared equals 5.527. This at the one percent level where chi- suggests that the hypothesis of a squared is 88. 379. normal distribution is very strong. Since letter recognition is Letter recognition test scores neither normally distributed nor ranged from 0.500 up to 0.99 3. All independent of Optacon reading speed, but four of the people tested had it did not seem appropriate to in- scores between 0.900 and 1.000. The clude in further analysis.

39 Correlation Coefficients The equation for the regression analysis will be ordered as in the In the regression analysis, a table above. Although the coeffi- set of linear equations is solved to cient for age of onset of blindness yield the coefficients. In solving has a slightly larger magnitude, sex these equations, it is helpful to the was solved for first on the basis of analysis of variance if the factors an earlier analysis and with a smaller most likely to be statistically sig- sample. nificant are solved for first. The correlation coefficients have been All of the t statistics have 39 calculated in an attempt to get a degrees of freedom. At the five-per- rough measure of the significance cent level, the t distribution has the of each factor. The factors will be value 1.6 85. From this, it appears ordered according to which one has that only for braille reading speed the largest absolute value of its would the hypothesis of a zero cor- correlation coefficient. These correlation be accepted. Braille read- relations are also of interest in ing speed would become significant themselves as they sometimes disagree at the 40-percent level where t equals witli the results of analysis of vari- 0.255. At the one-percent level, t ance . equals 2.426. Hence, sex and age of onset of blindness would have zero The standard estimator for the correlation at this level. At the correlation coefficient has been used. 0.05-percent level, t equals 3.561 so This estimator is derived from the bi- that age still has a non-zero correla- variate normal distribution. The t tion. statistic is used to test the hypothesis that the correlation coefficient is statistically equivalent to zero. Regression Analysis The variable for Optacon reading speed has already been shown to have To construct a linear model for a normal distribution. Although Optacon readina speed, coefficients tests have not been made, it appears of the normal equations had to be that age and braille reading speed calculated. For a number of reasons, are also normally distributed. On it seemed better to break each factor the other hand, the distribution of up into a set of variables which take sex within the sample has a demon- the values zero and one only. The strably binomial form. An examina- definitions of the controllable vari- tion of the data also suggests that ables are given below. age of onset of blindness would probably have a distribution represented Age . From earlier analysis with by a decaying exponential. In this a smaller sample, it was clear that case, the correlation coefficients there was an extremely sharp break and the statistical tests on them between people above the mean age and are only a very rough guide to the those below it. It seemed reasonable significance of these factors. then to use a set of discrete levels to more accurately represent the Table 6 shows the correlation situation. These variables take the with reading speed and the calculated value one in the indicated age range t statistic for each factor. and are zero otherwise as follows:

TABLE 6 of '11 those under 20 years age

20 30 Correlations with Optacon '12 those between and Reading Speed those between 30 and 40 Correlation '13 Factors Coefficient t '14 those over 40 Age - 0.6047 4.74 The reading speeds within each of Sex* - 0.3248 2.14 these age ranges appear to be more

Age of (Dnset of consistent than would be the case Blindne;3S - 0.3441 2.19 with a smaller number of levels. Braille Readi•ng Speed 0.0672 0.402 *Negative value indicates that females had higher reading scores. 40 : :

Sex. This factor has an obvi- '32 those blinded during the main ous two-level representation. For part of their education proc- this analysis as for the correlation ess, between 10 and 2 5 coefficients z-^^ those blinded after the major (1 for male portion of their education, over 25 (0 for female

Age of onset of blindness . Braille Reading Speed This factor was divided into three

levels. These levels were based on z . those at or below the meem the point in education at which peo- braille reading speed ple became blind.

z^, those blinded early in their Using these variables, the ma- education, under 10 years trix which results from the normal of age eauations is as follows

3.61 -1.95 -1.17 -0.49 0.17 -0.32 0.71 -0. 39 -0.15 20.74

-1.95 10.24 -5.85 -2.44 -3.15 1.41 0.54 -1.95 2.27 33.07

-1.17 -5. 85 8. 49 -1.46 1.51 0.05 -0. 88 0. 83 -1.44 -28.76

-0.49 -2.44 -1.46 7.39 1.46 -1.15 -0.37 1.51 -0.68 -25.70

0.17 -3.15 1.51 1.46 8.49 -2.05 0.88 1.17 -4.56 -27.86

-0.32 1.41 0.05 -1.15 -2.05 5.80 -2.49 -3.32 3.76 12.90

0.71 0.54 -0.88 -0.37 0.88 -2.49 2. 78 -0.29 -1.61 6.32

-0.39 -1.95 0.83 1.51 1.17 -3.32 -0.29 3.61 -0.02 -22.11

-0.15 2.27 -1.44 -0.68 -4.56 3.76 -1.61 -0.02 10.20 8.32

The first term of the linear model is equal to the mean 9.16. When the matrix is solved, the resulting linear model for Optacon readinq speed, x, is;

X = 9.16 - 2.84 (z - 0.098) 6.72 {z^2 " 0.488) 10.46 (z - .293)

13.31 (z 0.122) 0.74 [z^ - 0.707) 2.18 (z^^ - 0.829) 2.36 (z 14 32

• 0.073) 3.83 (z 0.098) + 0.08 (z^ - 0.0537) 33

41 Analysis of variance . It is pos- At the five-percent level, the pre- sible that some of the factors solved dicted value F £ = 2.6 8 so that the for do not, in fact, help in predict- 31 ing the person's reading speed. The coefficients correspondina to age are analysis of variance technique can be statistically different from zero. used to determine if some of the co- In fact, they remain statistically efficients are statistically equiva- significant up to the 0.1-percent lent to zero. The technique involves level where F — =6.0 8. the division of the sum of squares 31 into parts corresponding to each variable. The remainder is the error term, Since age is the only signifi- and can be used as an estimator for cant factor, the linear model for the standard deviation of the error Optacon reading speed, x, is: distribution. Table 7 gives the results of the analysis of variance. X = 17.27 - 2.84z^^ - 6.72z 12 - 10.46z,2 - 13.31z,^ + e

TABLE 7 where e is the error term which has a Analysis of Variance of Optacon normal distribution with mean zero Reading Speed Predictors and standard deviation 3.73.

SS is the sum of squares. DF is the Update of the Original degrees of freedom, which is the num- Analysis ber of variables representing a factor in the normal equations. MS is As mentioned above , data of the the m.ean square and is the SS entry same kind used in the main analysis divided by the DF entry. F is the was also gathered from the March, ratio of the MS entry for the error April, and May classes of 19 74. A term. total of 2 3 people were trained to use the Optacon during this time. Factor SS DF MS For a number of reasons, primarily previous experience in using the Mean 3440.13 1 Optacon, six of these people could not be used in a statistical analy- Age 363.64 4 90,91 6.09 sis. The remaining 17 data points Sex 21.00 1 21.00 1.41 were added to the original 41 to compute a new mean of 10.00 anc! Age of Onset standard deviation of 5.18 for Opta- of Blindness 18.50 6.17 0.42 con reading speed at the end of Braille Read- training. Because of the earlier ing Speed 0.03 1 0.03 0.002 analysis, it was assumed that age remained a significant factor in \ 462.55 31 14.92 Error predicting Optacon reading speed. Therefore, the linear model for mean reading speed was updated. The re- If a particular factor is such sulting equation is: that all of the coefficients in the linear model are zero, the F entry X = 14.35z 12.37Z '^•^^^12 should be small relative to the pre- + 2 + dicted value from the F distribution. + 4. 62z^ . + e For braille reading speed F _ = 4.16 31 where e, the error term, has mean zero the five percent level that the at so and standard deviation of 4.29. This hypothesis that braille reading speed equation is graphed in Fig. 1 which does not predict Optacon reading speed illustrates the dram.atic relation is accepted. For of onset of age between age and Optacon reading speed blindness, F — = 2.91 so that this at the end of the TSI trainina course. 31 factor is also not significant at the five percent level. For sex, the predicted value of F is the same as for the braille reading speed, 4.16, so that this factor is also insignificant.

42 that of the early group is accepted. This hypothesis would be rejected STANDARD DEVIATION at the 10-percent level where t 15 -- OF equals -1.330. At this level, it ERROR would appear that the mean reading speed of the later group was statistically significantly higher than that of the early group.

Since age is known to be a significant factor, a t test under the is same assumptions was used to deter- ^1 mine if a significant difference existed in the mean ages of the two groups. The computed value of the t statistic is -0.758. Hence, there is no statistically significant difference between the mean ages of the two groups.

20 30 40 AND OVER These two tests do not neces- AGE GROUP sarily provide an absolute answer. (YEARS) However, based on them., it appears that no statistically significant Figure 1. Optacon Reading Speed at change in mean reading speed has the End of TSI Training Course as occurred as a result of additional a Function of Age Group, Based on teacher experience. A more detailed Data from 5 8 Students Between Mid analysis of variance would be neces- 19 7 3 and Mid 1974. sary to understand all of the effects and interactions between teacher experience and Optacon reading speed. Effect of Teacher Experience on Student Performance Implications of the Analysis A question was raised concerning the effect of additional teacher ex- The most important result of perience on the student's perform- this analysis is the negative cor- ance. To test this, mean reading relation between the age of a person speeds from sample groups at the beat the time of training and his read- ginning and at the end of the data ing speed at the end. of that train- collection were compared, The test ing. This does not imply that an was based on the assumptions that older person cannot learn to read 1) the underlying distribution of with the Optacon. It simply states reading speeds was normal, and that the average reading speed that 2) the standard deviations of both can be expected at the end of train- samples, although unknown, were ing will be lower. There is, how- equal. These assumptions seemed ever, a reasonable probability that reasonable in view of the previous a person in a given age group will analysis. Ten students from the read more quickly or more slowly first four classes were used for than the predicted mean. It may be the early sample, and ten from the that, for people over the age of 40, April and May classes were used as a less intensive, more extended the last sample. The students in period of training would be more the first group had a mean reading appropriate. Also, any "screening"

speed of 9.86 words-per-minute , and test which attempts to predict a the students in the last group had person's performance in Optacon a mean reading speed of 13.35 words- training will have to take age into per-minute. The computed value of account as a concomitant variable.

the t statistic was -1.469. With Other\-7ise , the effect of this factor 18 degrees of freedom., the predicted may overwhelm the predictive value five-^percent level of the t distribu- of the test itself. tion is -1.734; hence, the hypothesis that mean reading speed, of a The fact that letter recogni- later group is less than or equal to tion scores do not have a normal

43 distribution is, on reflection, not Perhaps the most remarkable con- surprising. Reading with the Optacon clusion to come from this analysis is at the end of training is a slow, al- the complete lack of significance of most letter by letter, process. In braille reading speed. As indicated order to read, then, a person must be earlier, it was thought that this able to recognize individual letters. variable might give some indication Even an experienced Optacon reader of tactile sensitivity. Apparently, may be forced to examine an unusually it does not. VJeisgerber, et al. long or unexpected word in this man- (1974) obtained the same result.

ner. Perhaps a more appropriate Tobin, et al . (1973), however, did mathematical model for letter recog- find a statistically significant re- nition scores would be a Bernoulli lationship between this variable and trial: that is, either the person Optacon reading speed. recognizes letters or he does not.

In light of these remarks , the inter- Even with age as a factor, the dependence between Optacon reading error term in the predictive model speed and letter recognition should has a relatively large standard devia- be expected. The surprising part is tion. This implies that predictions the marginal quality of that relation- based on this model are relatively un- ship. certain. TSI has developed an assessment test battery which measures The correlation between sex and short-term memory, manual dexterity, Optacon reading speed was statistic- and tactile sensitivity. To date, ally significant. Yet, the analysis the effectiveness of this test as a of variance showed that this factor predictor of performance in Optacon had no predictive value. The dis- training has not been statistically agreement in the results obtained' analyzed. Hopefully, this test or from these two methods of analysis some similar instrument can be de- can be explained in terms of the ef- veloped to help assess a person's fect of age on performance. The aver- potential for effective use of the age age of the women was 26.5 years Optacon. while that of the men was 31.6 years. These means are on opposite sides of the population mean which is the most PART III-USE OF THE significant breakpoint for Optacon OPTACON ON THE JOB reading speed. As mentioned before, the probability of a certain number Part I gives some idea of the of men in a sample is given by a bi- occupations where the Optacon can be nomial distribution. The probability used as an effective tool. To try of choosing a man at random in the to find out how people in different entire U.S. population is 0.487. Us- occupational groups actually use the ing this value, the probability ac- Optacon, a telephone survey was con- cording to the binomial distribution ducted. A total of 17 people were of 29 or more men in a random sample interviewed in this manner. An at- of 41 people is 0.00 355. This would tempt was made to get two people from suggest that people selected by the each of the major categories shown in method of Optacon purchase has not Table 1, Part I. been a random process. Those women

who have obtained Optacons have , on The Questions asked were intended the average, been younger. to determine the amount of time the person used the Optacon each day, the There is an even simpler explana- kinds of materials read, and the im- tion for age of onset of blindness portance of the Optacon to the person having a statistically significant in performing his job. Information correlation with reading speed but not was also gathered about use of the being a significant factor in a pre- Optacon outside of the job, estimated dictive model. The older a person is reading speed, and some other details when he becomes blind, the older that about age and Optacon training. The person has to be when he is trained following are brief summaries of these to use the Optacon. That is, it is a telephone interviews. logical contradiction to have someone who lost his vision at 34 but is 25 Respondent 1 . when trained to use the Optacon. Again, the correlation coefficient is ex- Age estimated— 2 4-25 plained in terms of the significance of age. 44 .

Age of onset of blindness: un- plans to use the Optacon extensively known, apparently early during his studies. He states that he is very dependent on the Optacon Optacon training center: Vision and would feel lost without it. Center of Central Ohio, Inc.

Date of training: March 1974 Respondent 3.

The respondent is currently in Age : 26 training to be a medical transcriber. She has already graduated from col- Age of onset of blindness: 26 lege with a bachelor's degree in linguistics. Due to the recent date Optacon training center: TSI of her training with the Optacon, she has not been able to make full use of Date of training: June 19 73 it. She intends to use it as a selling point when seeking employment. The respondent is currently Currently, she uses it to read dic- studying for a master's degree in tionaries, personal correspondence, social work. She uses the Optacon and other similar personal materials. to read a wide variety of materials She is also using the Optacon to learn necessary in her studies. She de- to read different alphabets such as pends on the Optacon to read class Greek and Arabic. She indicated that handouts, notes, tests, and some Arabic presented serious difficulties. texts. She indicated that the Opta- Her reading speed at the end of train- con has been extremely useful in ing was 20 words per minute; she proofreading her typing of reports would not give an estimate of her and papers. She believes she uses present reading speed. the Optacon for an average of two

hours-per-day . She estimates her reading speed at 40 words-per-min-

Respondent 2 . ute

Age : 20 During her internship work at a hospital, she used the Optacon to Age of onset of blindness: 3 review charts and write-ups of previous interviews with patients. She Optacon training center: San also used the Optacon to read confi- Diego Unified School District dential case files. She felt that this was very important since it was Date of training: December 1971 difficult to obtain access to these files in any other way. She does not He is currently majoring in busi- feel she could function in graduate ness administration and Spanish. He school or in an employment situation uses the Optacon to read hand-outs of without the Optacon. Her ability to class materials, announcements, and use the Optacon will help her to some brief passages of text. During obtain a job. the last school year, he used the Op- tacon to read all of the materials

for a course in Spanish grammar. He Respondent 4 . indicated that he was able to do this

sufficiently well to complete the Age : 3 8 course with a high grade. He is very

interested in using the Optacon with Age of onset of blindness : birth a pocket calculator as he believes this will help him in his business Optacon training center: Prof. courses. He estimates his reading Robert Stilwell, Morgantown, speed to be in excess of 40 words- West Virginia per-minute , and says that he reads with it from an hour- to an hour-and- Date of training: early 1973

a-half-per-day . Outside of class, he uses it to read personal correspond- Respondent is a professor of po- ence and materials from groups in litical science currently using the which he participates. He will at- Optacon an hour or more each day. His tend law school after graduation and reading speed is between 40 and 50

45 words-per-minute. He uses the Opta- Optacon training center: TSI con for seve ral tasks related to his university w ork (memos , circulated Date of training: November 1972 articles, et c.) , but would like to increase his speed. He seemed inter- Respondent works as a programmer ested in the typewriter attachment as for an oil company and uses the Opta- he uses the Optacon to proofread his con three or four hours-per-day. She work . He re ads term papers and the- estimated her reading speed at 50 to ses and can easily understand the 6 words-per-minute but felt that the organization and thoroughness of a speed was not terribly important— it student's re search from looking at is sufficient to allow her to do her the format. job. She reads nearly all the materials for her job with the Optacon He does not feel dependent on (computer listings, program speci- the Optacon to do his job as he was fications, computer manuals, memos). in the same position prior to getting She also uses the Optacon with a CRT the machine. However, he does feel terminal, but usually cannot read he now works more effectively and can handwritten material. She mentioned read certain materials more quickly that the Optacon had not been a strong and organize his work more effective- selling point when she obtained her ly. current job but the independence and flexibility conveyed by its possession may have helped convince her employer

Respondent 5 . to hire her.

Age: 38 -

Respondent 7 . Age at onset of blindness: birth • • Age: 28

Date of training: December 1973 Age of onset of blindness: 4

Optacon training center: TSI Optacon training center: Vision Center of Central Ohio, Inc. Respondent currently works as an elementary school teacher. She uses Date of training: end of 19 73 the Optacon about four hours-per-day and estimated her reading speed at Respondent works as a computer

40 to 45 words-per-minute . She reads programmer for the State of Ohio, job-related material (memos, teach- uses the Optacon four to six hours- er's editions of textbooks) and per- per-day, and estimates his reading sonal items. Because elementary text- speed at 40 to 45 words-per-minute. books are frequently changed, the Although he has the option of receiv- teacher's editions are often unavail- ing braille output from the computer, able in other form. She uses the Op- he uses the Optacon to read computer tacon to find supplementary materials listings, program specifications, for the basic textbook of the course. computer manuals, memos and notices, She feels that scanning these materi- and only uses braille when he must als is one of the most important ben- scan a lengthy program written by efits to her. She also uses the Opta- someone else. At present, he feels con for work which she takes home he can skip between widely separated from school and feels strongly that sections of a program more easily in she is a better teacher because of braille; he can scan with the Optacon the Optacon, and wonders how she if necessary. managed without it before. He had his present job before obtaining the Optacon but now feels

Respondent 6 . he is very dependent on the machine to do his job properly and without

Age : 29 it he would not be nearly as produc-

tive . Age of onset of blindness: birth

46 . ,

Respondent 8 . Optacon training center: TSI

Age: 35-40 Date of training: April 1974

Age of onset of blindness: Respondent works as a lawyer for unknown a large automobile manufacturer. He uses the Optacon one-half to one Optacon training center: un- hour-per-day and estimated his read- known ing speed at 25 words-per-minute. He uses the Optacon primarily to read Date of training: 1972 summaries of court decisions, usually found in a weekly or biweekly journal, Estimates he uses Optacon about He reads his correspondence when his two hours-per-day on the job and has secretary is not available, but hopes a reading speed of 20 to 25 words- to make better use of the Optacon as per-minute. He uses the Optacon to his reading speed increases. He also look up information in references uses the Optacon for pleasure read- and to read correspondence. He would ing, such as sports news in the news- like to be able to read pocket calcu- paper. This is not readily available lators and would also like to look at to him in other forms . Since he was an oscilloscope display and other unable to use the Optacon for almost measuring instruments. He did not a month, his reading speed was prob- feel the Optacon was necessary in his ablv inhibited. He does not feel job as he had the job before obtain- the Optacon is essential for his job ing his Optacon; however, he did sug- at this time but said that if his gest that he now works more effi- reading speed were faster he would ciently . become more dependent on it. He values the independence the Optacon gives him.

Respondent 9 .

Age: 33 Respondent 11 .

Age of onset of blindness : birth Age: 2 8

Optacon training center: TSI Age of onset of blindness: birth

Date of training: December 1972 Optacon training center: Cleve- land Society for the Blind Respondent works as an electrical engineer and uses his Optacon approx- Date of training: November 1972 imately two hours-per-day on the job. He works primarily as a technical Respondent works as a lawyer in writer and estimated his reading a small law firm. He uses the Opta- speed at 40 to 70 words-per-minute. con about four hours-per-day, fre- He uses the Optacon to read inter- quently much longer, and estimated office materials , technical reports his reading speed at more than 50 references, to proofread his typing, words-per-minute. He uses the Opta- examine circuit diagrams and amateur con for almost all reading done on radio manuals. His reading speed is the job (trust agreements, federal higher in braille, but often the ma- statutes, law journals, contracts, terials he needs are not transcribed. correspondence) , but prefers longer He is therefore very dependent on the materials to be read to him. He is Optacon in his job, has no readers on completely dependent upon the Optacon his staff, and is not getting materi- in his job and could not have obtained als transcribed in any other way. He his present job without it. He en- feels the Optacon is sufficient in joys the independence the Optacon his j ob gives him.

Respondent 10 . Respondent 12 .

Age: 31 Age: 2 4

Age of onset of blindness: birth Age of onset of blindness : birth

47 : 1

Optacon training center: TSI incoming correspondence, but also reads confidential personnel files. Date of training: May 1974 He feels he could handle his job without the Optacon as he obtained Respondent currently works as a the job before obtaining the Optacon. dictaphone typist for the Canadian However, he believes he can work more Government and uses the Optacon about efficiently with it. one hour-per-day . She estimated her reading speed in excess of 20 words- per-minute and stated that she uses Respondent 15 . the machine to look up words in the dictionary and to proofread. She be- Age : 4 lieves the Optacon was helpful in obtaining her job. Her ability to read Age of onset of blindness: 21 print without assistance probably helped convince her employer. She Optacon training center: TSI now feels she could not do her present job without the Optacon. Date of training: January 1973

Respondent works as a high school

Respondent 13 . counselor in California and uses the Optacon about two hours-per-day . He

-' ' ' ' Age: 46 ' estimated his reading speed at 25 to 30 words-per-minute and stated that Age of onset of blindness: 12 he uses the Optacon prim.arily to read books and extended articles to keep Optacon training center: TSI himself up to date. For shorter articles, he uses readers and/or his Date of training: May 19 74 secretary. He feels it would be difficult to work without the Optacon Respondent is currently an ad- and as his reading speed increases he ministrator in a vocational rehabilita- hopes to use it for more of his job- tion department. He uses the Optacon related reading. He greatly values one-half to one hour-per-dav and es- the independence the Optacon gives timated his reading speed at 15 to 20 him. words-per-minute. He uses the Opta- con mainly as a sorting tool for correspondence at work and reads short Respondent 16 . material, such as one-page memos and correspondence, when his secretary is Age: 40 unavailable. He does not really need the Optacon to do his present job, but Age of onset of blindness: 36 feels it makes him more efficient and organized. He likes the added inde- Optacon training center: self- pendence it gives him. trained

Date of training: December 1973 Respondent 14. Respondent is employed as a

Age : 40 counselor in the psychological section of the Veterans Administration. He Age of onset of blindness uses the Optacon three or four hours- mid-20 's per-day in job-related activities and estimated his readina speed at 60 Optacon training center: unknown words-per-minute. He uses the Opta- con to read almost all of his job- Date of training: summer 1973 related reading as well as material related to his thesis, to draw and Respondent is currently employed read flow charts , and to proofread as a supervisor of caseworkers in an when he types. He firmly believes agency for the blind. He uses the he could not do his job without the Optacon about one-half hour-per-day Optacon as he would be much less ef- on the job and estimates his reading ficient and productive. He values speed at 15 to 20 words-per-minute. the independence the Optacon gives He uses the Optacon primarily to sort him.

48 .

Respondent 17 . Even for people who are not able to read as quickly with the Optacon,

Age : 39 it seems to give them an opportunity to organize their materials and time Age of onset of blindness: more efficiently. It permits them to birth choose which materials to have read or brailled and which to ignore. It Optacon training center: TSI also allows them to handle shorter passages of reading independently Date of training: February 19 74 while their secretary or assistant is working on more lengthy materials. Respondent works in the person- These individuals can deal privately nel department of a large chain store with confidential material which can and uses the Optacon an hour to an be of considerable importance to peo- hour and a half each day. He uses ple in senior administrative positions. the machine to read short memos , tel- With these peoole, the Optacon, al- ephone numbers, and journals such as though not "necessary," can be a tool the Journal of Rehabilitation , and which makes it possible for them to estimates his reading speed at 15 to become more efficient and more pro- 20 words-per-minute. He recently ductive employees. used his Optacon to analyze information on job applications rather than have his secretary do it. Although SUMMARY he had the job before obtaining the machine, he feels he is able to work The first part of this paper in- more efficiently now. He greatly dicates that, since mid-1973, more appreciates the added independence young people have been able to obtain the Optacon brings him. Optacons. This is apparently a result of large-scale Optacon dissemination programs such as that of the Conclusions Richard King ^''ellon Foundation. It also indicates that blind people in a The interviews suggest that the wide variety of jobs have felt that people who are making most effective the Optacon can sianificantly increase use of the Optacon on the job are the their efficiency as an employee. This younger people in entry-level posi- conclusion is substantiated by the in- tions. This can be partly explained terviews conducted with Optacon users. by the fact that younger people learn Some of these people indicated that to use the Optacon more rapic31v. they could not have obtained their Also, for people in entry level po- present jobs without it. Others in- sitions, it seems to be more import- dicated that, although not essential ant to demonstrate a capability in to their employment, the Optacon made advance in order to be able to per- them more efficient, better organized form the job with a minimum of addi- workers tional expense to the employer. Older people have, in general, already These interviews strongly sug- worked out alternative methods for gest that the Optacon, as an employ- obtaining information. They are usu- ment tool , cannot be limited to a ally in positions with high enough small set of occupations. Instead, rank that the employer is willing to it should be considered as a useful spend the additional money for read- tool in any job where retrieval of ers or special equipment. These in- printed information is essential. dividuals, therefore, do not have the incentive to perform their jobs with- The statistical section of this out special assistance. This implies paper indicates that young people read that those people who are most able more quickly at the end of Optacon and willing to use the Optacon have training than do older people. The not been the major group obtaining interviews show that young people are access to it. It seems imperative also more dependent upon it in their to find some means to allow younger emoloyment. The Optacon has given people to obtain Optacons ; this would them the ability to do a job without permit them to compete more effective- special concessions from their employ- ly for entry-level positions. er. Many of the people interviewed

49 . , . , expressed the belief that this made many young people in entry-level it possible for them to obtain a positions have found the Optacon job or to advance within their to be absolutely essential. present organization. In short.

REFERENCES

Baer, J. A., & Hill, J. W. Optical- an investigation of some learner to-taatile image conversion for variables and teaching methods the blind. Final Report on Con- Research Centre for the Education tract SRS 70-42 and Grant 14-P- of the Visually Handicapped, 55296/9-02 for Social and Reha- School of Education, University bilitation Service, U.S. Depart- of Birmingham, 1973. ment of Health, Education and Welfare. Stanford Research In- Weihl, C. The Optacon reading pro- stitute, Menlo Park, California, gram at Monroe Public School. June 19 72. The Hew Outlook for the Blind, May 1971. Bowker, A. H., & Lieberman, G. J. Engineering statistias Engle- V/eisgerber, R. A. , Eve re tt , B . E . ,

wood Cliffs, New Jersey: Prentice- Rodabaugh, B. J. , Shanner, W. M. Hall, Inc. , 1959 . & Crawford, J. J. Educational evaluation of the Optacon Eraser, D. A. S. Statistias--An ( optical-to-ta ctile converter) introduction. New York: John as a reading a id to b lind ele- Wiley & Sons, Inc., 1958. mentary and se condary students. AIP-34 500-9/74-FR. Goldish, R. H., & Taylor, H. E. Submitted to B ureau of Educa- The Optacon: A valuable device tion for the H andicapped for blind persons. The flew Out- U.S. Office of Education,

look for the Blind, February 1974. Washington, D. C. , under Con- tract No. OEC- 0-72-5180.

Teaching Guidelines (R17454-B) . Tele- Final Report, September 1974. sensory Systems, Inc., Palo Alto, California, 1974,

v.. Tobin, J., James, W. R. K. , McVeigh, A., & Irving, R. M. Print reading by the blind-- An evaluation of the Optacon and

50 .

SPELLED SPEECH AS AN OUTPUT FOR THE LEXIPHONE READING MACHINE AND THE SPELLEX TALKING TYPEWRITER

Ching Y. Suen*

Michael P. Beddoes**

Abstraat : This paper studies INTRODUCTION the applicability of a synthesized spelled speech code for two types of "How can one develop a set of speech aids for the blind: the Lexi- alphabet letter sounds on tape for a phone reading machine and the Spellex reading machine for the blind or for talking typewriter. In the investi- other instruments requiring an audi- gation of the perception of the tory readout?" The work of Metfessel spelled speech code, both the intel- (1963) and Metfessel and Lovell (1961 ligibility of letter sounds and the and 1972) indicated in aeneral terms effect of bandwidth variation were an answer to this auestion. Their examined. The results indicated aim was to produce a set of distinc- that all 26 synthesized letter sounds tive sounds very brief in duration for were recognizable at a bandwidth as each letter of the alphabet in order low as 3 kHz. after a short period of to facilitate a high reading rate. training The letter sounds were to be concaten- ated to form easilv readable words and The effects of presentation sentences. Using the tape recorder as speed, bandwidth, word length, and the chief tool, they described many pause between words on the intelli- techniques for removing redundant parts gibility of spelled sentences were of the letter sounds by the cut-and- also investigated using a computer splice method. In this pruning opera- simulated Lexiphone. Experimental tion they attempted to minimize dis- results indicated that an ordinary continuities within letters and be- young blind subject could read sen- tween letters. The task was difficult tences spelled out at 65 to 75 words- and time consuming, but Metfessel and per-minute with a high intelligibility Lovell came up with a set of alphabet score. Reduction in bandwidth, in- sounds which needed broad-band eauip- crease in presentation speed, and in- ment and very low distortion before crease in word length, all reduced subjects could read well. Work by the intelligibility of spelled speech. Beddoes (196 3) used a different scheme It was also found that a pause pro- to shorten the durations of the letter portional to the length of the pre- sounds. Samples were removed, from ceding word instead of a fixed one original recordings in a random fashion, did not increase intelligibility ap- and even so, comparable reading rates preciably. The application of the (about 80 to 100 words-per-minute) high-speed spelled-speech code to the were obtained. Spellex talking typewriter improved the accuracy of blind typists and in All the above work was based on many cases also improved their typing analog principles and was directed speed. more to test the idea of spelled speech, in particular how quickly *Department of Computer Science, Sir blind people could read with it, than George Williams University, Montreal, to develop a good machine to generate P.Q., Canada H3G 1M8. the spelled speech sounds; the **Department of Electrical Engineering, University of British Columbia, Van-

couver 8, B. C. , Canada. 51 . . technology' for which was not ready at of British Columbia. Except for the the time. As we see it now, digital few subjects who had been blind for circuits are preferable to analog only a few vears , the majority of ones in the development of speech subjects had acauired a good knowl- aids for the blind because of their edge of Braille (with an average compactness, high speed of operation, reading speed of 90 words-per-minute) and repeatability. Although digital All subjects were individually trained sampling will introduce a certain and tested. amount of quantizing noise, this effect can be minimized using high- This experiment was divided into resolution converters. two sessions of about one and a half hours each. At the beginning of the Assuming we store speech in dig- first session, the entire set of let- ital form, it is essential to mini- ter sounds was presented in random mize the number of bits of samples in order to the subject for identifica- order to cut down the cost. Several tion. The Durnose was to find out methods were employed to achieve this how natural and distinct the synthe- goal. First, vowels are auasiperiod- sized letter sounds were. Four sub- ic. By repeating the pitch period jects were tested for each bandwidth. (60 to 120 digital samples at 12.5- The result of this part of the ex- kHz sampling) 10 to 20 times, life- periment furnished scores which in- like vowels were produced. Second, dicated the subjects' proficiency at speech samples were economically identifying the synthesized letter stored as differences rather than sounds before training. Following their absolute magnitudes. Third, this, the subject was tauaht to since letter sounds can be synthe- recognize the letter sounds at the sized by concatenation of phonemes, 6-kHz bandwidth. Subseauently the only a limited number of phonemes subject was given control of the rather than the entire set of alpha- keyboard of the teletype so that any bet sounds were stored. Using these letter sounds could, be heard by and other techniques, the spelled striking the corresponding keys. speech code shown in Table 6 has been In this way, he learned to compare developed with the aid of a computer and contrast letters which could be

(Suen, 1972; Suen and Beddoes , 1973). easily confused. Short quizzes on It consists of only 18 basic phonemes, letter identification were given to namely. A*/, /s/, /d/, /dz/, /k/, the subject from time to time. In /P/, /t/, /v/, /w/, /i/, /e/, /e/, the course of the experiment, the /a/, /o/, /u/, /«,/, /m/ and /n/ bandwidth was gradually reduced to This code has a maximum output rate 3 kHz and words and sentences were of about 120 words-per-minute. The also introduced at a gradually in- construction of a digital spelled creasing speed. By the end of this speech generator for incorporation session, most subjects could recog- in the reading machine and the talk- nize all the letter sounds and had ing typewriter was found to be eco- been exposed to all -Four bandwidths nomically viable (Suen and Beddoes, and four speeds. The emphasis of 1973). Experiments indicated that the second session was on sentence spelled speech is suitable for both reading though letter sounds were machines. Further, it was shown also reviewed from time to tim.e. that, although synthesized speech At the end of the second session, does not generally sound very "natu- the subject was given a letter test ral," the subjects required only a to evaluate letter distinctiveness short period of training to master and learning effects. the code. The results of this experiment are shown in Tables 1 through 5, in PERCEPTION OF THE the form, of confusion matrices. The SPELLED SPEECH CODE total scores (out of 104) for each bandwidth are: 57 at 3 kHz, 70 at Perception of the spelled speech 4 kHz, 64 at 5 kHz, and 64 at 6 kHz. code was studied at the following The average score is 63.75, or bandwidths: 3 kHz, 4 kHz, 5 kHz, and 61.3 percent correct. 6 kHz. Sixteen blind subjects, aged 14 to 27, in eight groups, took part in the experiment. Most of the s\±>- jects were students of the University

52 TABLE 1

Confusion Matrix of Letter Sounds at 3-kHz Bandwidth

Response ABCDEFGHIJKLMN0PQRSTUVV7XYZ Stimulus

A 3 1

B 4

C 11 2

D 2 2

E 2 2

F 3 1 •

G 4

I 3 1

K .11 11

L ,1 12

M _ . 4

N 4 •

- . 4

P 4

Q 4

R 1 3 S 11 11 T 2 2

U 2 2

V 1 2 1

W 3 1

X 4

Y 4 Z 1 12

53 TABLE 2

Confusion Matrix of Letter Sounds at 4-kHz Bandwidth

Response ABCDEFGHIJKLMNOPORSTUVWXYZ Stimulus

A 3 1

B 2 2

G .3 ' 1

D 2 2

E 1 3

F 4

G 3 1

^ -, H 4 ^ .

I A J 13/.

K 3 . 1

L ^1 '3

M 1 . 12.:

N 3 1

P 3 1

. 1 Q . .3

4 R .

S 4

T 4 U 13 V 12 1 W 2 2

X .. 1 3

Y 4 Z 1111

54 TABLE 3

Confusion Matrix of Letter Sounds at 5-kHz Bandwic'th

Response ABCDEFGHIJKLMNOPQRSTUVWXYZ Stimulus

A 2 2

B 111 1

C 1 3

D 2 2

E 4

F 1 3

G 3 1

H 4

I 3 1

J 3 1 K 1 111 L 3 1

M 4

N 4

O 4

Q 2 2

R 4

S 4

fp 4

U 3 1 V 111 1

W 3 1 X 13 Y 4

Z 1 3

55 TABLE 4

Confusion Matrix of Letter Sounds at 6-kHz Bandwidth

Response ABCDEF5HIJKLMN0PQRSTUVWXYZ Stimulus

A 2 2

B 4

C 3 1

D 3 1_ E 12 1

F 3 1

G 3 1

H 4

I 4

J 3 - 1 K 1111 L 13' Ml 3 N 4 O 11 2 P 4

Q 4 R 13 S 4 T 13 U 2 2 V 11 2

W 2 2 X 13 Y 4

Z 1 1 2

56 TABLE 5

Combined Confusion Matrix of Letter Sounds for the Four Bandwidths

Response ABCDEFGHIJKLMNOPORSTUVWXYZ Stimulus

A 10 6

B 11 1 1 3

C 8 1 7

D 4 9 3

E 1 3 11 1

F 1 13 2

G 13 3

H 16

I 14 2

J 11 4 1 ,

K 12 2 4 3 4

L 2 8 6-

M 1 1 ,1^

N . :,. 15 , 1 , O 11 14 P 15 1

Q 13 3

R 1 ,1 .14 S 11 1 13

, 3 13 T ,

U 3 10 3 V 2122 324 W 10 6

X 3 13

Y 16

Z 2 3 11 2 7

57 The combined scores shown in tabl e, it can be seen that the syn- Table 5 indicate that some letters thes is (using /aee/ of diphthong could be identified more easily than /al/ which occurs in both Y and I was others. They also indicate that let- very successful. The low scores of U

ters having the same phoneme sound and ^ 3 indicate that direct combina- at the beginning (such as /ds/ in tion of /i/ and /u/ does not give a letters G and J, and /e/ in F, L, M, good sound of diphthongs /ju/ and

N, S, etc.) or at the end (such as /lU/ , which occur respectively in /i/ in B, C, D, etc.) are more easily thes e two letters. The low scores of confused among themselves. Natural- some other letters were mainly due to ness and distinctiveness of letters simi larity in sound, such as the let- can be measured according to the ter M was being misidentif ied as N identification scores gathered in (see Table 5) , letter J mistaken for this test. Letters ranked in this G, 1 etter U m.istaken to be E, etc. way are shown in Table 6. From this

TABLE 6 . J, Combined Identification Scores of Letter Sounds in Descending Order of Correctness (Perfect Score: 16)

Score 16 16 15 15 14 14 14 Letter H Y N P I OR Sound /eids/ /waee/ /en/ /pi/ /aee/ /ou/ /a/

Score 13 13 13 13 13 11 11

Letter F G S T X ' B E

Sound /cV/ /dsi/ /cs/ /ti/ /eks/ /bi/ /i/

Score 10 9 8 8 7 6 4

Letter A D. C L Z W J

Sound /ei/ /di/ /si/ /ei/ /se/ /dabiu/ /dsei/

Score 4 4 3 1

Letter K V U M Q Sound Aei/ /vi/ /iu/ /em/ /kiu/

58 , ,

Although the identification balanced sentences (IEEE, 1969) were scores of some letters (particularly used. In each list, there were 74 U, M, and Q) seemed disappointingly words contained in ten unrelated sen- low, it was foxmd that after a short tences. Prior to each test, practice period of training, most subjects sentences were given to faniliarize could identify all the letters with- the subject with the test procedure. out error. Test sentences were presented in the manner previously described in a pi- During the progress of the ex- lot experiment (Suen and Beddoes periment, it was observed that many 1973) and there was a rest period of subjects could easily learn to recog- five minutes between lists. The re- nize all the letters correctly. In sults of this test are summarized in the letter test towards the end of the Tables 8 and 9. second session, each letter sound was presented five times in random order It can be seen that the scores to the subject for identification. vary widely from subject to subject. The results of this test are shown A detailed analysis of the data in Table 7. In this test, 12 sub- showed that the subjects who were jects had perfect scores, 3 subjects poor in Braille (particularly those made only 1 mistake out of 130 re- who became blind in their adulthood) sponses, and 1 subject made 8. Of were also relativelv poor in spelled the total number of 11 mistakes, 9 speech; the results resembled those belonged to the 3-kHz bandwidth and obtained from sighted subjects.* 2 to the 6-kHz bandwidth. The ex- As a corollary, those who can read tremely high 99.5-percent average Braille well will also do well with correct identification score suggests spelled speech. We found a number that all letter sounds can be learned of exceptional blind people who could to perfect recognition after a short read spelled speech comfortably at period of training. It also indicates 80 words-per-minute. that the intelligibility of the constructed set of letter sounds is high- Table 9 shows that the average ly resistant to bandwidth reduction. scores of the blind subjects obtained were very high, ranging from. 83 to 94.15 percent correct. It must be TESTS ON A COMPUTER- borne in mind that this was not merely " ' SIMULATED LEXIPHONE ' an intelligibility test, but also READING MACHINE partly a comprehension test because the subjects were tested on sentences In order to test the applicabil- rather than on single words. When ity of the spelled speech code to the subjects were Questioned about con- reading machine, the Lexiphone was tent, they often showed that they simulated on a PDP-9 computer. Sen- understood the subject matter even tences were typed and stored in the when individual words were unintelli- computer and spelled sentences were gible. then generated for identification by the subjects.

The same 16 subjects were used in this test and this session lasted *A spelled speech experiment was about one hour and a half. The fac- conducted some time ago using a tors investigated were divided into a highly motivated group of sighted number of treatment levels. Presenta- subjects (Suen and Beddoes, 1972). tion speed was divided into four lev- The same testing procedure and the els: 45, 55, 65, and 75 wpm. The same type of testing materials (PB bandwidths studied were 3 kHz, 4 kHz, sentences) were used. For approxi- 5 kHz, and 6 kHz. Two variations in mately the same presentation speed pause duration after each word were (55 wpm. in the present experiment examined; in one case the pause dura- with blind subjects and 54 wpm. in tion was fixed, and in the other it the experiment with sighted sub-

was made linearly proportional to the jects) , the average percent correct length of the preceding word. A score of the blind subjects VN^as Greco-Latin square design (Winer, 12.9 3 percent higher than that of 1962) was used in this experiment, the sighted subjects'. the plan of which is shown in the Appendix. Four lists of phoneticallv

59 TABLE 7

Coitibined Confusion Matrix of Letter Sounds for the Four Bandwidths After Learning

Response ABCDEFGHIJKLMNOPORSTUVWXYZ Stimulus

A 80

B 80

C 80

D 2 77 1

E 80

F • :, - 80

' • ', 'G ., 80

H , • .... 80

'''" .' . i r/. , -,-i'' 80

J 80

.". . ._. . 80 ^ - -- - . -;3.i V.:.: :

" -'.---', ".':': :L , : ' . ,: \: m

' "^^ • •' - M . !. , 78 2 ,

N -, 80

• . . 'O . : ^ ..-f 80

'.-.. • 1.F-- ,; : >.-;-:../•; ,-:; 80

Q - ' 2 78

R 80

S 2 78

T 1 79

U 80 VI 79 W 80

X 80

y 80

Z 80

60 TABLE 8

Scores of Spelled Speech Experiment in Percent Correctness

80.3 8 77.22 82.91 87.97

95.57 79.11 85.44 76.58

96.20 96.20 84.18 91.77

88.61 93.04 96.20 99.37

84.08 79.11 90.51 90.51

94.94 91.77 89.87 86.71

92.45 9 8.10 89.87 90.51

81.65 82.91 90.51 95.57

TABLE 9

Summary Data of Spelled Speech Experiment in Percent Correctness

List

90. 19 86.39 87.18 87.92

88.31 87.97 90.19 90.82

Average 89.25 87.18 88.69 89. 87

Speed (wpm) a^ (45) a^ (55) a3 (65) ^4 (75)

93.51 89 .87 87.66 81.65

94.78 91.30 86. 87 84.34

Average 94.15 90.59 87.27 83.00

Bandwidth (kHz) b^ (3) ^2 (^^ b3 (5) ^4 <6^

85.60 87.34 90.19 89.56

88.61 89.87 89.56 89.24

Average 87.11 88.61 89.88 89.40

61 ' .

An analysis of variance of the Thus, it can be concluded that the data indicates that: bandwidth and effects of the four speeds of presenta- list of testing materials are both tion on intelligibility differ sig- significant (p < 0.05), the speed of nificantly from one another. For the presentation is highly significant four bandwidths, only 3 kHz differs (p < 0.01), whereas both the row po- significantly from 5 kHz and 6 kHz. sition and the interval of pause be- During the experiment, the subjects tween words are not significant. also often reported that the 3-kHz The bandwidth effect is significant letter sounds were less pleasant than because lowering the bandwidth re- those having higher bandwidths. duces the pleasantness and clarity of letter sounds. Since different The results of this experiment lists of test materials contain words indicate that the spelled speech code and sentences of varying lengths and is well suited as an auditory output familiarity, it is not surprising to for the Lexiphone. The intelligi-

find the list effect statistically (, bility score is much better when the significant. As expected, presenta- user is allowed to vary the presentation speed is highly significant be- tion speed and rescan the missing cause the subjects had a shorter time words to decode the sentences. The lack of statistical significance of row effect suggests that the order effect WORD LENGTH AND of subject groups in this test is not INTELLIGIBILITY crucial. As far as the interval of pause between words is concerned, al- In this spelled speech experi- though a pause proportional to the ment, many errors occurred with long length of the preceding word gives a words. A calculation was hence made slightly better result than a fixed on the number of errors observed at interval (average score: variable different word lengths. The result interval, 89.32 percent correct; is shown in Table 10. fixed interval, 88.17 percent cor-

rect) , this effect is not significant It can be seen that the longer at the 0.0 5 level. the word, the more likely an error will be made. This is true even when Since presentation speed and a longer pause was given for decoding bandwidth are the two factors in a longer word. In some cases, the which we were most interested, further subjects reported insufficient time tests were made to probe the nature of for the perception of long words. In differences among their treatment other cases, they forgot some of the means. The results of Newman-Keuls letters especially those which oc- test are summarized schematically as curred at the beginning or in the follows: middle of long words. These can be

7£in_ 45 55 65 75 kHz

45 5

55 6

65 4

75 3

p < 0.01 p < 0.05

62 "

TABLE 10

Overall Percent Correctness of Words According to Word Length

Word Length (No. of Letters) 4 or less

Percent correctness 92.13 83. 5f 80.21 76.84 62.50

attributed to two factors. First, typist to eliminate typing errors has since long words occur less fre- been added to the Spellex system. A quently in the English language, block diagram is shown in Fig. 2. It they were less familiar to the sub- operates in four modes of operation, jects. Second, some subjects per- "type," "search," "output," and "com- ceived long words letter by letter, mand." In the type mode, the typist some of which they forgot as the word can type in the text or a command length exceeded their immediate mem- string for mode chanaes. The search ory span. The latter factor became mode allows the operator to locate less important as the subjects became the position of a word or a letter more skillful in decoding spelled for insertion, deletion and other speech. It is expected that through corrections . The output mode com- practice, they will eventually per- mands the typewriter to print the ceive chunks of letters in the form corrected version of the typescript. of syllables or words, rather than one letter at a time.

( START ) TESTS ON THE SPELLEX TALKING TYPEWRITER ENTER AUDIBLE TEXT The applicability of the spelled speech code to the Spellex talking typewriter was tested using several groups of blind students from, a YES HIT RUBOUT BUTTON school for the blind. A block dia- TO REMOVE ERROR gram of the computer-simulated Spellex typewriter is shown in Fig. 1. It gives the typist an auditory feedback HIT CONTROL (J. each time a key is depressed. For ex- TO PLAY BACK LAST ample, on pressing the letter "a," 10 CHARACTERS the typist hears the voice say "a. In order to improve the performance, an editor which enables a blind NO

COMPUTER PRINTER TEXT STRING/ COMMAND TELETYPE

TYPEWRITTEN WRITER PRINT TEXT LETTER SELECTION MANUSCRIPT PHONEME DATA OR COMMAND AND RULES FOR YES SYNTHESIZING LETTER SOUNDS RECALL AUDIBLE TEXT

CORRECT ERRORS

TELEPHONE LINK

Figure 1. Network of the Spellex Figure 2. Operation of Spellex Automatic Talking Typewriter

63 . .

The command mode gives the typist full The encouraaing experimental re- control over the selection of operat- sults have led us to design and build ing modes a digital spelled speech generator for use with the Lexiphone reading machine Several Spellex networks have and the Spellex talking typewriter in been installed in Vancouver and ^''ont- place of the computer. This would real and we have been encouraged by make both speech aids economically blind typists' success with this ma- feasible. We have finished building chine (Beddoes and Suen , 1973). It the first prototype spelled speech has been found that using the Spellex generator and a picture of the unit is typewriter improved the typing accu- shown in Fig. 3. Presently we are ex- racy. For a group of more advanced ploring the applications of the spelled typists, both the accuracy and the speech unit to other instruments and typing speed improved providing furth- we shall report our progress in the er evidence that the spelled speech future code is a useful aid for blind typists, ACKNOWLEDGEMENT DISCUSSION This research was supported by A series of experiments have grants from the National Research been conducted to test the feasibil- Council and the Medical Research Coun- ity of using a synthesized spel.led cil of Canada, the Vancouver Founda- speech code as an output for both the tion, the Mr. and Mrs. P. A. Woodwards Lexiphone reading machine and the Foundation and the B. C. Telephone Spellex talking typewriter. The re- Company. Ching Y. Suen also wishes sults indicate that spelled speech is to acknowledge the financial support easy to master and it provides a use- from the Ministry of Education of ful communication link between the Quebec. blind and machines.

Figure 3. A Prototype Spelled Speech Generator and the Talking Typewriter

64 )

REFERENCES

Beddoes , M. P. Possible uses Suen, C. Y. & Beddoes, M. P. of a printed braille reader Dichotic presentation and time with spelled speech output, compression of the spelled Proc. of the International speech lainguage. Paper pre- Congress on Technology, The sented at the Fourth Ajinual American Foundation for the Audio-Tutorial Congress Con- Blind, 1963, pp. 325-42. ference, Columbia, South Caro- lina, November 2-4, 1972. Beddoes, M. P. & Suen, C. Y. Experimental study of the Suen, C. Y. Computer simula- Spellex editing talking type- tion, development and evalua- writer. Proceedings of the tion of a high speed spelled International Conference on speech code. Ph.D. Thesis, Cybernetics and Society, Department of Electrical En- Boston, November 1973. gineering, University of British Columbia, Vancouver, June 8, IEEE, 1965 revised list of 1972. phonetically balanced sentences (Harvard sentences). IEEE Suen, C. Y. & Beddoes, M. P. Trans. Audio and Electro-acous- Development of a digital tics, 17, 239-246, 1969. spelled speech reading machine for the blind. IEEE Trans, on

Metfessel, M. , & Lovell , C. Biomedical Enaineering, Vol. Spelled speech as output for Br''E-20, 452-459, November 1973. an automatic reader. Research Report, Psychology Department, University of Southern Cali- 10, Winer, B. J. Statistical fornia, Los Angeles, 1961. Principles in Experimental Design. New York: McGraw- Metfessel, M. Experimental Hill, 1962, 575-577. studies of human factors in perception and learning of spelled speech. Proc. of the 11, Beddoes, M. P., Suen, C. Y., International Congress on Tech- Dixon, B. A., & George, R. C. nology, The American Foundation Spellex: An automatic talking for the Blind, 1963, pp. 305-08. typewriter for the blind. 10th International Conference

Metfessel, M. , & Lovell, C. on Medical and Biological En- Guidelines for developing an gineering, Paper 38.11, alphabet to synthesize spelled Dresden, August 1973. speech. Bulletin of Prosthetics Research , Veterans Administra- tion, U.S.A., Spring 1972, pp. 154-76. (This is the m.ost complete account of Metfessel-

Lovell ' s work .

65 .

APPENDIX

Plan of Spelled Speech Experiment with 16 Blind Subjects

^3^3 ^4^1 ^1^4 ^2'"2

a b a b a b a b 1 2 2 4 3 1 4 3 a b a b a b a b 2 1 1 3 4 2 3 4

a b a b a b a b 4 4 3 2 2 3 1 1

a b a b a b a b 3 3 4 1 1 4 2 2

a b a b a b a b 1 2 2 4 3 1 4 3

a b a b a b a b 2 1 1 3 4 2 3 4

a b a b a b a b 4 4 3 2 2 3 1 1

Interval of pause between words,

I, = fixed interval = 4.4 T

Ij = interval increases linearly with word length = m L^ + 2 T

where T = pause between letters

L = word length

and m = 2.4 T^/4.4 = 0.545 T^

G: Subject group, there were two subjects per group. L: List of testing materials, lists 13, 16, 22, and 45 of PB sentences were used. a: Speed of presentation, four speeds were used, viz. 45, 55, 65, and 75 wpm. b. Bandwidth of letter sounds, four bandwidths were used: 3, 4, 5, and 6 kHz.

66 .

OBSTACLE DETECTION WITH AND WITHOUT THE AID OF A DIRECTIONAL NOISE GENERATOR

N. V. Clarke, G. F. Pick, and J. P. Wilson*

Abstract : The nature of obsta- It has been known for a long cle detection by facial vision is time that blind people can sense an discussed and the main underlying obstacle at a distance and move clues are identified as acoustic around it without needing to touch images and reflection tones. A di- it. The physical basis for this rectional sonic torch was developed ability, however, has only recently to enhance these natural clues. become well understood. The reason Static experiments and an obstacle for this would appear to be that the course were devised to assess per- subjective impressions elicited by formance in four groups of subjects. the proximity of an obstacle bear The long-time blind adults and chil- little relationship to the physical dren showed many more unaided static stimuli involved. Kohler (1952, detections than the newly blind or 196 7) , who was particularly inter- blindfolded sighted subjects. With ested in the basis of these subjec- the aid, many more detections were tive impressions, has reported de- made by all groups which then showed scriptions such as being "similar similar performances. On the obsta- to a slight touch or pressure," "a cle course the blind children and kind of tactile sensation somewhere blindfolded subjects detected most near the eyes," and "like bumping into obstacles and the newly blind subjects a spider's web, only much lighter"; least, but the blindfolded subjects blindfolded-sighted subjects and the were much slower in completing the recently blind spoke of "darkening course. Use of the aid approximately before the eyes." The nature of doubled both the number of detections these impressions has given rise to and the distance at which detection the term facial vision. However, occurred. Aided and unaided detec- Dallenbach and coworkers demonstrated tion improved throughout the tests. convincingly that the basis is auditory and not tactile. Earlier workers had apprently even entertained INTRODUCTION the possibility of some "sixth- sense" being involved. It is neces- The general aim of the investi- sary to understand the basis of ob- gation reported here was to obtain stacle detection if a device is going some quantitative measures of sonic to make optimal use of these natural obstacle detection with and without abilities the aid of a directional sonic "torch," and to ascertain whether practice with a device of this type, The Obstacle Sense which essentially enhances the natural clues used unconsciously by skilled Supa, Cotzin, and Dallenbach blind people, could lead to improve- (1944) provided a survey of the early ments in unaided performance. studies and presented their own series of experimental investigations. They showed that blind and blindfolded sub- *University of Keele , Department of jects were poor at detecting obstacles Communication, Staffordshire, England.

67 ,

when walking without shoes on a car- various self-generated noises such as peted floor, and were unable to de- footsteps. The performance of the tect obstacles at all with their ears blindfolded (sighted) subjects was blocked or masked by sound. On the slightly worse under some outdoor other hand they showed that covering conditions than indoors. This worsen- the exposed skin areas with felt, al- ing was attributed to the higher legedly to cut out any "pressure" ambient-noise level, and wind noise. clues, made no difference except that which could be attributed to the Apparently inspired by the im- slight acoustic absorption. More im- pressive echolocating abilities of portantly they demonstrated that sub- bats and porpoises , and also the jects in another room could detect technical capabilities of sonar and

obstacles over headphones , even mon- radar, Kellogg (1962) undertook an aurally, when the experimenter car- investigation of what he called the ried the microphone towards the ob- "sonar system of the blind." He stacle. Worchel and Dallenbach (1947) measured the just detectable change showed that deaf -blind subjects were in the distance of a small disc from not able to anticipate obstacles. the face of the subject. The subjects Thus it appears that audition is nec- were instructed to make their judg- essary and sufficient for obstacle ments with the aid of vocalizations. detection, although there m.ay occa- <- They noticed during these experiments sionally be special conditions where that the most effective vocalizations other clues such as wind patterns contained a number of sibilant and/or odors, heat radiation, etc. could be •- stop phonemes. Generally, in previ-

of s ome us e . ' ;- ous experiments little difference had been found between the echolocating Cotzin and Dallenbach (1950) in- performance of blind subjects and vestigated which frequency region was blindfolded subjects who had had. a most important by using pure tones minimal period of training. In this and a similar microphone and head- experiment, however, the trained phone arrangement. Continuous pure blindfolded subjects were almost in- tones between 125 Hz and 2 kHz yield- capable of estimating relative dis- ed few detections with an increasing tance, while the blind subjects could number for frequencies from 2 to detect differences in distance of be- 10 kHz. It would appear from the tween 10 to 20 cms at 60 cm for a descriptions of the subjects that 30-cm diameter disc. He also reported they were making use of the increas- " that subjects showed some ability to ing peak-to-valley ratio in the discriminate between hard and soft standing-wave patterns as the loud- textures. Rice, Feinstein, and

speaker and microphone approached Schusterman (1965) , and Rice (1967) the obstacle, although Cotzin and ' ' continued with the study of size and Dallenbach suggested that the clue distance factors in echo detection. was an increase of pitch which they Blind subjects were reauired to de- attributed to the Doppler effect. tect discs of different sizes at a Subjects found detections far easier variety of distances again using vo- when white noise was used as signal, calizations. It was found that each and reported a distinct increase in disc subtended an angle of approxi- pitch. Twersky (1951) investigated mately 5° at the distance at whicn the detection of obstacles with the 50 percent detections were obtained. aid of a directional high-frequency tone generator. Although he reported Next, Rice generated artificial that an interrupted tone yielded bet- hisses (continuous white noise) and

ter detection (see below) , it appears clicks (pulse trains) using a fairly that in the majority of his experi- directional loudspeaker placed in ments he used a continuous tone. He front of the si±iject's mouth. None stated that obstacles were detected of the artificial or oral signals was by means of the increasing intensity found to be consistently better than of the echo, together with an increase the others. After training, some of in pitch which he thought resulted the subjects were able to discriminate from this increased intensity (see slightly between circular, sauare, Stevens [1935] but also below). Am- and triangular surfaces at distances

mons , Worchel, and Dallenbach (1953) of less than 90 cm using vocaliza- carried out experiments on the detections. Rice also noted that, in gen- tion of obstacles with the aid of eral, detection was as good monaurally

68 as binaurally, but that a mild hear- It would appear, however, from the ing loss generally led to some de- results of Experiment III below that terioration of detection ability. this cannot be a particularly influen- tial factor although it may be the Although there have been observa- basis of the shape discrimination re- tions of a "pitch" effect, or "reflec- ported by Rice. The obstacle may also , tion tone " in the neighborhood of selectively absorb and reflect certain broadband sources of noise such as frequencies, and because of its con- waterfalls predating Dallenbach's struction exhibit resonances and anti- reports by a century, and the physi- resonances at certain freauencies. A cal principles involved have been galvanized iron pipe in the present understood for a similar period, this study could auite easily be discrimi- phenomenon does not appear to have nated from a plastic one of the same been specifically related to obstacle geometry because of the "tinny" timbre detection by the blind until the re- of its reflection. It is not known port of Bassett and Eastmond (1964). whether subjects are able to use They showed that interference between spectral information caused by dif- direct sound from a white noise source ferent surface textures but this and that reflected from a surface gave would appear to be of secondary im- a series of spectral maxima at multi- portance. ples of the frequency, c/2d, where, c, is the velocity of sound and, d, In addition to the various spec- the distance of the microphone or ob- tral clues which will be available server from the reflecting surface. whenever broadband ambient or impul- This arises because the time delay sive sounds are present there are between the direct and reflected temporal clues such as separately sound is 2d/c. Approaching the sur- perceived echoes when the delay is face shortens the delay and gives long enough and spatial clues such more widely spaced spectral maxima as acoustic images for both short and with a higher pitched sensation. long delays. Although acoustic images There has been considerable interest like optical images should be situ- in the auditory mechanism underlying ated as far behind a plane-reflecting this pitch sensation which Bassett surface as the source is in front; in and Eastmond incorrectly attributed practice they sound to originate from to difference tones due to intermodu- the surface itself. Presumably this lation distortion within the ear is because the rate of change of in- (Fourcin, 1965; Wilson, 1967, 1974; tensity at the ear produced by ob-

Bilsen, 1968) . It also appears to server and source moving up towards be related to the "sweep-pitch" ef- the obstacle would be the same as fects produced by pulse pairs (Thur- that for a stationary source at the low and Small, 19 55) . This signal surface. For an acoustic image to has the same power spectrum as noise be perceived it is necessary that the plus delayed noise and of course is ears should be able to resolve source relevant to echolocation when impul- and image spatiallv, and that the sive noises such as footsteps , hand Haas or precedence effect, in which clapping, or tongue clicks are uti- sound appears to originate from the lized. No pitch modification can, direction of the earlier of two sound of course, arise in this way for a sources (see Gardner, 1968) should pure-tone signal. Thus it would ap- not predominate. The latter effect pear likely that the slight pitch ef- may be a contributory factor towards fects with tones noted by Cotzin and localizing the acoustic image at, Dallenbach, and Twersky could have rather than behind, the reflecting been due to residual noise in the surface. The range of greatest prac- signal or spectral spread due to gat- tical interest to the blind person is ing the tone. Wilson (1967) showed that in his immediate neighborhood that the extreme limits of obstacle where echoes are not separately per- detection via the reflection tone ceived. Thus two clues, reflection would be 1 cm to 5 cm. tones and acoustic images, stand out as being chiefly responsible for ob- The reflector may also modify stacle detection. In general, geo- the spectrum of the signal in other metrical arrangements of source, re- ways. Small or medium sized obsta- flector, and observer which favor one cles will reflect a greater proportion of these clues tend to obscure the of high, rather than low frequencies. other. In an ingenious experiment

69 . that introduced both of these clues that is, at the same ear, although (though not separately controlled) the sound source need not be direc- Kohler (1967) was able to elicit the tional. Acoustic images are better sensation of facial vision. Subjects detected with longer delays, the were initially required to detect ob- minimum possible direct sound from stacles with the aid of a small-sound the source, preferably gated noise source. Kohler found that if instead or pulses, clear directional separa- of a real obstacle he ipoved up a section between source and image, and a ond sound source relaying the same fairly directional sound source. The sound they detected it as if it were directionality required is something a real obstacle and obtained similar of a compromise: a highly directional subjective impressions. He was able source gives a very long range and to show in other experiments that good spatial resolution when scanned anaesthetising the skin had no effect but increases the chance of missing on these sensations. a nearby small object because the chance of scanning it is lower. A It would appear plausible that directional source also contributes tactile sensations arise centrally ; to a reduced direct-sound path when by a process of association or con- pointed at the obstacle, favoring ditioning. Bumping into an object acoustic image perception. results in strong tactile sensations and therefore detecting or sensing an object by some unidentified means Previous Sonic might be expected to lead to a mini- Mobility Aids fied and perhaps graded tactile sensation, akin to pressure on the face, The discovery that humans can which increases as the object is ap- ',' echo- locate well in certain circum- pro ached. stances, albeit poorly in comparison to the ability of bats and dolphins, Now that the basic clues for ob- has prompted many people to devise stacle detection have been identified methods of helping blind people to it is possible to consider both the make greater use of this ability. limitations and how detection might Som.e of these aids have emulated be optimj-zed. For either clue it is bats to the extent of usina ultra- necessary that the obstacle should sonic signals, althouah of course the reflect sound into the ear of the ob- "displav" must be transposed into the server. The obstacle has to be suf- audible range (Kay, 1963, 1966; and ficiently large compared with the '' Beurle, 1972) wavelength of sound and appropriately orientated for reflection (i.e., for It is reported that Kohler used the acoustic im.age to be "visible" in a mechanical clicker (of the type its surface) . For small objects of used in a child's toy, a "cricket") a few centimeters in diameter, only in order to demonstrate the detection the highest audible frequencies will of echoes. Griffin, in 1943, elabo- be reflected and to detect wires it rated upon this theme by placing the is necessary to use ultrasonic fre- mechanical clicker at the focus of a quencies. Although, in theory, it parabolic reflector (see Griffin, 1960). should not be possible to detect an '"' This made the device more directional, external corner with self-generated giving better resolution of obstacles. or carried-sound sources , because the However, it would seem that neither sound will be reflected away from the of these devices received serious observer, in practice it is frequently evaluation as aids for the blind. possible due to ambient noises from The 10-kHz continuous tone "flash- other directions or indirect reflec- sound" of Twersky reported above was tions of the source. evaluated by a small group of blindfolded subjects who learned to per- The conditions which favor detec- ceive obstacles far faster with this tion of reflection tones tend to be highly directional aid than with am- complementary to those for acoustic bient sound alone. Twersky (19 53) images. The former requires shorter later went on to develop two further delay times, approximate equality be- types of echo-location aids. The tween direct and reflected sound lev- first was a directional, mechanical els, preferably continuous noise, and clicker, very similar to that described arrival of the sound from, approximate- by Griffin, the second being a whistle ly the same direction as the source. at the focus of a parabolic reflector.

70 blown by a finger-operated rubber the results of their experiments us- bulb. Twersky reported that his sub- ing an obstacle course that the de- jects, including one blind man, pre- vice aided those children with poor ferred the whistle to both the click- navigational ability, but provided er, and the flash-sound. They thought no improvement for those already adept that the clicker did not provide good at navigation. close-range detection, although it was better than the other devices in the region of 6-15 m from the obsta- Sonic Aid Used cle. They preferred the whistle over in the Present Study the flash-sound, as it provided a wide repertoire of signals, and it The sound generating device allowed rapid pulsing of the signal. used in the present experiments is They felt that the whistle was the shown in Fig. 1. It consisted of best device for use at distances of two parts: the hand-held "torch" less than 3 m. Beurle (1951a and (approximately 24 cms long) , con- 1951b) constructed an electronic aid taining the sound-source and the using a piezo-electric transducer associated electronics, connected situated at the focus of a parabolic by a lead to the battery case which reflector, actuated by capacitor dis- was carried on a shoulder strap. charge. Thus, this was an electronic The torch weighed approximately version of the clicker first used by 470 g. Griffin. Beurle evaluated the device using a large variety and number of The signal used in this study blind subjects. VJhile subjects were was broadband noise, obtained by initially very enthusiastic, many la- amplifying thermal and semiconductor ter reported that the device made noise. The transducer was an elec- little difference in the navigation trostatic loudspeaker element (10 x of familiar routes. Beurle felt that 12 cm) originally designed for use the device would not be of sufficient in a high quality headphone (Wilson, usefulness to be carried at all times 1968). This has the advantage of by the blind person, but might find producing a plane wave front for some use as a training device for sound wavelengths less than its blind children. Graystone and McLen- shortest dimension, giving a direc- nan (1968) also developed a clicker tional beam of sound. Larger and type of device using a small moving- smaller diameter radiators were also coil loudspeaker as a transducer, tried giving narrower and broader which was evaluated by a group of distribution patterns respectively. blind children. It was apparent from From these tests it appeared that

Pulsed/Continuous push-button

Electnsf otic loudspeaker

Battery holder

Figure 1. Sound Generating Device

71 .

the present radiator, giving a beam- holding the device at hip level

width with half-energy points sepa- proved to be most advantageous . When rated by 25° over the range 5-20 kHz, held at this level a large object at would be the best compromise. At a 1 m distance returned a first reflec- distance of 5 cm from the transducer, tion which is approximately 6 dB above on-axis, the energy spectrum of the the level of the direct plus multiple- noise was flat with a sound pressure reflected sound at the subject's ear level of 70 dB in the frequency re- (the difference can be as great as gion 1-20 kHz (±3 dB) . The noise 16 dB in anechoic surroundings) was normally gated, at a rate which was continuously variable from about A simple alternating vertical 1 to 20 pulses-per-second. The gat- and horizontal scanning procedure was ing function had an equal mark/space evolved covering that part of the en- ratio, but could be switched out for vironment which was in the direct path continuous noise by pressing a button of the subject. on the handle. Preliminary trials indicated that This facility was incorporated the prototype device did not perform

because Twersky's investigations had as well outdoors as indoors , partly shown that a click signal was most because wind could interfere with the useful at long range. At short range operation of the transducer. Also, the echo would return while the .noise because the much higher levels of was still present, while at long ambient noise frequently found out- range the echo would return during side impair detection as found by

the silent period. Twersky's experi- Ammons et al. (1953) . ments indicated that the transient signal was useful for ranges in excess of 7 m. Using a gated-noise EXPERIMENTS signal, the echo would occupy the 't silent gap at a range of 7 m if the The aim of the experiments was gating rate was approximately 14 to evaluate the device as a training pulses-per-second. In preliminary aid for echolocation and as an every- experiments it was found that a gat- day mobility aid. ing rate between 6 and 10 pulses-per- second provided the optimum compromise between long- and short-range TABLE 1 detection. In the majority of the experiments reported the rate was Obstacles set at 6 pulses-per-second. The ability to switch to a continuous noise was provided, as this gave the clearest reflection tone for short-

range detection. Shape Dimension Surface Material

The method of carrying the de- 180 cm high 90 cm wide

- Plane board - . fSOcm wide sized hardboatd vice had an effect on the ability to L detect obstacles. If the device was held too close to the subject's ear 160 cm high 23cm di am. galvanised iron Cylinders J the unreflected sound tended to mask 140 cm high 7cm diam polished plastic the weaker reflections, although as [20 cm pointed out above, the strongest re- Step ups/ _ ' 90cm long, 1 = 13 cm sized hardboard flection tones are obtained if the Kerbs 1 5 cm direct and reflected signals have equal intensities. In an earlier m prototype there was no rear cover on the generator. Such an arrangement gave a figure eight distribution pat- Details of the obstacles and the tern with no radiation to the ears subjects used are shown in Tables 1 when held by the side, and also gave and 2 respectively. The only obsta- a clear reflection tone when held at cle which subjects found very easy arms length. However the back radia- to identify was the 23-cm diameter tion proved to be confusing at other galvanized iron pipe, whose echo had positions and during scanning and was "tinny" timbre. eliminated. In the present form. a

72 .

TABLE 2 different obstacles by measuring how frequently they could be detected at Subjects different distances; both by using the aid, and by means of ambient noise alone.

No of ' Where Group Deicripton Age yean leifpd Method blind In a preliminary experiment, the Blind children 2 boyi-2gif|i 12-14 yrv >I0 Blind ichool placed near the center of library subject was the experimental room and asked to Blind «ofkhop' Long-time Hind 4 men middle >10 a\^mb\y holl rotate on the spot with the aid point- oged ofdce %u.te ing directly in front of him. He was asked to report any echoes that he Shorl-time blind 4 men m.ddle 1-3 Blindwo'ltihop' perceived. These echoes were, of oged - oHice iuite course, excited by walls and fixtures 4 men Keele ' Blindfolded 3 oulhori 25-45yfi Blind Mhool within the room. Steps were taken to iighted 1 technician library reduce them, either with the aid of curtaining, or with auxiliary re- flectors. In this way, the more * No subject was blind from birth prominent of the room-excited echoes

1st mentioned venue for eiperiments 1-111 were eliminated. The direction of

2nd IV any remaining echoes was then noted. These directions were never used for the siting of obstacles to be detect- The obstacles were designed to ed, and anv detections made in those simulate those found in a normal en- directions were ascribed to room re- vironment, and in particular street flections. There then followed a furniture such as telephone poles, short training period in which sub- stop posts, and curbs. bus jects were requested to report the moment at which an obstacle was si- The hearing of all subjects was lently placed before them. The aim One of the long-time blind tested. of this training was to reduce the persons showed a considerable hearing number of false detections made and loss at frequencies in excess of the results indicated that in the 4 kHz in both ears . One of the blind- majority of cases this was achieved. folded subjects showed in both ears a sharp notch of about 40 dB in his audiogram in the region of 4 kHz but Main experiment procedure . only a 10-dB loss for lower and high- While the subject was outside the frequencies. The results for both er experimental room, the experimenter of these subjects were examined indi- arrayed a random selection of one, vidually, and they showed a slight, two, or three obstacles (including but statistically insignificant, re- the possibility of two similar ob- duction in their ability to detect stacles) at randomly selected dis- obstacles by means of echolocation tances (in multiples of 60 cm) , and For the first subject this appeared at randomly selected directions rather surprising, considering the around the point at which the sub- findings of Rice. Experiments were ject was to stand. The front sur- out serially in the order carried faces of boards and kerbs were placed given, and required about two months in normal fashion to the subject. Each subject to complete the series. Occasional trials were interspersed two 40-minute ses- was tested during in which no obstacle was present. sions each week. The experiments Trials with and without the aid v/ere were carried out in ordinary rooms of interspersed (approximately two tri- medium reverberation, although, as als with the aid to one without) . On may be seen in Table 2 subjects were trials without the aid, subjects were the same location. not all tested at encouraged to make any noise they felt would be helpful. Subjects were on six occasions with each ob- -- tested Experiment 1 stacle at each of the distances. discrimination Static Once the obstacles had been placed, The object of this experiment was to the subject was brought to the test determine the range of detection for point by a circuitous route, and

73 Kerbi Boards & Pipai Boards & Pipes Aided Aided Unoid«d

O 1 2 3 4 O Distance between subject & obstacle

Figure 2a. Percentage of Obstacles Detected as a Function of Obstacle Distance Parameter: Obstacle Type; 30-cm wide board; D olS-cm wide board; • f 23-cm diameter pipe; o o 7-cm pipe; A 20-cm kerb; A A 13-cm kerb; 5-cm kerb. The bars indicate mean standard error. rotated on the spot in order to dis- blind people, the experimenter could orientate him. He was instructed to usually ascertain that the subject rotate on the spot and report any ob- was perceiving the relevant reflec- stacles that he detected. The experi- tion by hearing its echo himself. menter, wearing soft soled shoes, stood behind the subject, and noted detections as they were made, during Results one revolution of the subject. It is interesting to note that the experi- The percentage of detections for menter often heard echoes which the each obstacle, at each distance, was subject missed. When a subject had calculated, and is plotted as the mean made a detection, he was asked to stop over each group of subjects in Figs. and point towards it. Although it was 2a and 2b. As might be expected, those found that the ability to point pre- obstacles which reflected the most cisely at an obstacle is poor in some sound, due to their dimensions and

74 Aoordt & Pipvi K ef bi Boordi 4 Pip«» Aidad Aided Unoidad

lOO

No obstacl*» Ntwiy d«f«cted unaided Blind b y newl y blind

lOO

Blindfolded

S ighted

Distance between subject & obstacle

I in metres)

Figure 2b. Percentage of Obstacles Detected as a Function of Obstacle Distance orientation, were detected most easi- had much less experience of blindness ly, the 30-cm wide board being de- than the other. Hence, it might be tected on the greatest percentage of postulated that those who had been trials, and the 5-cm high kerb the blind for some time had developed the least. The hypothesis that the sub- ability to detect obstacles by means jects when using the aid, were men±>ers of ambient sound. This skill appar- of the same statistical population was ently takes some time to develop when not disconfirmed at the 5-percent lev- no explicit training is provided, as el of significance. Similarly, the the newly blind (blind from 1 to 3 hypothesis that each of the groups years) had not yet acquired it. It were from the same population was not is interesting to note that those un- disconfirmed at this level. However, skilled in unaided echolocation , per- without the aid the groups fell into formed just as well as the other two populations: the newly blind plus groups when using the aid. the blindfolded, who detected very few obstacles unaided, and the blind If one assumes that the limiting children plus the long-time blind, factor for perception of the boards who did rather better. No subject is their width, then it is clear, detected a kerb unaided. The most especially when the aid is used, that distinctive difference between the a given percentage detection is not two groups of subjects is that one achieved for the same angle of

75 . subtense when differing widths and a blind person in detecting, for ex- distances are employed. This appears ample, partially open doors. to be contrary to the results of Rice, Feinstein, and Schusterman (1965). The reason for this might well be that Method the obstacles used in the present study are somehwat larger than those The maximum angle of tilt was used by Rice et al . , and are easily measured for the 90-cm and 30-cm wide detected at a greater distance. Above boards at distances of 1.2, 1.8, and a certain size a further increment 3.6 m. The board was placed at the would not be helpful except for very given distance in front of subject, low-frequency sound. At distances at standing at the center of the room, which detection becomes uncertain, and perpendicular to the line between perhaps the limitation is imposed by the center of the board and the sub- factors other than the just percepti- ject. A 360° protractor was placed ble audible angle, that is by the re- beneath the center of the board, and flection being masked by ambient and the board was slowly rotated about unreflected noise. this vertical axis until the subject reported that he could no longer de- There is some indication that tect the board. The angle was noted. the detectability of kerbs may be The board was then slowly rotated greatest in the region of 1 m. ' This back as the subject continuously observation prompted a subsidiary scanned across the target area, un- experiment using the 13-cm high kerb, til the subject once again reported which showed, for a single subject, detection. The mean of these two that detection was considerably worse angles was taken as the threshold. at a distance of 0.2 m than at a dis- Each board and distance was selected tance of 1 m. This is presumably beat random for measurement. The thres- cause the comer between the vertical hold for each subject at each distance face of the kerb and the floor re- was the mean of six such measurements. flects less energy when the subject is close to the obstacle, than when

"'-, - -; , , \'" he is at some distance. More in- Results " forma'-ion will be given on this point in the report of Experiment IV. The resulting threshold is shown in Fig. 3. This figure shows the results averaged over all subjects, as Experiment II — the x^~test showed no significant dif- Maximum Angular Displacement ference between the results for the four groups (at the 5-percent level

It is apparent that a plane ob- of significance) . The results are stacle which is detected with the aid rather disappointing for these two of spe aular re flection may not be de- obstacles, as even at a distance of tected if it is displaced by even a 1 m the 90-cm board became undetect- small angle from the perpendicular to able when turned through an angle of the listener. The exact angle at 20°. Calculation of the just per- which detection is lost depends upon ceptible angle expected from specular the geometry of the reflector and the reflection indicate that the empirical relative positions of the sound source, threshold is somewhat in excess of and the listener's ears. However, that calculated (ranging from about specular reflection is not the only 10 percent at the shortest distance means whereby sound could reach the to about 80 percent at the longest) listener. The angle at which the Hence, it would seem that non-specular plane may be turned may be increased reflection does play a role in the because the obstacle is to some ex- detection. This inability to detect tent a diffuse reflector of sound, or angled surface would seem to be a because diffraction has led to side- limitation of acoustic obstacle de- lobes of sound being produced at the tection, including ultrasonic fre- edge of the plane. Hence, it was of quencies . some interest to investigate the maximum angle of tilt for some of the plane boards used in the present study. The ability to detect at relatively large angles could be of interest for

76 s .

' — ' ! 1

- 1\ - - \l -

- - ^^^"^-^

- I i 10 If T ^"^^^T o^ « i " 1 '^^ T I

- 1 ~~"~~^--. """ — -6

- -O i f- .1

1 1 , .. 1 _1 1 —

Dislaoce between subject and obstacle, x

(in metres]

Figure 3. Maximum Angle at Which Obstacle Could Be Detected as a Function of Distance. Inset indicates experimental set-up. Upper curve: 90-cm board; lower curve 30-cm board. Bars indicate the standard error about the mean.

Experiment III -- and a 7-cm pipe. The comparison of Relative Distances obstacles was carried out in the order given above, but all other vari- The aim of this experiment was ables were randomized. Subjects were to determine how well the different not told which obstacles were being groups of subjects could determine used or at what distances. Each sub- the relative distances of various ject made each comparison on six sepa- obstacles, the size of which they rate trials. The task of the subject did not know. With perception of was to report whether the right-hand acoustic images it is possible that obstacle appeared further, equidis- simply the strength and direction of tant, or nearer. A "right, further" an echo is perceived so that obsta- response was scored +1, "same dis- cles at different distances having tance" as 0, and "left, further" as different sizes, geometries, and sur- -1. Although, unfortunately, the de- face textures may be indistinguish- sign of this experiment was faulty in able. If, however, reflection tones that the right object was never near- are being used effectively, distance er, and subjects were allowed to re- judgments should be correct and un- spond "same," it appears that sub- influenced by obstacle size. jects were not greatly influenced by these shortcomings and the results are worthy of comment. Method

Two obstacles were placed at Results about 15° on either side of the subject's midline and perpendicular to The mean score for all of the him (while he was absent) . The left- subjects is presented in Fig. 4. The hand obstacle was placed at one of ordinate corresponds to the propor- three distances: 1.2, 1.8, and 2.4 m. tion of "right, further" responses. The right-hand obstacle was placed at Once again there was no significant an equal or greater distance than the difference between the different left-hand one. The obstacles to be groups. This is contrary to the re- compared were: 1.) two 30-cm boards, sults of Kellogg (196 2) who had shown 2.) two 15-cm boards, 3.) a 30-cm and that experienced echolocators could a 15-cm board, or 4.) a 23-cm pipe make fairly accurate serial discriminations of whether an obstacle had been

77 .

clue used to make the judgment of 1 1 1 1 A relative distance. When the two ob- *l ~ .—-^* stacles are identical (as in Fig. 4A /'' for two 30-cm boards, and Fig. 4B for - two 15-cm boards) then sx±ijects have - little difficulty in reporting the / 1 true relationship of the obstacles. / 1 / However, when the 15-cm " board is com- - pared with the • 30-cm board, subjects " /^^-> almost always report the larger board - ' to be nearer, when in the A^ ^~~~~V fact, two - >-(§ - boards are at the same distance. The — boards are reported to be at the same distance when the larger board is at -1 . 1 ^--T"^ twice the distance of the smaller. This result would suggest that the distance judgment is related to the solid angle subtended by the obstacle and that under these static conditions reflection tones do not play an important role. This conclusion is, at first sight, contradicted by the result of Fig. 4D. Here the 23-cm diameter pipe is compared with the 7-cm diameter pipe. To a far greater extent than for the two boards of different size, the siabjects were capable of reporting the true relation between the distances of the pipes. Two explanations may be suggested for this observation. IX 2x 3x 4 X 1x 2x 3x 4x The first is that the energy re-

Distonce of obstacle to ttto right, y flected radially by a cylinder less

(in mattes) than doubles as the diameter of the cylinder is doubled. Thus, the two cylinders might appear more similar Figure 4. Relative Distance Judg- than the boards of different size. ments for a Pair of Obstacles. A simpler explanation is that in The left-hand obstacle was set previous experiments a few of the at one of three randomly se- subjects had commented upon the lected distances, 1.2 m, 1.8 m, "tinny" timbre of the 23-cm diameter or 2.4 m, while the right-hand galvanized iron pipe, and this might obstacle was varied in multiples have provided them, with clue as of this. Inset indicates a to arrange- which obstacles were being tested. ment for (A) two 30-cm boards, It would seem from the results of (B) two 15-cm boards, (C) a 15- this experiment that blind people cm board (left) and a 30-cm board although accurately detecting the (right) , (D) a 7-cm pipe (left) presence of an obstacle, do not have and a 2 3- cm pipe (right) a clear idea of its distance if its size is unknown and, therefore, would be unable to build up an accurate moved closer, or further away, while mental map from a single "viewpoint." an inexperienced echolocator (blindfolded subject) lacked this ability almost entirely. A result which is Experiment IV-- common to all of the comparisons test- Obstacle Course ed is that the proportion of judgments: "right, further" a function as The obstacle course was devised of the relative distance of the right- to evaluate the device in a situation most obstacle is fairly independent more closely approximating everyday of the absolute distance of the left- situations. It was also hoped that most obstacle. This would seem to be this might be useful in allowing com- an indicator that the solid angle parisons with evaluations of other subtended and therefore perhaps the mobility aids. The performance of relative sound intensity might be the the subjects was determined when

78 negotiating the course both with and practical use of the aid, and with without the aid. There were two ma- the experimental room which was later jor reasons for this. One was to to be used for the obstacle course. provide a base-line level for com- Once the subject had become reason- parison of performance with the de- ably practiced and knew what path he vice, and the other was to provide was to follow, he was led out of the a measure of whether experience gained room, and obstacles were arrayed with the aid led to an improvement in around the course. A typical obsta- performance using unaided echoloca- cle course layout is shown in Fig. 5. tion. In studying results presented On each trial one of each of the fol- for this experiment, it should be lowing obstacles was used: a 30-cm borne in mind that all subjects had board, a 15-cm board, a 23-cm pipe, already gained quite considerable ex- a 7-cm pipe, a 20-cm kerb, a 13-cm perience in using the aid while act- kerb, and a 5-cm kerb. The sequence ing as subjects in Experiments I to and placement of the obstacles was III. randomly selected on each trial from twelve predetermined plans. These plans were designed to avoid very Method difficult or ambiguous situations, such as where the subject might con- In preliminary training subjects fuse an obstacle with a wall behind were required to walk around the perim- it, or where a difficult obstacle eter of the experimental room using might be confused with an easier ob- the aid, and attempting to maintain a ject beyond it. In half of the tri- distance of approximately 0.5m between als the start was as marked in Fig. 5; themselves and the wall without any in the other half the subject went collisions. The purpose of this was in the opposite direction around the to familiarize the subjects with the room and started at the point marked

Metres

1 3 cm Kerb

7 cm Pipe

20cm Kerb 23 cm Pipe ^_^ 15 cm Boord o -• 1

30 cm Board

B - Bookcose

T - Toble 5cm Kerb

Figure 5. A Typical Obstacle Course Layout.

79 ,

"finish." Once the course had been set out, the subject was led into the room. He was taken to the start of the course, where he received his instruction. He was required to walk around the room, maintaining a constant distance from the wall. When he detected an obstacle in front of him he was required to stop and report the detection to the experimenter. The experimenter walked quietly behind the subject, and satisfied himself that the subject had indeed detected an obstacle. He then measured the distance between the sub-

- ject and the obstacle, it was removed, 100 - D -B L rxiqT ,„ B .nd W,n (JlolH ^1 rfhlCfl

and the subject was asked to continue. - 11 In general the time taken between the as subject stopping and being requested m il . to continue was between 10 and 20 sec- ii If - onds. Subjects were allowed to col- 1 lide with the obstacles if they failed 50 to detect them. On collision, the ob- -^ stacle was caught and removed, and ' the subject requested to continue. ^ 1 • 1 i The time taken to complete the course F 1 ^ - was recorded. Trials with the aid i i_i uL 1 1 were interspersed with trials without L., M ^ s £: it. The long-time and newly-blind ^ Q. adult groups both performed six trials with, and six without, the aid; the blindfolded, four each, and the ' blind children; 10 with and six without the aid. Subjects were given instruction in echolocation with the Figure 6. Proportion of Obstacles aid, but not without. Subjects adopt- Detected per Subject for Each ed various techniques for getting Group. Shaded Bars Indicate around unaided (e.g., shuffling the Detections Made with Aid. feet, sharply bringing down the heel of the shoe onto the floor, jingling coins in the pocket, clapping hcinds experiment. Although they were less and clicking fingers and tongue). familiar with their experimental course which was also a little more complex, this is not directly rele- Results vant to the task of detecting obstacles within it. It appeared, however, The proportion of obstacles de- that some of the middle-aged blind tected per siibject for each group is subjects lost motivation in the rath- shown in Fig. 6. The shaded bars er long series of experiments, and show detections with the aid, and the were not as careful as they had been insert bars those without. The blind- earlier. In support of this, some folded subjects performed surprisingly of them had indicated that they well in this task, both with and with- thought it unlikely that an elec- out the aid, while the long-time, and tronic aid would ever replace their newly-blind adults did rather poorly, present mobility aid (guide-dog in especially unaided. The blindfolded one case, and white stick in the subjects were, of course, particular- others) and once this indeed proved ly highly motivated and not a matched to be the case, motivation would control group. The blind children presumably drop. performed at about the same level as the blindfolded group. It is not al- In almost every case it is evi- together clear why the two groups of dent that the subjects were able to blind adults did rather poorly, by detect obstacles better with the de- comparison, in the obstacle course vice than without. The order of

80 likelihood of detection for the dif- technique. The distance at which the ferent obstacles was similar to that detection was made was measured only observed in Experiment I. The blind- for the blind children and the blindfolded group had learned to recognize folded adults. The mean distance of the galvanized iron pipe, and, rather detection is shown in Fig. 7, both curiously, detected the 15-cm board with and without the aid. It is ap- more frequently than the 30-cm board. parent that detections were made at The adult blind subjects appeared to a far greater distance with the aid have some difficulty in detecting the than without. This increased dis- 7-cm pipe. This might be a result of tance of detection with the aid would presbyacusis , as there was some evi- allow the blind person time to take dence of high-frequency hearing loss avoiding action while walking at a in their audiograms. On the other higher speed. Further weight is add- hand, these subjects seemed to show ed to the suggestion noted in Experi- an enhanced aibility to detect the ment I, that there was an optimum kerbs , especially in comparison to distance for the detection of the the children. This might be associ- kerbs when using the aid. It seems ated with the greater difficulty ex- that if the subject had not detected perienced by some of the children in the kerb at a distance of 0.5 m, then maintaining the prescribed scanning there was little likelihood of him detecting it at all.

Wind Children

C Wilhoul A.d

2 -

o a. a. ^ ^ X IS ID S b E E o O u E u F E u in u to P^ lo m O m O in (M

Figure 7. Mean Distance of Detection for Blind Children and Blindfolded Adults With and ''^.chout Aid. Bars Indicate Standard Error About the Mean.

81 .

With Aid Without Aid

lOCl A &l>od ChJd-en

""*^\ .- .V ,• 50 / V^•^/ — O"""^' /:>

100 •B Long I " Bl.nd Blindfolded S.gHFd

50 /K " /

' r- '-v ' -

2 4 6 10 2 4 6 8 10'

Trial number

Figure 8. Percentage of Obstacles Detected per Siobject as a Function of the Trial Number, With and Without Aid.

Figure 8 shows the percentage of ob- about the possibility of colliding stacles detected per subject as a with obstacles. They would often function of the trial nuirber. Almost stop and scan the environment ahead, all subjects showed an increase in before continuing for a few more percentage detection as the trials steps and were thus much less likely progressed for both aided and unaided to miss an obstacle. It is also note- performance, although the performance worthy that subjects took considerably of the long-term blind eventually de- longer to complete the course when us- clined (see above) . None of the s\±>- ing the aid than when without it. This jects had previously received any ex- is a similar result to that observed plicit training in obstacle detection by Beurle (1951b) for blind children and it is notable that even those navigating in a familiar outdoor en- blind for a long time could have their vironment. A slight contributory unaided performance improved. It is factor for the extra time taken to also notable that the blindfolded sub- complete the course with the aid was jects made very few unaided detections that more detections were made, and so in Experiment I , but by the end of more time was taken in recording and Experiment IV were performing well measuring the results. Also, the unaided. The same trend is apparent blindfolded subjects, when deprived but less marked for the newly blind of the aid, were less ready to stop group and "scan" the environment than when using the aid. The comparatively high level of performance of the blindfolded sub- TABLE 3 must be viewed in relation to jects Mean Time to Complete Course (Min.) the mean times to complete the course, 9 and Table 3). The blindfolded (Fig. Blind with aid 3.5 considerably longer to subjects took children without aid 2.2 complete the course than their blind counterparts, because they were far Blindfolded with aid 12.0 more cautious and much more conceined sighted without aid 6.0

82 DISCUSSION (1968) suggest, most blind people prefer to have both hands free. This The results of Experiment I in- is not a serious objection, as the dicate that subjects find the aid to device could easily be mounted else- be of considerable assistance in de- where, such as on the chest, although tecting obstacles. The ability to this might require automatic scanning. detect obstacles with it appears to Head mounting would facilitate scan- be independent of any previous experi- ning but would be objectionable from ence of echolocation. This result by the cosmetic aspect, as the source itself suggests that the device might size required could hardly be accom- be useful as a mobility aid, especial- modated within spectacle frames. Al- ly for those blind people who have not ternatively, a sound source might well become proficient in echolocation be mounted in the handle of a long

(those who are newly blind) . However, cane, or even at its tip. as Beurle {19 51b) and Gray stone and McLennan (196 8) have suggested, there Experience gained in the study are several factors which militate reported above has indicated possible against the use of such a device as improvements to the device for more a general mobility aid. Blind peo- general use. The aid could be made ple, like the rest of the population, considerably smaller, with its own dislike attention being drawn to them power supply incorporated. It could particularly if it reflects a dis- then be stored in pocket or handbag ability, and might well reject the until required for use. Although a carrying of an audible sound source. form of electronic, or mechanical The device becomes far less effective scanning might be advantageous, there when there is a high level of ambient is merit in the user taking an active noise, thus reducing the advantage of role in data acquisition. It would the device in situations where it is be worthwhile to include an automatic most needed (busy streets , noisy volume control to adjust to environ- shops, offices, or municipal build- mental noise level.

ings) . Also as Graystone and McLennan

W.th ad

Wiihout aid

-ei.nd folded O C Sighted

*- C

, -^._,.^'-

Triat number

Figure 9. Time to Complete Obstacle Course With and Without Aid.

83 , , ,.

The device provided a greater In conclusion, we would like to range of perception of the world which reiterate the little heeded words of would tend to increase confidence in Beurle (19 51a) , written over 20 years the less mobile blind. The experi- ago, with regard to echolocation ments suggested that the aid would be "... the desirability of deliber- of considerable assistance in detect- ately training blind people in the ing a wide variety of street "furni- art of ' getting about ' cannot be too ture," although probably not a down- highly stressed." step or a parked bicycle. The implications of Experiment II, which showed that obstacles became "invisi- ble" if tilted, might not be as serious in a real-life situation where reflections are also obtained by means of ambient sound and multiple reflections, and the user is not stationary within the environment. The implications of Experiment III, show- ACKNOWLEDGEMENTS ing poor absolute distance judgment, indicated that acoustic images tend to predominate over reflection, tones but this may not be the case when the The work de scribed in this paper subject is in motion and dynamic clues was supported by a grant from the are available (Wilson, 1966). The Wo Ifson Foundati on. We would also surprising and encouraging result of like to thank J. B . Ruscoe for tech- Experiment IV, that children who had nical assistance and Drs. D. P. An- been blind for many years more than drews and W. A. Ainsworth for help- doi±iled their original unaided detec- ful comments on the manuscript, and tions on the experimental course sug- the personnel an d many willing sub- gested that explicit training in jects from the B lind Workshops echolocation using an aid such as Stoke-on-Trent and Lickey Grange that described here might be of use School for the B lind, Bromsgrove. to many blind people.

REFERENCES

Ammons , C. H., Worchel, P., & Dallen- Beurle, R. L. Pulsed stereophonic bach, K. M. Facial vision: The ultrasonic aids. In Proceedings perception of obstacles out of of the Leonard Conference on Re- doors by blindfolded and blind- search into Visual Handicap folded deafened subjects. Ameri- London: Southern Regional Asso-

can Journal of Psy ahology , 1953, ciation for the Blind, 1972, 19-25. 66, 519-53. Bilsen, F. A. On the interaction of

Bassett, I. G. , & Eastmond, E. J. a sound with its repetition. Doc- Echo location: Measurement of pitch toral Dissertation, University of versus distance for sound reflected Delft, 1968. from a flat surface. Journal of the Aaoustiaal Society of America Cotzin, M. , & Dallenbach, K. M. Fa- 1964, 36, 911-16. cial vision: The role of pitch and loudness in the perception of ob- Beurle, R. L. Electronic aids for stacles by the blind. American blind people. British Journal of Journal of Psychology , 1950, 63,

Psychology , 1951(a), 42, 164-71. 485-515.

Beurle, R. L. Electronic guiding Fourcin, A. J. The pitch of noise aids for blind people. Electronic with periodic spectral peaks. In

Engineering , January 1951(b) , 2-7. Rapport 5^ Congres International d' Aaoustique ^ Liege: la, B42, 1965,

84 , , ,

M. B. Gardner, Historical background American Journal of Psychology , of the Haas and/or precedence ef- 1944, 57, 133-83. fect. Journal of the Acoustical

Society of America, 1968, 43, Thurlow, W. R. , & Small, A. M. 1243-48. Pitch perception for certain periodic auditory stimuli. Grays tone, P., & McLennan, H. Eval- Journal of the Acoustical Society uation of an audible mobility aid of America, 1955, 27, 132-37. for the blind. Research Bulletin 17, New York: American Foundation Twersky, V. On the physical basis for the Blind, 1968, 17, 173-80. of the perception of obstacles by the blind. American Journal Griffin, D. R. Echoes of Bats and of Psychology , 1951, 64, 409-16. Men. London: Heinemann, 19 60. Twersky, V. Auxiliary mechanical Kay, L. Active energy radiating sound sources for obstacle per- systems: Ultrasonic guidance for ception by audition. Journal of the blind. In Proceedings of the the Acoustical Society of America International Congress on Tech- 1953, 25, 156-57.

nology and Blindness . New York: American Foundation for the Blind, Wilson, J. P. Obstacle detection us- Vol. I, 1963, 137-56. ing ambient or self-generated noise. Nature, 1966, 211, 218. Kay, L. Spectacles for the blind. In Sensory Devices for the Blind. Wilson, J. P. Psychoacoustics of London: St. Dunstan's, 1966, 275- obstacle detection using ambient 92. or self-generated noise. In

R. G. Busnel (Ed.) , Animal Sonar

Kellogg, W. N. Sonar system of the Systems . Jouv-en-Josas , 1967, blind. Science, 1962, 137, 399- 89-114. 404. Wilson, J. P. High-quality electro- Kohler, I. Der "Fernsinn" des Blind- static headphones. Wireless en. Die Umschau, 1952, 15, 449-51. World, 1968, 74, 440-45.

Kohler, I. Facial vision rehabili- Wilson, J. P. Psychoacoustical and tated. In R. G. Busnel (Ed.), neurophysiological aspects of

Animal Sonar Sy stems . Jouy-en- auditory pattern recognition. In Josas, 1967, 187-96. F. O. Schmitt and F. G. Worden

(Eds.) The Neuros ciences : Third

Rice, C. E. , Feinstein, S. H., & Study Volume. Cambridge, Mass.: Schusterman, R. J. Echo-detection MIT Press, 1974, 147-53. ability of the blind: Size and distance factors. Journal of Experimental Psychology, 1965, Worchel, P. & Dallenbach, K. M. 70, 246-51. Facial vision: Perception of obstacles by the deaf-blind. Rice, C. E. Human echo perception. American Journal of Psy chology Science, 1967, 155, 656-64. 1947, 60, 502-53.

Stevens, S. S. The relation of pitch to intensity. Journal of the Acoustical Society of America 1935, 6, 150-54.

Supa, M. , Cotzin, M. & Dallenbach, K. M. Facial vision: The perception of obstacles by the blind.

, "

RECOMBINATION AND VISUAL PERCEPTION

Robert B. Welch* Charles Hallenbeck** Editors

PREFACE More generally, it is assumed that by disturbing normally accurate The Conference on the Recombina- perceptual capacities and forcing tion Procedure as a Tool for the their reconstitution (adaptation) it Study of Visual Perception was held is possible to see more clearly how at the University of Kansas, Lawrence, these capacities operate under normal Kansas, on November 16-17, 1973. Our circumstances. A corollary to this theme for the present conference was, assumption is the belief that the as the title indicates, an examination process of adaptation in adult human of the recombination procedure as a beings provides a close analog to the technique for the better understanding original development of normal visual of visual perception. Because our perception in the human neonate. term for this procedure is little Kohler said (1964, p. 23), ". . . fac- known outside a relatively small tors which help a system overcome a

group of investigators , some clari- disturbance are no different from fication is in order. It was appar- those which play a role in normal ently Weiss (19 41) who first used the development.

term re aomhination procedure . He meant by it a technique whereby the While the connection between re- normal relationship between percep- combination research and blindness tion and action is disrupted. A good may not be an obvious one, we regard example is the rearrangement of the it to be fundamentally important. The visual field by means of optical de- fullest understanding of visual loss vices (wedge prisms) placed before the can only be achieved after the simpler eyes. Of interest to investigators problem of normal vision is fully un- who make use of this technique is derstood. To assert that visual sci- 1.) the way in which the organism re- ence has no points of contact with sponds initially to the recombination, blindness research is to suggest that 2.) whether or not the organism adapts a separate and unique science applies to it (and if so, to what degree) to a class of individuals. 3.) the nature of the adaptive end state, assuming that adaptation oc- Perhaps an example will help to curs, and 4.) the effect of certain clarify the importance of a sound variables upon the rate and/or final visual science as a kind of context level of adaptation that may result. for current work on blindness problems. The discovery that phosphenes can be produced by electrical stimulation of the occipital lobes has led Department of Psychology, University some enthusiasts to adopt a cinematia of California, Riverside, California. interpretation of the visual cortex. They contend that an artificial sens- **Department of Psychology, University ing and preprocessing system could of Kansas, Lawrence, Kansas. apply the correct combinations of " s , ,

locus, intensity, and frequency of examination of the present status of stimulus to support subjective visual research on the recombination pro- experience in individuals who have cedure . lost ordinary vision. The attainment of this goal clearly requires that we A second reason for calling the know the precise nature of the media- conference was that it had been eight tions that support the simple fact of years since the last conference of normal sight. Which mediators are this sort was held (Freedman, 1968). retinal? Which are motor? How much A great deal of relevant research has of the process of normal vision takes taken place since the earlier confer- place by means of functions with no ence at Tufts University, and we hoped clear locus in the occipital cortex? that our meeting would serve as a If "edge detection" is enhanced by means of assessing what the recombina- the action of peripheral ganglea, tion procedure has taught us to date that fact is essential in the quest about perception and, as well, the to evoke a subjectively similar ex- present-day evaluation of its useful- perience by electrical events farther ness as a tool for investigating per- "up-stream. ception.

The effects of the recombination Finally, it was an opportune time of sensory and perceptual elements to look to the future. Had the tech- constitute a truly critical set of nique outlived its usefulness? Had phenomena for the long-range success certain assumptions or expectations if not the short-term goals, of seri- of the technique been unfulfilled? ous work on visual loss. The m.eans What was the future role of the pro- by which the adapting organism, re- cedure? These were some questions we stores a sense of ordinary to genu- hoped would be dealt with during the inely extraordinary experiences, tech- conference. niques for the facilitation of the process, and discovery of barriers to Dr. Charles Kiesler, Chairman of its completion are the kinds of ques- the Department of Psychology at the tions which have important conse- University of Kansas gave the opening quences for work with the blind and ' address. Ten papers were presented visually impaired. A student in a in sequence over the next few days course in computer methods in psy- each followed by discussion. The chology once asked what pattern recog- closing address was given by the guest nition had to do with blindness. The of honor, Ivo Kohler. answer to these questions is the sam.e; one cannot possibly claim to under- Discussants included Drs. Mal- stand the one without an intimate colm Cohen, Thad Cowan, Hubert F. knowledge of the other. Dolezal, Wayne Hershberger, Lawrence E. Melamed, Gordon Redding, William B. There were several reasons for Templeton, James Walker, Benjamin Wal- our decision to hold a conference on lace, George Woods, and Fred Yaffe. the recombination procedure. First and most important, was the fact that There follows summaries of the Ivo Kohler of the University of Inns- ten papers presented at the confer- bruck, was visiting the University of ence. Because they are summaries, Kansas during the Fall 19 73 semester there was liberal condensation, and as the University Rose Morgan Profes- many details were omitted. We take sor. This is an endowed professorship, full responsibility for any inadver- awarded to one foreign scholar each tent misrepresentation or oversimpli- semester. The research of Dr. Kohler fication. The papers varied in con- (together with Erismann) on adaptation tent. Some dealt with a few studies to optically rearranged vision is con- and others with many. Research varied sidered by most investigators to be from previously published studies to the primary stimulus for the veritable unpi±iiished or preliminary data. In explosion of research on this topic several cases a significant propor- which was generated among American tion of the paper entailed a discus-

psychologists during the late 50 ' sion of proposed research. and the 60 's. We sought to capitalize on Dr. Kohler 's presence on our cam- Each summary is followed by a pus by inviting a number of investi- brief presentation of discussion gators to a conference devoted to an elicited by the talk. Because of

88 .

the nature of tape recordings of ACKNOWLEDGEMENTS discussions it was not always clear The editors who was talking, therefore, the extend thanks to Renate Welch, Diana Killey, and William names of the discussants have been Chest- nut for their omitted. invaluable help. Thanks for financial aid is due the University of Kansas and most especially to the American Foundation for the Blind.

An Introduction to the Typical Paradigm of a Study Using the Recombination Procedure

Because the reader may be unfa- all feedback (perceptual or otherwise) miliar with the procedural details of is precluded, as was true in the pre- studies involving the recombination exposure phase. procedure it is probably useful to describe a typical paradigm, the one Typically, the dependent variable used (with minor modifications) in is the difference in perceptual and/or most of the research presented at the perceptual-motor performance between conference the pre-exposure phase and the postexposure phase. If the shift is sta- Typically, testing begins with tistically significant and expecially measures of S's normal perceptual and/ if it serves to compensate for the or perceptual-motor capacities such effects of the rearrangement, it is as target-pointing accuracy, prior to assumed to have been caused by the re- imposing a rearrangement. This, then, arrangement and is termed adaptation. represents a baseline measure and is Naturally, it is necessary to include usually taken in such a manner that S a control group (for example, one ex- receives no feedback about his per- posed to a normal perceptual environ- formance. In some instances this ment) for the purpose of comparison. "before" measure is taken just after If £"s perceptual field is not rear- the perceptual rearrangement has been ranged at the time that the pre- and imposed, but again no feedback is al- postexposure m.easures are taken, a lowed. After the "pre-exposure phase" "pre-post" shift is often referred to is completed, S is exposed to the con- as the "negative aftereffect." If, sequences of his actions in the per- on the other hand, measures are taken ceptually rearranged environment. This at the beginning of the period of re- represents the "exposure," or "adapta- arrangement and again at some later tion," phase, its length varying in time while the rearrangement remains, different studies from a few minutes an observed adaptive shift may be to many days. Depending on the study, referred to as the "reduction of ef- the postexposure measures may be taken fect. " It is not clear that these at intervals during the exposure peri- two kinds of shifts are directly com- od, at the end of the exposure period, parable . and/or at various intervals after exposure is terminated. In any case.

89 , .

Prism Adaptation: Learning Aspects and Proprioceptive Contributions

Edward Taub and Israel A. Goldberg

Institute for Behavioral Researah, Inc.

In this paper Taub discussed a In order to see if the preceding series of studies which he and his conclusion also held for immature or- associates have carried out over the ganisms Taub et al. subjected four last decade. The studies have in moneky infants to bilateral forelimb common the aim of assessing, by means deaf ferentation within hours after of various types of recombination birth. These animals spontaneously procedure, the role of sensory feed- developed the behaviors of ambiula- back upon the learning and perform- tion, climbing, and reaching for ob- ance of motor skills. Three comple- jects, all of which were as good at mentary research strategies were three months of age as in monkeys de- used: 1.) somatosensory-deaf ferenta- afferented in adolescence. Further- tion, 2.) adaptation-to-sensory-re- more, with special training it was arrangement and 3.) combined deaf- possible for grasp, individual finger ferentation and sensory rearrangement. use, and precise hand-eye coordina- Both human beings and monkeys served tion to be demonstrated. On the day as 5s in this research. of birth, two additional rhesus infants were given both bilateral forelimb deafferentation and occlusion The Somatosensory - of vision (sewing the eyelids closed) Deaf ferentation Experiment The elimination of vision had only a small effect (1-2 week retardation) Taub and his colleagues surgic- on the development of general motor ally eliminated sensation from vari- coordination of the forelimbs such ous portions of the monkey's body as supporting the body weight, walk- and then, after recovery, assessed ing, or clasping objects, as compared the amount of movement and motor to the nonblinded deafferented ani- learning demonstrated by the animal mals. On the other hand, the blinded

in various situations. With adoles- deaf ferents , not unexpectedly, recent monkeys he (and others) found vealed an inability to engage in dis- that if one forelimb is deafferented crete reaching responses (hand-to-

(by means of dorsal root section) , it mouth) . However, special shaping is effectively useless unless there procedures led very rapidly to ac- is sufficient motivation and/or the curate performance. intact limb is restrained. Interest- ingly, Taub et al. showed that if Finally, Taiib and his associates both forelimbs are deafferented (by have attempted to perform limb de- means of bilateral dorsal rhizotomy) afferentation on the unborn fetus. the animal demonstrated spontaneously The dorsal rhizotomies were performed quite adequate (although not complete- at the end of the second trimester of ly normal) usage of these limbs for prenatal life. The results of this grasp, ambulation, and climbing, both study were eaui vocal, however, due to while able to see and while blind- the fact that the operation caused folded. These capabilities remained the spinal cord to become flattened intact even following complete de- and otherwise deformed, thereby re- afferentation of the spinal cord. sulting in what was essentially a "spinal preparation." Nevertheless,

Thus , in contradiction to some these animals developed the ability well-known theories concerning the to use the forelimbs for pushing nature of movement and learning, it themselves into a sitting position, seems clear that somatosensory feed- and for crude ambulation. Further back is not necessary in order for research will include techniques adolescent monkeys to learn and to designed to prevent damage to the perform movement. spinal cord.

90 Adaptation to Sensory event. In this experiment S was re- Rearrangement Studies quired to move the right arm from side to side. As he viewed the limb In this second type of study in- through a prism whose strength could tact human subjects were confronted be varied. In one condition the de- with an optically rearranged visual gree of optical displacement was re- environment and their response to the duced if S increased the rate of arm situation noted. More specifically, movement; in the other, condition the Taub and his colleagues exposed 5s reduction in sensory discordance was in rate of to a visual field which, by means of contingent upon a decrease If there was no change in wedge prisms , was displaced laterally movement. by approximately 11°. The ability to rate or a change opposite to the de- adapt to the visual rearrangement and sired one, the result was an inoreaae was found the nature of this adaptation were in optical displacement. It of arm measured in a variety of conditions. that (unbeknownst to S) rate movement would either increase or de- Taub's laboratory had produced crease when such a change led to a strength, thus a number of studies designed to test reduction in the prism the view that adaptation to optical providing support for the hypothesis rearrangement is an example of learn- that such a reduction represents a , which ing and not (as some have argued) a rewarding state of affairs special or "more primitive" process. strengthens a response tendency. The research by Held and Freedman of evidence (196 3) suggested that prism adapta- Another category the con- tion is not learning, because it was which can be used to support an found to occur even when S received tention that prism adaptation is no error-corrective feedback. Ss in exam.ple of learning is a dem.onstra- many of Held's studies were allowed tion'that it is affected by certain to observe through the prism their "learning variables" in the same man- actively moving limbs, but no target ner as are "traditional" learned be- was provided. Because they were pre- haviors. One such variable is the cluded from making errors (punish- distribution of practice. It is well ment) or correcting for them (reward) known that "massed practice" leads, there appeared to be neither a motive at least initially, to poorer per- trials nor a reward contingency for the prism formance than is the case if adaptation that nevertheless resulted. are distributed over a longer time interval. Furthermore, at least for Tavib , however, suggested that the motive might be the avers ive character certain motor tasks, intermanual of sensory discordance. That is, the transfer of a skill is retarded or presence of sensory rearrangement is eliminated by the massing of trials. aversive to an organism because in the past this state has been associ- Taub reported a study in which ated with punishing consequences prism exposure trials were either (misreaching, running into obstacles). massed or distributed. The results Avoidance or escape from the presence indicated that distributed practice evenly of sensory discordance is, therefore, (ten 30-second exposure periods secondarily rewarding. Since adaption distributed over 9.5 minutes) led to serves to reduce or eliminate the dis- greater adaptation than did massed cordance, it represents a response practice (ten 30-second exposure rest that leads to a reward and therefore periods with no intervening predicted by the fits neatly into the traditional learn- periods) . Also, as paradigm. Crucial to this argu- learning theory hypothesis, inter- ing adaptation ment is the fact that adaptation will manual transfer of prism prac- be induced merely by 5's perception occurred only with distributed that a sensory discordance exists; tice. it is not necessary that S actually experience an error in reaching or In another experiment, Taub et well-known in navigating obstacles. al. demonstrated that the absence of prism adaptation when 5's One study reported by Taub which limb is moved passively (Held and represents a direct test of his learn- Hein, 195 8) does not occur if the ing theory of prism adaptation showed passive movement is coupled with dis- that the reduction of sensory distributed exposure periods. Neverthe- cordance is, indeed, a reinforcing less, active movement (whether massed

91 or distributed) induced more adapta- in the normal animal inhibits or de- tion than did passive movement, thus lays adaptation. That is, the less indicating that active movement en- there is of position sense the easier hances, but is not (as Held and Hein it is to alter. have maintained) necessary for adaptation. The evidence from this study clearly supports the view that, at least in the prism-wearing situation, Combined Deaf ferentation adaptation is primarily or totally a and Sensory Rearrangement Studies matter of recalibrating the body position sense (which could include

In this third type of experiment felt eye position) . A vision change Ss (monkeys) were both deafferented hypothesis would predict less, not

and exposed to a prismatically dis- more , adaptation after deaf ferenta- placed visual field. The specific tion, since this results in making aim of the study which Taub described felt body position less salient and, was to examine the hypothesis of Har- therefore, presumably less effective ris (196 5) and others that the prism- in modifying visual localization. adapted state is one of altered proprioception (or, more appropriately, position sense), rather than of- a change in visual perception. If this hypothesis is correct, then the surgical elimination of proprioceptive afference should significantly affect the rate and/or final level of prism DISCUSSION adaptation. During the discussion period The monkeys had been deaffer- several issues were raised with re- ented in the neck, shoulders, and spect to the deaf ferentation studies arms for a year prior to being trained (with and without visual rearrange-

to reach for a food pellet while seat- ment) . Of particular interest was ed in a restraining chair. S was pre- the question of how it was that a cluded from viewing the limb through- deafferented monkey could learn to out each reaching response and at all reach for unseen objects (as when other times as well. All of the de- also blinded) or to adapt to prism afferented animals quickly learned displacement. Taub argued that in to make this response with surprising both instances it is clear that a (although not normal) accuracy. After central (efferent) form of position an accurate and stable performance sense must be involved, in that pro- level was reached, each of the deaf- prioceptive afference from the body ferented 5s, as well as each member parts in question had been eliminated

of a group of normal animals , was by means of the operation. If a de- fitted with prism goggles for about afferented and blinded monkey can 24 hours. They were not allowed to learn to reach accurately for its see any part of the body, but the mouth it must be because the efferent head could move freely. As measured commands for the correct response are by aftereffect, adaptation was 39 per- registered centrally and reinforced cent of full prism displacement for (by ingestion of the food object in

the normals but 100 percent for the the monkey's hand) . If the deaffer- deafferents. ented monkey demonstrates adaptation, it cannot be due to a change in pro- Dr. Taub argued that there are prioception but rather in the central two forms of felt body position: form of position sense. Because only 1.) proprioception or somatosensory one of the normally two types of po- feedback, and 2.) central feedback sition sense is present, prism adapta- or central efferent monitoring. The tion should be enhanced— and, in fact, first is afferent, the second effer- it was 100 percent. ent. Clearly, the surgical operation served to eliminate only the afferent Discussants expressed uncertain- type of body position sense. Because ty regarding the precise body part this procedure led to facilitation of that was involved in the adaptation prism adaptation Taub et al. argue and the nature of the cues of sensory that the presence of proprioception discordance that induced it.

92 . .

Because the head was free to move and parameters as exposure trials and was not deafferented (while the neck intertrial intervals and has gener- forelimbs, and upper extremities of ated a number of interesting experi- the trunk were) Taub suggested that ments. Questions were raised, how- adaptation may have involved a re- ever, with respect to its usefulness calibration of (central) position as an explanatory concept. In par- sense in terms of the relationship ticular, several of the discussants between the head and the remainder of were unclear as to why distributed the body. Possibly, the neck muscles practice should facilitate ipsilateral represented the specific locus of adaptation and lead to intermanual adaptation. Such a change, if com- transfer. Merely noting the fact pletely adaptive, would lead the that the results are the same or animal to reach accurately for ob- similar to those found with "tradi- jects (with either limb) . Because tional" tasks, such as the pursuit the monkey was precluded from a rotor, does not explain the finding. view of any part of its body, the It was suggested by one discussant sensory cues for adaptation must that a useful explanation might en- have resulted from the various vi- tail the notion that different kinds sual effects that occur during head of exposure condition, such as massed movement. vs. distributed practice, lead to differential selective attention to There was general agreement visual and proprioceptive cues, which magnitudes that Taub ' s learning model of prism in turn lead to different adaptation has been useful in that and/or types of adaptation. Uhlarik it has forced investigators to ex- and Canon (1971) have proposed such amine and operationally define such a mode 1

Processes Involved in Behavior Observed in Recombination Studies

George Singer, Meredith Wallace, and Margaret Austin

Latrobe University

George Singer's address consist- reinforcement. The amount of BC which ed of a detailed presentation of a will occur in a given situation de- number of studies carried out in his pends upon 1.) the extent of the trans- laboratory, supporting a three-com- formation (if the rearrangement is ponent model of adaptation to optic- relatively minor, and therefore un- ally-induced tilt. Singer began by detected, no change in motor response stating that the end product of adap- "rules" will be elicited), 2.) the na- tation to rearranged vision is not a ture of the task (e.g., the presence unitary entity, but rather the sum of error-corrective feedback facili- (or product?) of several qualitative- tates motor learning), and 3.) the ly different components: behavioral number of trials (the more trials,

compensation (BC) , sensory spatial the better learned the response)

adaptation (SSA) , and visual shift

(VS) . BC is the acquisition of an SSA, unlike BC , can occur as a adaptive motor response and is, there- result of prolonged, unchanging stim- fore, an example of motor learning. ulation of a single sensory system, It should reveal the characteristics dissipates relatively rapidly, and of learning. That is, it should: does not depend on dis aordant informa- 1.) require knowledge of results, tion. The "figural after-effect" 2.) increase with practice, 3.) trans- (Kohler and Wallach, 1944) is an ex- fer to other situations, and 4.) per- ample of this form of adaptation. sist for some time in the absence of Another, more relevant, example is

93 . .

the reduction in perceived curvature a 20° tilted (up or down) bar, plus of objectively straight lines when visual stimulation from the same bar passively viewed through a prism for made to appear horizontal by means of a period of time. This sensory adap- the prisms. This condition was ex- tation leads to a negative aftereffect pected to induce both SSA and BC in upon removal of the prism--the lines the same direction; they should sum- appear curved in the opposite direc- mate. Group 6 was exposed to kines- tion. These aftereffects are short- thetic stimulation from a 10° tilted lived (2-3 minutes) . The same ef- bar plus visual stimulation from the fects occur with the inspection of same bar made to appear tilted 20° objectively curved lines, which dem- from the horizontal in the same di- onstrates that sensory rearrangement rection as the 10° physical tilt, is irrelevant for this form of adap- giving a net sensory discordance of tation, and that it is probably con- 10°. It was predicted that this group fined to the visual system. Analo- ' would reveal SSA and BC, but in oppo-

gous results have been reported in ; site directions (thus partially or the kinesthetic modality (Heinemann, totally cancelling each other) 1961) . Finally, Group 7 received kinesthetic stimulation from the bar tilted phys- VS , the third component of adap- ically by 10° from the horizontal tation, is an apparent reduction of (either up or down) plus visual stimu- the visual rearrangement (and/or a lation from the same bar optically visual aftereffect upon removal of tilted 10° from the horizontal in the optically transforming device) the direction opposite the physical It is a purely visual change with tilt (a net discordance of 20°). In a longer duration than is the case this condition the prediction was with SSA. VS may occur in the ab- that SSA and BC would both develop sence of BC and vice versa. and be in the same direction.

In the first experiment reported As can be seen by Table 1, all in detail by Dr. Singer the rearrange- predictions were confirmed. These ment used was optical tilt, effected data, then, provide support for the by means of two Dove prisms in tan- existence of two of the three prism- dem. The S's task before, during, adaptive components postulated by Dr. and after exposure was to set (by Singer. As he mentioned in his talk, means of a control knob) a bar to the both SSA and BC probably occur (and apparent horizontal in terms of kines- summate) in m.any prism adaptation thetic information. The exposure studies, since both sensory discord- period involved stimulation of kines- ance and prolonged exposure to atypical thesis alone, vision alone, or the (tilted, curved, asymmetrical, etc.) simultaneous stimulation of both sen- stimulation are usually present. sory modalities. There were seven

conditions in the experiment. Group 1 The third component, VS , was ex- received kinesthetic stimulation from amined in another experiment. Dr. a horizontal bar but no vision, while Singer observed in his talk that the Group 2 received both kinesthetic and elicitation of VS appears to require, visual stimulation from the horizontal among other things, a relatively large bar. Neither of these two groups was visual field containing various ob- expected to reveal either SSA, since jects. In order to overcome the con- the bar was horizontal, or BC, due to striction of the visual field that the absence of sensory discordance. occurs when Dove prisms are used, a Group 3 received kinesthetic stimula- 6' X 6' X 4' room was built which tion from a 20° tilted bar but no vi- could be physically rotated about the sion and was therefore expected to horizontal axis. Thus an upright S reveal SSA but no BC. Group 4 experi- who views this field is confronted enced kinesthetic stimulation from a with a discordance between visual and horizontal bar plus visual stimulation nonvisual cues since the visual frame from the same bar, optically tilted of reference gives information of by 20°. For this group the fact that tilt, while the vestibular, kines- the bar was horizontal led to the pre- thetic, and proprioceptive systems diction of no SSA, while the presence give information for upright body of the 20° sensory discordance meant position. Two factors were examined that BC should result. Group 5 re- in this experiment: trials (either ceived kinesthetic stimulation from one or ten) and exposure posttest

94 .

TABLE 1

Mean Aftereffect and Standard Deviations Obtained from 7 Different Prism Exposure Conditions

Conditions of Discordance Aftere ffect Exposure between vision Predictions Mean S.D. Group Vis. Kin. & kinesthesia for adaptation (deg)

1 absent 0° None No adaptation -0.65 2.63

2 0° 0° None No adaptation 0.24 2.07

3 absent 20° None SSA 2.95 2.33

4 20° 0° 20° BC 1.97 1.60

5 0° ±20° 20° SSA + BC 4.92 1.81

6 +20° + 10° 10° SSA - BC -0.64 2.58

7 tlO° il0° 20° SSA + BC 5.71 1.93 delay (adaptation measured immec?iately the effect for trials was statistic- after exposure, 5 minutes after ex- ally significant (more adaptation for posure, or 15 minutes after exposure). 10 trials than for 1). Finally, there Different Ss were used in each group. was a drop in level of adaptation between the 0-minute exposure-posttest The room contained various ob- interval and the 5-minute interval jects, a chair, a picture, and was but not between the 5- and 15-minute tilted laterally by 22°. On the intervals. The drop over the three back wall of the room was a 24" X intervals is more apparent for the 1.5" luminous bar, which could be ro- 1-trial groups than for the 10-trial tated (by means of a rotary knob) by groups either S or E . VJhen seated directly in front of the tilted room, S could Dr. Singer concluded that this see nothing else. During the pre- experiment supported the existence of and posttests S was required to set the VS component of prism adaptation, the luminous bar to apparent visual a component that is Qualitatively vertical (with respect to gravity) distinct from a visual sensory after- in the otherwise dark room. During effect of the Cibson variety since, the exposure phase the lights of the in contrast to the latter, it did not tilted room were switched on and S dissipate even after a 15-minute made either one or ten adjustments period. VS is also separable from (at 30-second intervals) of the bar behavioral compensation. Singer ar- to vertical. Between all adjustments, gued, because in the present experi- view of the room was precluded by ment there was no behavioral inter- drawing a curtain. Two of the groups action with the sensorily discordant were given the posttest immediately field. Another characteristic of the after the exposure trial(s), two adaptation found in the tilted room, were tested after a 5-minute interval situation is the fact (based on a and two after 15 minutes. For the pi±)lished study) that it leads to two delay conditions, S wore trans- complete intermanual transfer of lucent goggles between completing the motor responses, clearly supporting adjustments and undertaking the post- the notion that this component consists test. of a chanae in the visual system.

The results are seen in Table 2. All conditions produced adaptation and

95 . .

TABLE 2

Mean Aftereffect and Standard Deviations Following Exposure to a Tilted Room

No. of Delay before Mean Standard Group Adjustments Posttest a/.=f feet (deg) deviations

1 1 No delay 3.93* 2.00

2 10 No delay 4.14 2.34

3 -s • -1 5 min. 2.20 2.52

4 ' ^ 10 " 5 min. 3.37 2.77 ''- 5 1- 15 min. 2.54 2.39

6 10 15 min. 4.05 4.64

*A11 deviations were in the same direction as the room tilt.

DISCUSSION prism-displacement study, since visual asymmetry is typically controlled for Singer was asked if he felt that (and curvature of tilting of contours the different components of tilt is irrelevant for the types of prism-

adaptation ever interaated. In par- ' adaptive measure usually taken) ticular, would an increase in one Singer's response to this question was component lead to a balancing de- that indeed there are sources of rele- crease in another? He said that he vant sensory spatial adaptation in the has not yet specifically investigated prism-displacement study, in terms of

this possibility. Thus, as it stands kinesthesia . That is, when S inter- now, his is an additive model. acts with a prism-displaced field by, say, reaching for objects, he may well In response to another question, ' develop a chance in kinesthesis which he argued that it was quite unlikely will persist to influence (add to or that in the VS study (the second ex- subtract from) the adaptive shift periment described) any sensory adap- \ measured. tation to the visual tilt could have taken place along with the VS. The -' Finally, it was pointed out that reason for this belief is that the a change (from fatigue, muscle poten- actual exposure to the tilt (approxi- tiation, etc.) in felt head and/or mately 5 seconds for the 1-trial groups eye position could and apparently and 50 seconds for the 10-trial groups) does lead to a change in visual was too short to lead to much if any straight ahead, although it is not sensory spatial effects. clear that this is identical to the visual shift which Dr. Singer has An important question was raised measured. Relevant to this point. concerning the generalizability of Dr. Singer discriminated between a Singer's results and model from vi- muscular adaptation and adaptation sual tilt adaptation to the more typi- of the joint receptors (which may cal experiment in which the visual cancel each other) field is laterally displaced. More specifically, it was asked if he felt that SSA was really a factor in the

96 )

Some Effects of Exposure to Optical Disarrangement on Adaptation to Optical Rearrangement

John J. Uhlarik

Kansas State University

Two recent, unpublished studies The first experiment compared

on adaptation to prism-displaced vi- two conditions : pre-exposure to dis- sion are described in detail here. arrangement and pre-exposure to normal Rearrangement of the visual field viewing. Following the pre-exposure represents a stable and relatively period, 5s in both conditions were isomorphic change in which the normal exposed to 11° lateral prism-displace- correspondence between felt and seen ment and then measured on their level body position is disrupted, at least of adaptation in terms of 1.) vision, initially. Adaptation to such a 2.) felt limb position, and 3.) tar- situation is a possibility. Dis- get-pointing accuracy (the visual-

arrangement, on the other hand, proprioceptive measure) . During both represents a situation in which the the pre-exposure and exposure viewing nature and/or extent of optical dis- periods S reached for visual targets tortion is in a constant state of at a steady, fixed rate while continu- flux. That is, the relationship be- ously viewing the moving limb. Thus, tween felt and seen body is ambiguous both error-corrective feedback and or unspecified. If this change is experience of the relationship between random, it is impossible for adapta- felt and seen body position were pre- tion to occur. Held and Freedman sent. Short rest periods occurred (19 6 3) report a study which found every 2 minutes. The disarrangement that even if the changing state of pre-exposure viewing situation in- optical rearrangement is systematic, volved observing the targets (and the no adaptation will result. The aim target-pointing attempts of the arm) of the present studies was to examine through two counter-rotating Risley some implications of the "comparator" variable prisms, placed in tandem in or "coordinator" model of prism, front of the right eye. The net ef- adaptation. In brief, the role of fect was of a visual field continu- a comparator is to monitor, and hence ously changing in its lateral dis- to control, the level of adaptation placement from 17° to the left through in terms of the relationship between 0° (straight ahead) to 17° right and 1.) the stored correlations (between back again. The full cycle took ex- felt and seen body position) from actly 1 minute. In addition to the previous viewing situations and factor of type of pre-exposure view-

2 . ) the inputs from the present viewing condition were the variables of ing condition. In order to further exposure duration (10 min./20 min . specify some qualitative and quanti- during the pre-exposure phase and tative properties of this kind of prism orientation (base left/base mechanism the studies reported by right) during the rearrangement Uhlarik examined the effects of ex- period. posure to conditions of disarrangement on the level of adaptation to subse- Measures of vision, felt limb po- quent conditions of rearrangement. sition, and target-pointing accuracy Various models and assumptions lead were taken before and after the pre- to predictions that pre-exposure to exposure viewing period and immedi- disarrangement will facilitate, will ately after the rearrangement period. retard, or will have no effect upon During each of these three test peri- the level of subsequent adaptation to ods vision was normal and feedback rearrangement in comparison with a about performance was precluded. The condition of pre-exposure to normal measure of visual location entailed viewing conditions. Clearly, the out- the setting (by E) of a dot of light come of the present experiment has to S's perceived straightahead posi- important theoretical implications. tion in the otherwise dark room. The

97 , measure of felt limb position required were greater in the normal viewing that S place his unseen arm in such a condition than in the disarrangement manner that it was felt to be straight condition, especially when exposure ahead of his nose. The final measure was for 20 minutes. The two condi- required 5 to point at a series of tions did not produce differential visible targets, again with no visual visual adaptation. The apparent feedback from the pointing limb. The interactions between the type and order of the visual and propriocep- duration of the pre-exposure viewing tive measures during a given test periods were not statistically sig- period was counterbalanced across Ss nificant for either the propriocep- while the target-pointing measure al- tive or target-pointing measures. ways came last. Finally, it was found, contrary The data of primary interest to results reported by Held and Freed- were the differences between the man (1963) , that exposure to disar- measures taken before and after re- rangement did not lead to an increase arrangement. These scores represent in variability of target-pointing ac- the prism-adaptive aftereffects and curacy (or in the other two measures are presented in Table 1. While both either). Furthermore, no relation- proprioceptive and target-pointing ship was found between 1.) the amount shifts generally were significant, of change in variability between the only in one condition was there a tests taken before and after exposure measurable change in vision. The to disarrangement and 2.) the level general failure to find visual adap- of subsequent adaptation to rearrange- tation agrees with previous studies ' ment. which, like the present one, used a concurrent display condition. The In sum, the data of the first table shows that the adaptive pro- study suggested to Uhlarik that the prioceptive and target-pointing shifts comparator must act as an autocor- relator. Disarrangement serves to

TABLE 1

Mean Adaptive Shifts in Degrees Between Test 2 and Test 3 for the Concurrent Display with Target Exposure Conditions

Measures Visual- Visual Proprioceptive Proprioceptive

Preexposure Conditions 10 min. 1.4 2.1* 2,5^

Disarrangement 20 min. 2.2* 0.8 2.1*

10 min, 1.1 3.1* 4.2*

Normal Viewing 20 min. 1.1 3.9* 4.7^

Denotes values significantly different than zero, p < 0.05 (based on two-tailed t-tests)

98 introduce noise or error into the The results are seen in Table 2. system which interferes with the Both the visual and visual-proprio- autocorrelation process. ceptive measures were statistically different from zero in the disarrange- It is possible that the failure ment condition, while none of the to replicate the finding of increased measures was significant in the normal dispersion of localizations as a reviewing condition. However, in no sult of disarrangement was due to the case v/as there a significant differ- presence of error-correcting feedback ence between disarrangement and normal in the present experiment. Conse- viewing. Furthermore, no reliable in- quently, a second study was carried crease in variability of localization out in which S was exposed only to was noted subsequent to disarrange- the discrepancy between felt and seen ment for any of the measures. It is position of the limb. That is, no unclear why it was that even with no- targets were present, thus precluding target exposure there was a failure error-corrective feedback. S merely to replicate previous findings of moved the limb in and out, viewing it systematic increases in the varia- continuously during both pre-exposure bility of pointing localizations fol- and exposure viewing situation. This lowing exposure to disarrangement. is similar (although still not identical) to the procedure used in the To conclude, the data from the disarrangement study reported by Held first experiment indicate a retarding and Freedman (1963) . The rest of the effect of exposure to disarrangement design and procedure were the same as upon subsequent adaptation to rear- in the first experiment, except that rangement, while the second study all 5s received a 10-minute exposure foimd no such effect. This differ- period. ence in results suggests that there

TABLE 2

Mean Standard Deviation of the Visual, Proprioceptive, and Visual-Proprioceptive Measures for the Three Tests for the Concurrent Displav with Target Exposure Conditions. (Numbers in parentheses indicate correlation coefficients between the magnitude of the adaptive shifts [Test 3 - Test 2] and changes in variability due the pre-exposure periods [SD for Test 2 - SD for Test 1])

Measures Visual- Visual Proprioceptive Proprioceptive

Preexposure Conditions 10 min. Test 1 1.9 1.2 1.9 Test 2 2.1 (0.01) 1.2 (0.21) 1.4 (0.18) Test 3 1.6 1.0 1.5

Disarrangement 20 min. Test 1 2.0 1.6 1.8 Test 2 2.4 (0.08) 1.3 (0.20) 1.6 (-0.67) Test 3 1.8 1.4 1.6

10 min. Test 1 1.5 1.4 1.9 Test 2 1.5 (-0.50) 1.2 (0.48) 1.6 (0.28) Test 3 1.5 1.3 1.3 Normal Viewing 20 min. Test 1 2.0 1.6 1.6 Test 2 1.7 (-0.07) 1.3 (0.07) 1.2 (-0.26) Test 3 1.5 1.2 1.1

99 . might be important differences in the Because it is not clear that manner in which information regarding Uhlarik was successful in creating intermodality discordance and target- a "decorrelated" state in Ss exposed pointing error is processed by the to disarrangement, several alterna- comparator. tive procedures were proposed. One was to use a randomly (rather than systematically) changing prism strength while simultaneously making the target always appear to be straight ahead (by physically changing its position in the direction opposite the prism displacement) DISCUSSION This would make it impossible for S to use a conscious strategy of This paper elicited a very ac- pointing straight ahead. A second, tive discussion, both during and after related, technique is to allow S presentation. One of the major is- to "hit" the target on every trial, sues raised was Uhlarik's failure no matter where he points. This (even in the no-target situation) to situation could be effected by replicate the previously reported physically moving the target to (Held and Freedman, 1963) increase wherever S's finger had aimed. If in variability of localization due one or the other of these procedures to exposure to disarrangement. It induced increased variability in was suggested that this failure might target-pointing localization, one have vitiated the study. However, it could then repeat Uhlarik's design is more important to consider possi- with more confidence that the com- ble reasons for this failure to repli- parator had been affected in some cate. A number of suggestions were way. offered. The second experiment was designed to handle many of the obvious differences (target-pointing er- The fact remains, however, rors, use of strategies, possible that in the first experiment ex- failure to pay close attention to posure to disarrangement appeared the limb as it moves) between the to retard adaptation to rearrange- first experiment and that reported ment when compared to the situation by Held and Freedman (196 3) . Never- in which prior viewing was normal. theless, some potentially important Several potential explanations of differences remained in the second this result were put forth. Uhla- study. One of these is the fact that rik suggested that this may support Uhlarik used far fewer trials during the hypothesis that adaptation is the postexposure tests than was the a process of autocorrelation and case in the original disarrangement that disarrangement adds "noise" to study (12 target-pointing measures vs. the system, making autocorrelation 40). It is possible that the increase more difficult. Another possibility in localization variability induced that was raised is that the actual by disarrangement requires a longer discordcince experienced was greater period of postexposure time to be for 5s with prior normal viewing than manifested than was provided by Uhla- for 5s in the disarrangement condition rik's procedure. If this point is because the latter situation gave Ss valid it suggests that Uhlarik's experience with displacement (albeit,

studies were, in fact, successful in constantly changing) . Thus , there is their aim of inducing variability in a greater "contrast" between the ex- localization, but that their measures periences of normal viewing and rear- were unable to detect it. rangement than between disarrangement and rearrangement. This "contrast effect" might induce a more vigorous adaptation response.

100 The Subject: A Neglected Factor in Recombination Research

David H. Warren and Bruce B. Piatt

University of California, Riverside

David Warren began by pointing various sources of adaptation. Thus, . out that the contradictory results one 5, perhaps due to a particular often reported in the area of recom- perceptual history, may be quite sen- bination research (particularly prism sitive to the discrepancy between adaptation) may be due to individual seen and felt limb position, while differences in perception. In addi- another is attuned to target-pointing tion, he argued that by looking for errors. Both the amount and type of correlations between "perceptual adaptation achieved by these two hy- traits" and level and type of adapta- pothetical individuals might be quite tion to rearrangement one might come different. to a better understanding of the nature of the adaptive process. A dis- The experiment described by War- turbing possibility is that such hy- ren involved the evaluation of the pothesized perceptual characteristics adaptation of the right hand to later- interact with the independent vari- al displacement of the visual image ables of a given experiment. As an and of the transfer of that adapta- example of a relevant individual dif- tion to the left hand. The 5 sat at ference, when 5s are allowed to point an apparatus which was similar to the at a target with continuous view of type used in most previous studies of the moving limb, some individuals prism adaptation. During prism ex- (for whatever reasons) may fixate the posure he viewed the scene through a hand throughout its movement, while 10° base right-wedge prism and made others might spend most of the time pointina responses under a horizontal fixating the target. In effect, the shelf at various fixed targets. Visu- 5s would not be in the same condi- al feedback occurred only at the tion, since the former individual terminus of each pointing response. would receive truly continuous ex- A major addition to Warren's testing posure while the latter experienced apparatus was a "trolley" which ran something akin to "terminal expo- along a track placed beneath and sure. " parallel to the horizontal shelf. The track was attached to a boom Studies that have deliberately that rotated around an axle mounted examined individual differences in under the chinrest. 5's right index adaptation to perceptual rearrange- finger rested in this trolley and as ment are rarely found in the litera- he moved his pointing hand in and out ture. Subject differences may be at the various targets a potentiometer relevant for the "multi-component" system recorded continuously the ra- models of prism adaptation which dial direction of the boom while, at several investigators have proposed the same time, a series of micro- (Hay and Pick, 1966; Wilkinson, 1971; switches along the boom recorded the Singer, this conference; Welch, this proximal-distal position of the conference) . For example, it may be "trolley." Thus, it was possible to true that prism-induced aftereffects record continuously (on a Beckman of setting a bar to horizontal are dynograph) the details of 5's point- the sum of behavioral compensation ing responses throughout the experi- and sensory spatial adaptation (Sing- ment. Also used was a Beckman direct er) . Yet, some 5s may tend to acquire current electro-oculography system to a great deal of the first component monitor lateral eye position continu- and very little of the second; for ously. others just the reverse may be true. Similarly, it is possible that 5s The step-by-step procedure of the differ in their sensitivity to the experiment entailed eight temporally

101 ordered tasks. In Task 1 S pointed have occurred during Task 4. It con- with his right hand at each of six sisted of a maximum of 12 right hand visual targets, with vision displaced, adaptation trials, as in Task 3. but no feedback provided. Task 2 was Task 6 consisted of four trials point- identical to Task 1, except that the ing with the right hand, with dis- left hand was used. In Task 3, S placed vision, but no feedback. Com- pointed repeatedly with the right hand parison of Task 6 with Task 1 thus at a visual target which was placed provided a measure of right hand to the right of center but appeared adaptation. After Task 6 the prism to be straight ahead because of the was replaced with a clear glass aper- 10° base right—wedge prism which was. ture. Task 7 was identical to Task 3 now before S's right eye. Visual except that there was now no discrep- feedback was provided in this task ancy between the actual and apparent and, hence, adaptation to the prism locations of the target. A maximum displacement was expected to occur. of 20 trials was given, and the same If 5 made five consecutive pointing criterion of accuracy was used as in ' responses within 2° of the visual Task 3. Task 7, then, represented an target, he was then given three addi- "unlearning" or "readaptation" phase. tional trials and Task 3 was ended. Finally, Task 8 required four trials A maximum of 20 trials was allowed. pointing with the left hand to a Task 4 consisted of four trials point- visual target with no feedback. ing with the left hand, with vision still displaced but no error-correc- A graphic representation of the ' tive feedback provided. Thus a com- group results appears in Fig. 1. parison of Task 2 and Task 4 perform- The curve in the left-hand side of ance provided a measure of the amount the figure indicates the development of adaptation acquired by the non- of adaptation from Trial 1 to Trial prism exposed limb. Task 5 was in- 19. It is clear that S's initial cluded to regain any loss of adapta- error of pointing straight ahead (90°! tion of the right hand that might at a target actually located 10° to

* R ,95 / i

a 94 a ^ \ i

/ ' * R \ L

• 1.9° ' 77° L \ 92 I 3.5° UJ ir 9i

5 L tlj t ^ / CRIT CRIT 89 / 10.3 3.9

88 R

I I I I I I I I I I I t I t I I

4-5 8-9 10-11 18-19 I 2-3 4-5 6-7 8-9 10-11 I 2-3 6-7 12-13 14-15 16-17

Figure 1. Mean Pointing Response Over Trials for Adaptation to 10° Prism (Left Portion) and for Readaptation to Clear Glass (Right Portion)

102 " . the right (100°) was quickly overcome, 5s began by making a large error the average number of trials to cri- which on the next microanalyzed trial terion being 10.3 (SD = 6.2). Adap- was altered significantly in the di- tation as measured in terms of the rection of the actual target displace- difference between pre- and post- ment. By the criterion trials, the adaptation measures of no-feedback initial error was almost entirely target-pointing accuracy is indicated overcome by the two vertical arrows. It can be seen that adaptation occurred for both the exposed (right) and nonex- For the readaptation to zero posed (left) hand, but was signifi- displacement phase, five trials were cantly greater for the former. The microanalyzed, the first three tri- shift for the left hand represents als and the pair of trials that intermanual transfer. The right half began criterion. The mean paths of Fig. 1 depicts the "unlearning" for 22 5s appear in Fig. 3. Clearly, or "re adaptation curve" which result- the initial response is in the di- ed when the prism-adapted S was ex- rection of the former placement of posed to normal vision and error- the target, but by the second micro- corrective feedback (Task 7) . The analyzed trial is essentially back curve falls quickly, mean trials to to straight ahead. criterion being 3.9 (SD = 2.9). The heavy vertical arrow depicts the dif- More relevant to Warren's pa- ference between postadaptation tar- per than the group means are the get pointing accuracy for the left individual differences revealed hand and accuracy after the "readapta- by the microanalysis. Fig. 4 shows tion" period. That is, it represents the group mean path for a midway the mean intermanual transfer of "re- trial on the initial adaptation adaptation" to the nonexposed limb. condition, as well as the indivi- This transfer was significantly dual paths for three 5s. It is greater than zero. clear that the individual paths vary markedly. Of interest is the In addition to exposing 5s to possibility that the differences the eight tasks, detailed recordings are related to other perceptual and analyses of the exact nature of traits. In order to examine this his hand and eye movements before, question, 5s were administered during, and after each pointing re- eight pretests which tapped three sponse were made by means of the classes of 5 characteristics. In trolley-boom apparatus and electro- all, there were 2 6 submeasures oculography system described earlier. taken from the eight pretests. These measures represented what Dr. These pretest data from a total Warren referred to as the "micro- of 4 8 5s were submitted to factor analysis." Secondly, the testing analysis. The varimax rotated period was preceded by a battery of factor solution produced nine fac- pretests designed to assess various tors, of which six (accounting for perceptual characteristics. 81.4 percent of the variance) have clear conceptual meaning and were The microanalysis of the hand used in further aspects of the activity was examined first when S analysis. The factors, very was adapting to the prism displace- briefly, were 1) eye control and ment and again during "readaptation. accuracy, 2) hand-pointing accu- During each of these periods it was racy, 3) hand- tracking (of a the initial two trials, a pair of visual target) smoothness, 4) eye- trials about midway, and the first tracking (of the unseen hand) smooth- pair of criterion trials that were ness, 5) right-hand accuracy in microanalyzed. For these trials the reproducing a right-hand position exact path of the hand was charted. and 6) left-hand accuracy in re- For the prism adaptation period the producing a left-hand position. A mean paths for 19 5s are seen in factor score for each 5 on each of Fig. 2. Four paths are depicted in the six factors was calculated by this figure: Trial 1 (1), the sec- summing his Z-scores for the sub- ond or third trial (2) , the midway measures that loaded on that factor. as trials (M) , and the first pair of These factor scores were used criterion trials (C) . It is clear covariates in a set of stepwise from Fig. 2 that, on the average. multiple regression analyses, one

103 88 90 92

MEAN RESPONSE PATH FOR:

1 FIRST ADAPTATION TRIAL 2 SECOND ADAPTATION TRIAL M| MIDWAY ADAPTATION TRIAL C; CRITERION TRIAL

MEAN RESPONSE PATH FOR; FIRST ADAPTATION TRIAL THIRD ADAPTATION TRIAL

C, CRITERION TRIAL

Figure 2. Mean Response Paths for Selected Trials from the 10' -Prism Adaptation Phase. Top Half— Hand Boom Set to 5's Left; ' Bottom Half—Hand Boom Set to S's Right. analysis for each of several depend- trials to criterion for either adapta- . ent measures. These dependent mea- tion or "readaptation " On the other sures included the number of trials hand, the multiple correlation of the to criterion in both adaptation and factors and amount of right-hand " re adaptation " phases, the amount of adaptation was 0.52, which is sta- right-hand adaptation, the amount of tistically significant. The two left-hand adaptation, and the pro- factors of eye control and accuracy portion of left-hand shift relative (15 percent) and hand-tracking (of to right-hand adaptation (intermanual a visual target) smoothness (9 per- trans fer) . cent) accounted for most of the variance. Thus, the better the eyes are The results revealed no correla- at fixating or replicating a former tion between the six factors and position, the more adaptation occurs;

104 the smoother the hand tracking, the adaptation showed a statistically less adaptation occurs. In brief, significant (p < 0.05) correlation. the better the eyes and the worse the hand is , the more adaptation oc- The last dependent measure, pro- curs . Conversely, the more the hand portion of the right hand adaptation can resist the conflicting visual in- shown by the left hand (intermanual formation, the less the adaptation. transfer) , proved to be multiply correlated 0.76 with the six factors. Multiple correlation of the fac- By far the most important factor here tors with left hand adaptation was was eye control and accuracy (r = 0.42, but only one of the 15 pretest 0.67). According to the submeasures submeasures that might on intuitive of this factor, the worse S was at grounds be relevant to left-hand maintaining fixation on a stationary

MEAN RESPONSE PATH FOR

1 FIRST ADAPTATION TRIAL 2 SECOND ADAPTATION TRIAL C, CRITERION TRIAL

88 90 92

MEAN RESPONSE PATH FOR

I FIRST ADAPTATION TRIAL 3 THIRD ADAPTATION TRIAL C, CRITERION TRIAL

Figure 3. Mean Response Paths for Selected Trials from the Clear-Glass 5 Readaptation Phase. Top Half—Hand Boom Set to ' s Left; 5 Bottom Half — Hand Boom Set to ' s Right.

105 88 90 92

Figure 4. Group Mean Path (X) and Individual Paths for Three 5s on a Trial Midway Through the 10°-Prism Adaptation Phase. target, the more intermanual transfer DISCUSSION he revealed (r = 0.51). Similarly, the worse S was at redirecting his Because of the pressure of time, eyes to a previously demonstrated the discussion period was cut short. direction of gaze, the more transfer A major point was raised concerning he showed (r = 0.62) . It is import- the implicit assumption that the ant to note that these relationships "perceptual traits" measured in the are in the direction opposite that pretests were, in fact, relatively for the right-hand adaptation measure. stable S characteristics (possibly Warren's explanation for this appar- the result of particular developmental ently anomalous result was that inter- histories) . Clearly, since they were manual transfer is the manifestation measured only once (during the same of a change in vision (rather than session as prism-adaptation measures felt-limb position) and that it is occurred) there is no evidence to sup- reasonable to assume that the better port this contention. Furthermore, the eye control the less the visual one of the discussants reported un- adaptation and hence the less the published data of his and several transfer. That is, for Ss with poor other investigators indicating the eye control most of their ipsilateral lack of a correlation between adapta- adaptation would be visual in nature tion to leftward prism displacement and all of this visual shift would and adaptation to rightward displace- affect left-hand pointing responses. ment, measured a week later. Thus, doubt is cast upon the notion of a Dr. Warren concluded by pointing trait of "prism adaptability." On out that his study, in contrast to the other hand, it was pointed out most previous ones in this area, was that the value of Warren's and Piatt's able to account for a significant findings does not depend upon whether proportion of the "error variance." or not such a trait exists. That is, This was done by an assessment not of even if a given 5's performance on the more global traits, such as "field some or all of the pretests were to dependence" (Witkin, 1949), but by vary from v;eek to week , the continued means of performance on simple per- presence of the correlations reported ceptual tasks. Furthermore, certain in the present paper would justify the independent variables (degree of hand use of this research strategy for pre- control) for future research have cisely the reasons which were stated. been suggested. Finally, this approach to the study of prism adaptation should, in principle, be applicable to most, if not all, complex perceptual phenomena.

106 The Possible Role of Eye Muscle Potentiation in Several Forms of Prism Adaptation

Sheldon M. Ebenholtz

University of Wis aonsin

The "Kohnstamm phenomenon," is muscles were "instructed" to move. the tendency for a stressed muscle Ebenholtz reported observing such an to remain innervated for some period effect in several studies in his lab- of time after issuing the voluntary oratory. Another prediction stemming signal to relax. An example of this from the previous evidence and argu- effect is seen when one stands close ments is that if the eyes are made to to a wall and presses the back of the converge for a period of time and hand hard against it; if one then then relaxed, there should be an af- moves some distance from the wall and tereffect of continued, non-monitored allows the arm to relax, it will rise convergence. If subjects are then involuntarily. The present research required to fixate a visual target, was based on the premise that the eye they will be forced to overcome the muscles are also subject to this convergence caused by the EMP. Thus, "post-tetanic potentiation," which according to outflow theory, this in this instance may be referred to extrano2rmal innervation should cause as "eye m.uscle potentiation" (E>^) . the target to appear farther away Thus, for example, the continued m.ain- than it actually is. Likewise, pro- tenance of a rightward eye position longed divergence should lead to ex- followed by relaxation of the eye periencing the target as closer than muscle should lead to an involuntary its true position. These predictions drift of the eyes to the right. How- were tested by Ebenholtz (and David ever, since there is no propriocep- Wolfson) in a study which he described tive feedback from the eye muscles, in detail. In this experiment, opti- the subject shold incorrectly feel cal devices were used which permitted that his eyes are still in the pri- both level of accommodation and tar- mary position. Several studies get im.age size to be held constant. (MacDougall, 1903; Park, 1969) have During the pre- and postexposure peri- reported just such an EMP effect. ods S demonstrated the apparent dis- Ebenholtz described a successful tance of the target by moving his un- replication of these results in his seen finger away from his body lontil own laboratory, using both eye rota- he felt it to be at the same distance tion and eye plus head rotation. as the visual target. During a 6-min- Similar effects have also been re- ute convergence (exposure) period S ported by other investigators for fixated a luminous target which was atypical eye convergence. Further- set to produce either near-distance more, if a visual stimulus (a dot of convergence (16.47°) or far-distance light in an otherwise dark room) is convergence (6.37°). For both groups present when EMP is causing the eyes the distance of the target during the to drift, the stimulus should appear pre- and posttests was intermediate to move in the direction opposite the to these two positions. eye drift. In order to maintain fixation in this event, motor com- The results supported the predic- mands must be issued to the eye mus- tions, in that near fixation led to cles, serving to counteract the EMP. postexposure overestimation of the However, according to outflow theory target distance, whereas far fixation (Festinger, Bumham, Ono, and Bamber, produced underestimation. In other

1967) , such centrally monitored com- studies, exposure periods of 2-, 3-, mands to the eye muscles should lead and 4-minute duration proved insuf- to the experience of visual movement ficient to produce these visual ef-

(or change of apparent visual locus) fects . in the same direction as the eye

10 7 .

Dr. Ebenholtz went on to point was above the pre-adaptation levels, out that adaptation to prism displace- suggesting that if 5 remained in the ment has in some cases been found to dark, adaptation would persist indefi- be due to a change in registered eye nitely. Furthermore, the asymptote position (Craske, 1967; Kalil and was somewhat higher for the higher Freedman, 1966). Thus, it is reason- prism strengths. Future studies will able to ask if these shifts are in- examine the EMP decay process in more stances of EMP and if EMP effects detail. underlie prism adaptation. Thus, for example, one can assume that during a According to Ebenholtz, the most prolonged period of prism exposure important implications of his studies the position of the eyes would, on bear on our understanding of the prism- the average, be turned in the direc- adaptive state. If theories which ar- tion of the prism apex (the direction gue for the existence of a truly visu- in which the objective straight ahead al form of adaptation are to be main- has been optically displaced) tained, efforts will have to be made to exclude the possibility of EMP. A second experiment was designed On the other hand, it may be the case to investigate this hypothesis, using that the visual aftereffects of most optical displacement in the distanae if not all previous prism adaptation dimension, effected by the use of a studies can be explained entirely in pair of base-out wedge prisms, which terms of this oculomotor phenomenon. cause the visual field to appear Ebenholtz emphasized the fact that the closer than it really is. Prism proposed component of prism adapta- strengths of , 6, 8, or 10 diopters tion does not represent a reoalibra- were used. During the exposure peri- tion, but merely the visual outcome od 5s walked around in a hallway for of an attempt to overcome a load 15 minutes, not being allowed to see which has been placed on the eye their bodies. The same measure of muscles. the apparent distance of a target was used here (during the pre- and posttests) as in the first experiment. Two target distances, 30 cm and 40 cm, were used during these tests. ,..

The results indicated the presence of a visual aftereffect for all DISCUSSION but the 0-diopter group. This is, upon removal of the prism spectacles A major point raised was that, the visual target appeared farther according to Dr. Ebenholtz, the pres- away than it actually was for the ence of an actual rearrangement of three groups whose vision had been the visual field (or even the presence optically foreshortened. There was of any visual field at all) should be no difference in the amount of adapta- irrelevant for the type of adaptation tion manifested by the 6-, 8-, and being proposed since all that is nec- 10-diopter groups. essary is the continued placement of the eye in a particular position, such A follow-up study demonstrated as off to one side. Conversely, if 5 that the prism adaptive change was is exposed to a prism-displaced visual not in the arm (proprioceptive) since field in which a fixation target is prism exposure did not lead 5s to mis- made to appear straight ahead, no vi- reach for a remembered 1-foot distance sual adaptation should result, since in the dark, while at the same time 5 is holding his eyes in a symmetric, the estimate of a 1-foot visual dis- straight-ahead position. Dr. Eben- tance was affected. Thus, the adap- holtz agreed with both of these points. tive component was visual, not pro- Another discussant suggested that Dr. prioceptive. Ebenholtz 's evidence that EMP is oper- ative in the "normal" prism adaptation Also measured in the main study situation is merely circumstantial. was the potential decay of the after- Ebenholtz 's reponse to this suggestion effects during a 30-minute period was that if the effects of, say, prism subsequent to the removal of the adaptation are not discernible from prisms. Decay occurred in all cases those found when 5 merely holds his and was found to reach asymptote after eyes to one side, and if the variables 10 to 15 minutes. Remarkably, asymptote

108 that affect prism adaptation also af- component of prism adaptation and fect EMP in the same way, then one eye muscle potentiation effects are can conclude that at least the visual one and the same.

Motor Transformation Learning

John C. Hay

Vntversity of Wis aonsin, Milwaukee

When a person turns his head or on stimuli by the motor activity. eyes to one side, the visual field These transformations must be rule remains apparently stationary despite following, rather than requiring a the fact that the image is sweeping listing of every possible stimulus across the retina in the opposite di- and its transform. Furthermore, it rection. This "constancy of visual is not necessary that the input to position" clearly requires some sort the "correlation memory" be the re- of mechanism by which muscular acti- sult of kinesthetic feedback from vity of the head and/or eyes is cor- motor activity; it may also come related with the visual input ("re- directly from the output of the cen- afference") which it produces. Stated tral motor areas, regardless if this differently, when a person moves his output eventuates in a motor response. head, say to the left, he "expects," possibly from past experience, to see In Hay's recombination procedure, a particular form of optical motion, a movement of 5's head caused a spot to the right, and, under normal cir- of light in an otherwise dark room to cumstances , this expectation is con- move by an amount and in a direction firmed. Stratton (1896) , Kohler determined by a small, on-line com-

(19 64) , and others have found that puter (PDP 8/L) . Thus, the computer the initial loss of visual position calculated a particular optical trans- constancy that human beings experience formation from the head-movement sig- when exposed to optical inversion or nal and displayed it in the form of right-left reversal is soon overcome. the spot on a cathode-ray oscillo- Thus, it is clear that human beings scope. In most of the experiments are capable of modifying the efferent- reported by Hay, the rules of trans- reafferent correlations necessary for formation were either 1) head up, position constancy. This suggests spot right; head down, spot left or (but does not prove) that the normal 2) he ad- up, spot left; head down, relations between motor activity and spot right. Typically, half of the its visual consequences are acquired 5s in an experiment were exposed to by the human neonate in the process 1) and half to 2 ) . The initial re- of making movements and observing the sponse to this situation is to see resulting optical motion. the light moving from side to side as S nods his head. However, after as Hay's research involves the use little as 30 seconds 5 perceives less of a rather novel form of recombina- of the lateral movement, which means tion procedure to examine the "rules" that adaptation is occurring and that underlying these ef ferent-reaf ferent visual position constancy is being re- correlations. His central thesis is gained. that organisms learn not the particular stimuli associated with motor ac- In the first experiment described. tivity, but the transformations imposed Hay attempted to see if Ss can adapt

109 . to a constant transformation between In another preliminary study S head nodding and optical motion when was confronted with a 4 X 4 matrix there is simultaneously present a while a spot of light appeared in a rar.d:^'n (up-down and/or right-left) random sequence in each of the squares, movement of light. This situation As the spot appeared, a number for is comparable to the "real world," in that cell was called off by E. After which an organism is exposed not only repeated practice, S learned the ap- to reafference but to optical motion propriate number for each cell. Sub- from other objects in the environ- sequent to these learning trials S ment. This latter motion, which is was asked various questions , such as independent of the organism's own "What is the sum of column 3?" or bodily movements, is referred to as "What is the difference between the " "exafference . sums of column 1 and row 3?" For half of the 5s these questions were With 72 5s Hay found, after a asked when the matrix was physically 6-minute exposure period, just as present, including the numbers (stim- much adaptation to the optical trans- ulus condition) ; for the remainder of formation in this situation with the 5s the matrix was present but simultaneous exafference as in a there were no numbers in it (memory condition in which no exafference was condition) . The results indicated present. That is, after the exposure that the time required to answer vari- period, when S was tested on the ous questions varied as a function of transformation alone, position con- the amount of arithmetic computation stancy adaptation was demonstrated, required but was always less in the indicating that the random exaffer- stimulus condition than in the memory ence had not interfered with the condition (by about 2.25 seconds). adaptive process. More im.portantly , however, the reaction time profiles over the various In a second experiment exaffer- questions for the two conditions were ence was introduced into the trans- nearly identical. Thus, it appears formation exposure in terms of a that either eye movements or rem.em- variable lag of time between head bered eye movements are used to ac- nod and optical motion. This, too, cess the memory of the visual targets. is analogous to situations in every- This, then, is analogous to sending day life, such as the case in which efferent impulses to the "correlator" the optical motion of an arm movem.ent and accessing the expected visual does not begin as soon as usual be- transformation cause the hand has been unexpectedly loaded by some weight. In the current experiment the time lag varied from 4 to 22 5 msec, with a mean value of 118 msec. The only group to show adaptation in this experiment was the one in which the pre- and posttests DISCUSSION were taken with a delay of 118 msec.

According to Hay's model, if the Many comments were elicted situation is such that S decides to throughout this presentation. A num- issue motor commands but then fails ber of questions, however, involved to execute them, the expected visual the need for clarification, probably transformation will be imagined. Dr. due to the large number of studies Hay reported a preliminary study in which were described. which S was asked to count "in his head" the stoplights that he passes The "matrix experiment" seemed every day when going to work. He to be of most interest. One question was required to press a lever as soon asked was if the results of this study as he thought of a stoplight. The implied that everyone has good visual prediction was that the tim.e between imagery. Hay responded by saying lever presses would be linearly re- that he did find differences between lated to the true physical distance people who said that they have good between the stoplights. Tentatively, visual imagery and those who said that this hypothesis was supported by Dr. they do not. In particular, poor vi- Hay's data. He argued that S accesses sualizers reported the need to make target items by employing covert motor overt eye movements over the empty activity.

110 . " matrix (the memory condition) while What is important for his model is good visualizers said that they did that the time for a visual search best on the task if they avoided (stimulus condition) be propor- making such movements. Apparently, tional to the time for a memory there was no difference in mean research (memory condition) . It v;as action times for these two types of not expected (and not observed) S. One discussant asked if it had that S would recapitulate the se- actually been necessary in the ori- quence in which the cells were original learning trials of the "matrix ginally presented. Rather, S was experiment" that the cells of the expected to move the eyes in what- matrix be presented to S in a rigid ever way he wished in either condi- (although random) order. Hay implied tion and was also allowed to use that this was not actually necessary. "short cuts .

Proposals for the Study of Adaptation to Anomalous Causal Schemata

' Ian P . Howard > -

York University

This presentation consisted of was visual feedback provided. Thus, two parts. Howard began by discuss- S was measured on 1) adaptation in ing briefly some recent (in press) the ipsilateral arm, 2) potential research on prism adaptation and then intermanual transfer, 3) felt limb concluded with a presentation of a position, and 4) felt direction of number of research ideas in the area gaze, respectively. of perceptual-motor schemata which he planned to implement in the near fu- It was found that for 9 (out of

ture . 22) Ss there was both a shift in felt limb position and a change in direc- In the first prism adaptation tion of gaze and that the two com- study described, the issue was "What ponents summed to the total correc- is it that adapts?" 5 was adapted tive target-pointing response of the by pointing at a target repeatedly prism-exposed limb. Furthermore, while the visual image was gradually intermanual transfer occurred only displaced to one side, by means of for these latter 5s, and it was equal rotary prisms, until a total displace- in magnitude to the change in vision. ment of 16° was achieved. This For the remaining 5s only felt limb "prism-shaping technique" precludes position was modified and its magni- S from becoming conscious of the op- tude was less than the total prism- tical rearrangement. At regular in- corrective pointing response. Thus, tervals throughout the adaptation per- Howard provided convincing evidence iod S was tested on various capaci- that prism adaptation, at least for ties: 1) pointing at a target with some 5s, consists of visual and pro- the prism-exposed hand, 2) pointing prioceptive components that combine at a target with the non-exposed hand, additively 3) pointing at the (unseen) big toe with the prism-exposed hand, and In the second prism adaptation 4) directing the eyes toward the study the aim was to test the hypothe- (unseen) toe. In none of these tests sis that prism adaptation will exist

111 . as a change (re calibration) in that generally absent in young children. part of the body that was active Only after the age of about six years during prism exposure. Howard had are children able to see that pouring 5 wear a long extension on his nose, water from a short, wide container with a light on the end of it. S's into a long, narrow container does not right index finger also had a light lead to a change in total volume. Ac- attached to it. In the dark and wear- cording to Piaget, one acquires con- ing prism goggles 5 either pointed servation and other schemata relative- to his stationary "nose light" with ly early in life, continues to hold his right or left index finger, or them as an adult, and is able to gen- pointed to his stationary right eralize them to new situations. How- or left index finger with his ex- ever, Howard said that he believes, tended nose. Because of the optical unlike Piaget, that adults can be like rearrangement he initially made er- young children with respect to situa- rors, which he was allowed to see im- tions with which they are unfamiliar. mediately after each response had As evidence for this position, Howard been completed. As a consequence of gave a demonstration of a common faulty this feedback, S eventually was able schema in adults. If a continuous to make the response accurately. loop of string is held in the form of a square and then changed to a rect- Based on the previously stated angle, most adults, when asked, will hypothesis, Howard predicted that 5s say that the total area within the who pointed with the hand should re- loop has remained unchanged. As the veal a prism-induced change in the rectangle is made narrower they will felt position of that limb, while Ss continue to hold to this conservation who pointed with the "nose" should of area until it becomes quite obvi- experience a change in felt head po- ous that the area is, in fact, dimin- sition. The results were, however, ishing to zero. Another common incor- in exactly the opposite direction as rect adult schema is that an object predicted. Dr. Howard argued that falling vertically will reach the this finding actually makes sense. ground before an object shot hori- That is, a stationary limb undergoes zontally, as from a gun. sensory adaptation which makes its felt position more variable (less Thus, according to Howard, it precise) , and there is evidence (from appears that rather than postulating his laboratory and others) that the the existence of developmental stages less acute a system the more adapt- in cognitive growth one must look for able it is. the kinds of experience which lead children (and adults) to acquire or The last half of the talk was unlearn their schemata. devoted to a discussion of a program of studies just begun, aimed at an Various kinds of recombination investigation of people's generative procedure can be used to reveal, modi- perceptual schemata (cognitive struc- fy, and reinforce a person's schemata. tures, assumptions, etc.). Schemata In one such procedure Howard has in- are assumptions about the nature of vestigated the schema people generally the world which people hold and typ- hold with respect to the acceleration ically never question. Often, one is of falling objects. 5 is confronted unaware that he holds these assump- with a vertical screen having three tions unless they suddenly fail or small windows located 1, 2, and 3 feet lead to contradictory experience. The below the top. He is asked to hold an "phantom limb" experience is a good iron ring at the top of the screen, example of a situation in which the directly behind it (and thus not visi- existence of a schema (body position) ble to him) . He is to release the ring is not revealed until its fit with and watch it pass by each of the three reality slips (an arm is removed) windows. His experience, typically, Thus, the "recombination procedure" is one of surprise, as the ring seems may be used as a means of revealing to appear in the lower windows sooner the existence and nature of schemata. than expected.

Piaget has carried out well-known A variation of this technique is studies of the development of schemata to arrange the situation so that the in children. He showed, for example, falling ring is stopped (by a ledge) that the principle of conservation is and a second ring is made to fall past

112 . . . .

the window at whatever latency E de- It would , of course, be possi- sires. Thus, E can vary the apparent ble to re info rce incorrect schemata. falling acceleration. By asking S For example, E could catch the fall- when the arrival of the ring at the ing ring and then let another one various windows "looks right" it is fall in such a manner that 5's possible to measure precisely the na- (incorrect) e xpectation of the time ture of S's schema. of appearance was confirmed, allowing him to be accurate- from the very In a preliminary study described start in his attempts to catch the by Howard, S was trained to stable, ring accurate performance at catching the falling ring with a rod at the one- The presentation was concluded foot distance. Then S's attempts were with some suggested reasons for com- measured at the 2- and 3-foot windows, mon incorrect schemata held by adults, with the ring actually stopped by E For example, the pre-Galilean assump- after passing the first window. 5's tion that heavy objects fall faster errors again served to reveal his in- than light objects might represent a correct schema (an underestimation of generalization from the fact that gravitational acceleration) . In a heavy objects in a viscous fluid do further study S was tested (no knowl- fall faster. Similarly, the force edge of results) on an 8-oz ring and with which a spring in a pinball on a ring of half the weight dropping machine is loaded is proportional through the same height. It was to the speed with which the ball found that most 5s behave as if they travels assume that the fall time decreases by a factor of about /2 for a halv- ing of weight. This is the function that follows from the (incorrect) assumption that acceleration is proportional to force (weight) DISCUSSION Thus , the preceding techniques allow for the measurement of schemata. More important, perhaps is the fact Several discussants pointed out that by aon fronting S with the in- that the assumptions (correct or in- correctness of his schema (by allow- correct) held by most individuals are ing him to reach for the falling ring undoubtedly the result of previous at the second and third windows) it interactions with the world. Thus, may be possible to correct them. A for most people, experience has taught number of training techniques for this that heavy objects do fall faster than purpose were suggested. An example light objects (due to air resistance) described by Howard is one in which Thus, the "real world" is much "dirt- two identical self-energized toy cars ier" than the idealized world of the are run, either singly or linked in physicist, and maybe we should be pairs, under a horizontal screen to studying "dirty schemata. " Howard emerge at the far side. S must anti- apparently had no quarrel with this cipate the time of emergence of single as being a worthwhile topic of study, and linked cars. The appropriate but nevertheless felt that it is use- schema for success at this task is ful to know what people's schemata that linking the cars (doubling the are and to see if these schemata can weight) does not affect their speed. be modified by experience and specific Most adult 5s would probably see this training techniques. immediately and might then be able to apply the principle to the falling- ring situation.

113 The Effect of 5's Interpretation of "Straight Ahead" Upon Measures of Prism Adaptation

Charles S . Harris

Bell Laboratories

Two coiimionly used prism-adaptive immediately report a new position for measures are the pre-post shifts in straight ahead, one that is in line pointing straight ahead with eyes with the apparent orientation of the shut and reporting where the vi- visual field. Because Es have tended sual straight ahead appears to be lo- to assume that 5s were responding in cated. The first is considered to be terms of the mid-saggital plane, this a measure of change in felt limb po- change has been incorrectly interpret- sition and the second a measure of ed as an immediate visual shift. Sec- visual shift. Harris pointed out that ondly, if S were then to close his when S is viewing a structured envir- eyes and point straight ahead, he onment, one that includes walls and would probably point in the remembered other rectilinear objects, E's com- direction of the visual axes of the mand to point (or look) straight room. This change in arm placement ahead can have either of two meanings (as compared to the pre-exposure re- to 5. On the one hand, S may assume sponses) will appear to £ to be an that E is asking that he point (or an ti -adaptive shift in felt limb po- look) straight ahead of his body. It sition. Harris mentioned a number of is, in fact, this interpretation that studies (some from his own labora- all Es assume (or hope) S is making. tory) which have reported either an On the other hand, S may interpret "immediate correction effect" (Rock, the command to mean that he is to Goldberg and Mack, 1966) or maladap- point (or look) in a direction that tive reaching responses (McLaughlin, is in line with the walls of the room Rifkin and Webster, 1966). (or toward the center of the wall which he is facing) . If 5's vision The point is that what appear to is undistorted and he has been seated be immediate prism-induced changes in in line with the sides of the testing vision and/or proprioception may be room and/or the sides of the table in only the manifestation of S's inter- front of him, his behavior of point- pretation of the concept of "straight ing (or looking) straight ahead will ahead." As such, these changes are be the same, regardless of which as- both trivial (if one is interested sumption he holds. However, if prism in perceptual modification) and mis- goggles are placed on his head, the leading. Harris pointed out that in nature of his assumption can have an the research of Singer (this confer- important bearing on his response. ence) 5s whose visual fields were op- If the structured visual field is tically tilted responded to the com- prismatically displaced to one side, mand to indicate the apparent vertical S will have the same visual experi- by showing an "immediate correction ence as he would if he were sitting effect." However, with eyes shut and at a slant with reference to the sides asked to set a rod to kinesthetic ver- of the room. If in this situation he tical they placed the rod at a tilt interprets "straight ahead" to mean a that, if indicative of perception, point in line with his mid-saggital would represent a maladaptive shift plane (as E assumes) , then (at least in kinesthesis. It appears likely, initially) he should continue to look then, that here too 5 was using the (or point) in the same direction as visual framework, rather than his before the prisms were placed in front body, as the basis for his judgments of his eyes. On the other hand, if he and that what looked like changes in uses the visual framework as his basis perception were not. for judgments of straight ahead, two things will happen. First, S will

114 .

Now it is likely that many Ss if means of avoiding the latter possi- tested in a structured field will bility, Harris suggested the pro- make their judgments of straight ahead cedure of leading S into the room in terms of a aompromise between the blindfolded, having him set in a two contrasting interpretations chair, and then slowly rotating him Furthermore, if the exposure period into position for testing. continues for some time (and involves active visuomotor behavior along with visual feedback), "real" adaptation would occur, which will tend to aug- ment (or replace) the spurious change in visual straight ahead. Likewise, the apparent maladaptive shift in pointing straight ahead might be ne- gated, leaving what appears to be the total absence of proprioceptive adaptation.

Because Harris' entire argument is based on the assumption that S is DISCUSSION using a structured visual frame of reference to determine (or strongly influence) his judgment of straight In the brief discussion that ahead, the obvious precaution is to followed Harris' talk, questions omit such a field. This is easier centered on whether or not he thought said than done. Even if a dark room that asking S to look or point straight is used, S is likely to be able to ahead could ever be a trustworthy mea- see the testing apparatus during the sure of perceptual adaptation. He exposure period, and if the apparatus said that he thought it could if the is rectilinear it may have the unde- precautions he described were imple- sired effect on judgments of straight mented. Several discussants asked ahead. Of course, if the testing ap- why it was that one couldn't merely paratus is designed to present a instruct S on what E meant by "straight curved array (Uhlarik, 1972), the ahead" and even give him a demonstra- problem is reduced. Nevertheless, tion and/or pictorial representation. one still must be wary of the possi- Harris responded that while the use bility that S may remember how the of appropriate instructions seems to testing room appeared from when he be an obvious solution it will prob- first entered it and that this memory ably fail in many cases, since 5s do may influence his judgments. As a not generally seem able to switch consciously from one assumption to the other.

115 . . . .

Speculations on a Model of Prism Adaptation

Robert B. Welch

University of Kansas

Welch gave a short presentation fixation response. Finally, although of some ideas and data of his which the RR component can be elicited by suggest that adaptation to prism- non-active prism exposure (Howard, displaced vision entails, at least Craske , and Templeton, 1965) or con- in some situations, three different current exposure without a target, components the best possible situation is one in which S is confronted with a tar- If S points at a prismatically '•* get and exposure is of the terminal displaced target he 1.) experiences variety. In this situation S will a discrepancy between felt and seen make several initial target-pointing position of the limb and 2.) makes errors and then, typically, engage in target-pointing errors. If, on the a conscious correction and continue other hand, S views his optically to practice this correction until it displaced and actively moving limb becomes "automatic." What is crucial in the absence of a target, then the here is that 5 ultimately make the only cue to the presence of the re- prism-adaptive target-pointing rearrangement is the discrepancy between sponse "without thinking." seen and felt limb position (a discrepancy which may exist even if limb Thus, according to this model, movement is passive, but is probably NA = PS + VS + RR. That is, the to- more salient with active movement) tal target-pointing aftereffect is the combined result of feeling that The behavioral measure of prism the arm is displaced by a certain adaptation in most experiments is the amount in the direction of the prism- pre-post shift in target-pointing displacement plus seeing the image accuracy without optical displace- i of the target displaced toward its ment or visual feedback. This is objective position plus a rule for referred to as the "negative after- correct pointing. effect" (NA) . However, several investigators (Harris, 1965; Wilkinson, The existence of PS and VS has 19 71) have argued that NA is actually already been demonstrated by a number the sum (or product) of several dif- of investigators. It is the third ferent components. VJelch went on to component, RR, for which supporting describe a proposed multi-component data must be provided. Welch presented model of prism adaptation entailing some data (mostly unpi±)lished) indi- three different components: 1.) a cating that when independent measures change in felt limb position, or of PS, VS , and NA are taken from the "proprioceptive shift" (PS), 2.) a same Ss (and the order of postexposure change in vision, or "visual shift" measures is counterbalanced) NA sig-

(VS) , and 3. ) a prism-corrective mo- nificantly exceeds PS plus VS. Furth- tor response, or a "rule for reach- v ermore, the correlation between NA ing" (RR) and (PS + VS) is quite low (approximately + 0.35). Both of these find- Welch suggested that the stimu- ings support, in an indirect way, the lus for PS was the conflict (direct existence of a third component. That or indirect) between felt and seen this derived third-end state of prism limb position. VS may result from adaptation is, in fact, a motor learn- an asymmetry of the visual field (the ing component is supported by the fact

"immediate correction" effect) , from that when prism exposure does not en- very well-practiced, prism-corrective tail a target there is, according to visuomotor responding, or from the Dr. Welch's data, little or no dif- automatization of an asymmetric eye ference between NA and (VS + PS)

116 .

Welch suggested that the proposed this appeared to be contradictory to motor learning component is more sus- his notion that a third prism-adaptive ceptible to "learning variables" (see component exists. Several people men- Taub , this paper) than are the other tioned the fact that since decay of two components. Evidence for this prism adaptation seems to be rather hypothesis comes from several studies. slight, at least for 10 to 15 minutes In one it was found that the RR com- after the adaptation period, it may ponent occurred when prism-exposure be that the procedure of counterbal- trials were distributed, but not when ancing the order of postexposure mea- they were massed. In a second study, sures is not necessary. On this same it was found that brain-damaged 5s line another discussant argued that failed to acquire RR, while non-brain- the best solution would be simply to damaged 5s did. measure 5 on only one of the types of adaptation. Welch agreed that this A final point made by Welch was would solve the problems of 1.) po- that whether or not an investigator tential decay of prism adaptation, will actually be able to measure all 2.) potential differential decay of three prism-adaptive components may the various components, and 3.) ob- depend upon the nature of the tests taining the precise measure of the used. For example, a slow target- absolute magnitude of each component. pointing response in the postexposure A distinction was raised by one of test should maximize the contribution the discussants between 1.) the ef- of the PS component to NA, while a ferent (motor learning component) ballistic response should reduce its proposed by Welch, and 2.) the effer- importance , except for that amount ent sense of body position postulated which exists at the starting position by Taub (this conference) with respect of the limb to his deafferented 5s. The latter is not what Welch was talking about. In defense of a three-component model, one of the discussants argued that even if the sum of PS and VS is found to equal NA, there might still be a third component since it is not DISCUSSION clear that the measures of PS and VS can actually be independent of each other. That is, it is possible that A number of interesting points the individual measures of VS and PS were raised concerning Welch's pro- may be overestimates of their ef- posed three-component model of prism fects during the total visuomotor adaptation and the ways in which it response used as a measure of NA. should be tested. One discussant If that is true, then when the sum said that Hay and Pick (1966) had of PS and VS equals NA, it really found that PS + VS = NA in a study in means that a third component (RR) which 5s wore prism spectacles contin- is present. uously for 42 days. Welch agreed that

117 .

Past, Present, and Future of the Recombination Procedure

Ivo Kohler

University of Innsbruak

The closing address of the con- One of the problems encountered ference was designed to review the by Kohler (and Erismann) is the fact origins of the use of recombination that lenses used for tilting or in- procedures, to tie together, or at verting the field also constrict the least comment upon, the previous size of the field. Two ways for over- talks, and to suggest the directions coming this problem were devised. One in which future research will go. involved the use of mirrors, since they can be made as large as desired. Kohler began by acknowledging a For example, if a mirror is worn like debt to the pioneering studies of op- the brim of a hat and 5 looks up into tical inversion by Stratton (1896) , it, he will see the floor and his body not just in terms of his findings inverted. The disadvantage here is but because of his conscious attempt the fact that S must place his eyes to restrict as little as possible the in an atypical position. A second range of relevant observations. The technique for providing S with a large technique of having S wear distorting rearranged field is to build an envir- spectacles for prolonged periods of onment (such as a small room) and then time in a "real life" environment tilt or even invert it. while reporting all relevant perceptual and perceptual-motor experi- Kohler discussed several unpub- ences was the basis of the famous lished studies using the recombination "Innsbruck studies" by Kohler (and procedure. In one study a S wore Erismann) . Such an "openness" led prism goggles for eight days while to the discovery of the "situational engaging in everyday activities. At aftereffect." An example of this the end of the 8-day period he was re- phenomenon is the finding that after quested to directly face a wall. It wearing spectacles with left halves was found that S could comply quite blue and right halves yellow, 5 comes accurately with this request, suggest- to experience a complementary color ing that he had adapted visually to aftereffect with spectacles removed, the prismatically induced rotation of which is contingent upon the direc- the fronto-parallel plane. However, tion in which he looks. Analogous when he was asked to face straight findings have been reported more re- ahead with eyes shut, it was seen that cently (McCollough, 1965; Hajos, 1970). he held his head off to one side rela- He argued that much of our everyday tive to the rest of his body. Thus, experience may involve various con- this result supports a much later sug- tingent-produced visual perceptions gestion of Harris (1965) that what at the existence of which only become ap- first seems to be an example of visual parent with optical rearrangement. For change may actually be a manifestation example, if the visual field is in- of a change in felt eye and/or head verted, falling objects seem to move position much faster than if seen to fall with vision normal. Evidently, the experi- The great i nterest in adaptation ence of the apparent speed of objects to rearranged vi sion that has existed depends (is contingent) upon the di- in the United St ates since the re- vs. rection (up down) in which they search of Held ( 19 55), was attributed move. Another example is the fact not only to Kohl er and Erismann 's work that the perception of facial expres- but , more import antly, to the ideas of sion is drastically altered if the Hoist and Mittel stadt (1950) . Ameri- face is seen upside down. cans have been p articularly interested

118 , , .

in breaking down adaptation into its found that people who were very good

various components , as has been seen at avoiding obstacles were also quite in several of the talks at this con- sensitive to differences in the tim- ference. This, however, is a diffi- bre of sounds (echoes) cult task, since these components are probably not independent of each In response to Ebenholtz's talk, other. he asked how eye muscle potentiation could explain the adaptation that has Kohler commented briefly on each been demonstrated for split-field

of the previous talks. He suggested prism goggles (Kohler, 1964) . That that with respect to Taub's paper, is, in the case of adaptation to spec- the finding of 100 percent adaptation tacles which displace the upper half in deafferented monkeys might be in- of the field but not the lower half, terpreted in terms of an increase in it does not seem possible to invoke

the length of the feedback loop . Thus eye muscle potentiation since the lat- feedback in the deafferented monkeys ter implies an all-or-none state. may have been delayed, and it is well Rather, it seems necessary to assume known that delayed feedback circuits that adaptation to split-field prism are accompanied by an increasing in- spectacles is gaze contingent. stability (and therefore adaptability) of the system. Kohler related two observations from his laboratory that were similar Kohler disagreed with the cri- to several of the studies which Hay terion used by Singer to distinguish described in his presentation. In sensory spatial adaptation (SSA) from one it was observed that if 5 wears

behavioral compensation (BC) . It wedge prisms and moves toward a fixa- will be recalled that the distinction tion point straight ahead, the point was made partly in terms of the lon- will appear to move laterally toward gevity of the aftereffects — those due the prism base. However, after a few to SSA decaying in 2 to 3 minutes days this illusory movement will dis- while BC lasts much longer. He point- appear. If the goggles are then reed out the existence of very long- moved, the fixation point will again lasting aftereffects that, neverthe- appear to move, but now in the direc- less, appear to be examples of SSA, tion of the prism apex. A second, because they result from passive ob- related observation is the case of servation of unchanging stimulation; a person accustomed to wearing eye an example is the "McCollough ef- glasses. If the lenses are removed " fect . from the frames and the person moves these empty frames about while look- Kohler was very interested in the ing through them, a fixation point study described by Uhlarik because he will appear to move. However, this had done an unpublished study himself will not happen if someone else moves many years ago in which Ss walked the frames in front of the person's around while viewing a visual field eyes. In both the prism-wearing and that was undergoing constant (and lenseless-spectacles situations, per- quite rapid) change due to counter- sons acquired a constancy of visual rotating prisms. Unfortunately, 5s position as the result of prolonged became so nauseated that the study experience with a particular trans- had to be discontinued before any information. When the prism goggles terpretable data could be gathered. or spectacle lenses were removed, this constancy was no longer appro- Referring to Warren's presenta- priate and, hence, resulted in after- tion, he said that he was very im- effects. pressed with the general technique of "microanalysis." There are a few In response to Howard's discus- previous studies (Eysenck, Granger, sion of causal schemata, Kohler de- and Brengelmann, 1957) of individual scribed some observations that he had differences, but typically they used made of the experienced acceleration "abnormal" 5s (e.g., neurotic, schizo- of falling objects in an inverted

phrenic) . Kohler also referred to a visual field. If one is confronted study in which the perceptual traits with a room that is truly upside down of blind individuals were correlated but which appears, because of roof with their ability to move around prisms, to be rightside up, objects without bumping into objects. It was will be seen to "fall" toward the

119 .

ceiling. Furthermore, the accelera- behavioral measures , as seen in the tion of these objects will be per- research on "backward masking" ceived as much greater than in the (Gilinsky, 1967) case of objects that are falling, as they are supposed to, toward the The recombination procedure may floor. Kohler concluded that we may lead to a further understanding of underestimate acceleration to which the perceptual analyzers whenever it we are adapted (as Howard showed) and is found that some sorts of rearrange- overestimate acceleration that is ments are easier to adapt to than atypical. others. For example, the failure to find any adaptation whatsoever to a Finally, Kohler spoke of his particular form of optical rearrange- views with respect to the future of ment would undoubtedly have implica- the recombination procedure. One of tions for the nature of the perceptual the aims of future perceptual re- analyzers involved. search, he believes, is to come to an understanding of how the central Kohler concluded by saying that nervous system processes, simplifies, he believes that two important lines or otherwise makes sense out of the of future research in the area of myriad of stimuli bombarding 'the sen- recombination are an examination of sory receptors, usually from several "contingent adaptations" (of which different modalities at the same time. he spoke earlier) , and an examina- There is an increasing tendency for tion of how the nervous system re- psychologists (and others) to view sponds to the sensory invariances the brain as an exceedingly complex that demonstrably exist on the re- computer, one which has evolved to ceptor organs (Gibson, 1966). its present state in man as a result of a series of successes and failures to meet the demands of the environment. It is possible to examine the nature of the "analyzers" by means of physiological techniques (Hubel and Weisel, 1959) or by inference from

120 , , ,

REFERENCES

Craske , B. Adaptation to prisms. Held, R. , & Freedman, S. J. Plasti- Change in internally registered city in human sensorimotor control. eye position. British Journal Science, 1963, 142, 455-62. of Psychology , 1967, 58, 329-35.

Held, R. , & He in, A. Adaptation to Eysenck, H. J., Granger, G. W., & disarranged hand-eye coordination Brengelman, J. C. Pe raeptual contingent upon reafferent stimula- processes and mental illness. tion. Perceptual and Motor Skills, New York: Basic Books, 19 57. 1958, 8, 87-90.

Festinger, L., Burnham, C. A., Hoist, E. von, & Mittelstadt, H. Das

One, H., & Bamber , D. Efference Reaf ferenzprinzip. Naturwissen- and the conscious experience of schaften, 1950, 37, 464-76. perception. Journal of Experi- mental Psychology Monograph Howard, I. P., Craske, B., & Temple- 1967, 74 (4, Whole NO. 637). ton, W. B. Visuo-motor adaptation to discordant ex-afferent stimula- Freedman , S. J. (Ed.) The neuropsy- tion. Journal of Experimental Psy- chology of spatially oriented be- chology , 1965, 70, 189-91. havior. Homewood, 111.: Dorsey Press, 1968. Hubel, D. H., & Wiesel, T. N. Recep- tive fields of single neurones in Gibson, J. J. The senses considered the cat's striate cortex. Journal

as perceptual sy stems . Boston: of Physiology , 1959, 148, 574-91. Houghton Mifflin Co., 1966. Kalil, R. E., & Freedman, S. J. Per- Gilinsky, A. S. Masking of contour- sistence of ocular rotation follow- detectors in the human visual ing compensation for displaced vi-

system. Psy ahonomic Science , sion. Perceptual and Motor Skills, 1967, 8, 395-96. 1966, 22, 135-39.

Harris, C. S. Perceptual adaptation Kohler, I. The formation and trans- to inverted, reversed, and dis- formation of the perceptual world. placed vision. Psychological Psychological Issues, 1964, Mono- Review, 1965, 72, 419-44. graph 12.

Hajos , A. Sinnespsychologische Kohler, W., & Wallach, H. Figural Untersuchungen zur Invarianzbildung after-effects: An investigation im visuellen System des Menschen. of visual processes. Proceedings Paper read at Kybernetik-Kongress of the American Philos ophi cat So-

Berlin, March, 19 70. ciety , 1944, 88, 269-357.

Hay, J. C, & Pick, H. L. , Jr. Visual McCollough, C. Color adaptation of and proprioceptive adaptation to edge detectors in the hioman visual optical displacement of the visual system. Science, 1965, 149, 1115- stimulus. Journal of Experimental 16.

Psychology , 1966, 71, 150-58. MacDougall, R. The subjective hori- Heinemann, E. G. Figural after-effects zon. Psychological Review Mono-

in kinesthesis : Effects of object graph. Supplement , 1903, 4. width and repeated presentations. Journal of Experimental Psy ahology McLaughlin, S. C. , Rifkin, K. I., & 1961, 61, 51-56. Webster, R. G. Oculomotor adaptation to wedge prisms with no part Held, R. Shifts in binaural localiza- of the body seen. Perception & tion after prolonged exposures to Psyahophysics, 1966, 1, 452-58. atypical combinations of stimuli. American Journal of Psychology, Park, J. N. Displacement of appar- 1955, 68, 526-48. ent straight ahead as an aftereffect

121 ,

of deviation of the eyes from nor- Uhlarik, J. J., & Canon, L. K. In- mal position. Perceptual and Mo- fluence of concurrent and terminal tor Skills, 1969, 28, 591-597. exposure conditions on the nature of perceptual adaptation. Journal Rock, I., Goldberg, J., & Mack, A. of Experimental Psychology , 1971, Immediate correction and adapta- 91, 233-39. based on viewing a prismatic- tion Weiss, P. A. Self-differentiation of scene. Perception ally displaced the basic of coordination. 351-54. patterns i Psychophysics, 1966, 1, Comparative Psychology Monograph 1941, 17, 1-96. Stratton, G. M. Some preliminary on vision without experiments Wilkinson, D. A. Visual-motor control the retinal image. inversion of loop: A linear system? Journal of Psychological Review, 1896, 3, Experimental Psy chology , 1971, 89, 341-60, 611-17. 250-57.

J. A device for presenting Uhlarik, Witkin, H. A. Perception of body po- targets and recording positioning sition and the position of the vi- Be- responses in one dimension.- sual field. Psy chological Mono- Methods and In- havior Research graph, 1949, 63, No. 7. strumentation, 1972, 4, 15-16.

122 , ^ .

THE LIGHT PERMEABILITY OF THE RETINA: EXPERIMENTS WITH BLIND, COLOR-BLIND, AND SEEING SUBJECTS*

Johan Weckroth** Pentti Miettinen** Eila Weckroth**

The most essential functional Correspondingly, a predominant characteristic of a sense modality, anatomical characteristic is a cer- the eye in particular, is its ability tain duplexity of retinal substric- to achieve within enormously varied tures which form a bipolar modal unit stimulus and background conditions (such as the two responsible for

an adequate level of habituation and photopic and scotopic vision) . On to maintain its sensitivity in both the level of neural events we recog- directions. This ability is obviously nize this bipolarity in concepts such not based on characteristics of a as "excitatory and inhibitory zones single receptor-neuron unit, but is of a receptive field," "simple and achieved by, or is dependent on, in- complex field," "on-off response chan- teraction between neural units. This nels," and, finally, in the most gen- dependency has been especially empha- eral form at the behavioral level in sized by Pribram (1971) and also by the reciprocal phenomena of "contrast

Glezer, Bertulis , Ivanov, Kostelyan- enchangement " and "adaptation." This ets , and Bodvigin (1971). On the bipolarity is further reflected in basis of recent results Pribram has reference concepts such as "jump," also been able to ascertain that in "shift," and "neutral point" when this connection an efferent control reactions of the eye are illustrated mechanism also operates within the graphically. visual system (Glezer, et al. , 1971, p. 44) . On the other hand the Glezer Consideration of the eye as a group has emphasized that the inter- servo system implies that the basic action may be based on different mech- receptor units or their interaction anisms at different retinal levels, have not always been the same. It thus indirectly stressing the gener- further implies that the eye as a ality of this functional character- part of the brain has been able to istic of the visual system (Hartline, develop its receptor system which 1940; Hubel & Wiesel, 1962; Ratliff, originated either from a polygenetic 1965; Kuffler, 1953; Pribram, 1971, or monogenetic initial stage." From p. 64) . the above, it seems logical to assume that the bipartite systems once developed, were apparently founded on some base, a "reference background," *The investigation is included in "point of support," "pivot" or what- project 417-4 551-3 18017620-9 ever the name may be for a center financed by the Finnish Academy. around which the bipolar system could be built. As a matter of fact, this type of development seems impossible **University of Tampere, unless the system can establish or Finland. find a "hinge" around which the element can turn

123 Because we imagine t±ie develop- characteristic and the light perme- ment on this basis, the choice be- ability of the retina. tween monogenetic and polygenetic views naturally influences our con- Surprisingly little has been cepts regarding the potentialities said in the literature^ regarding of the primary elem.ents , and the the permeability of the retina or its manner of the development. The dif- connection with the fundus in spite ference is perhaps most distinctive of, say, the "perverted" posture of with respect to the concept of a the receptors, with the pigment epi- "hinge," and its location and char- thelia being the last element from acteristics. The simpler postulation the point of view of the source of of one single primary unit leads even light. The pigment epithelia actu- here to a simpler model of thought, ally lie as if their aim were to so far as it seems reasonable to lo- prevent most free re- flow of light cate the point of support outside of from the background of the eye to the element. But the differences the nerve endings and perhaps through must not be exaggerated, not even in the whole retina. other respects. The difference between the microanatomical or- spectral Is the above-mentioned "gain" structures of what we think are basic connected with this turning of the receptor elements may also be osten- receptor elements? If so, there are sible. Recent analysis of inverte- losses as well in the form of reduced brate retinae by Wasserman (1973) general sensitivity. Can it be that may be of central relevance here. color perception as an independent He presents findings against what he entity of sensation would not be pos- calls the "tuned receptor paradigm" sible without this "perverted" pos- by listing examples of invertebrate ture? Could this base be considered retinae (crab, crayfish, bee, etc.) as a "conditio sine qua non" for all which have non-tuned receptors. He integrative functions based on neural actually draws the picture of the interaction or dealing with percep- basic element perhaps most urgently tion of color, just because it is needed in the development outlined homogeneous and colored? The func- above, when he describes the non- tions we have in mind are those such tuned, double peaked, bipolar "beta" as the efferent control mechanism, cell as being able to "change the I - E components of the receptive relative sensitivity of the two field, on-off components, etc. And, peaks" and states, "that at least finally, does this base and turning some 'alpha' cells are really beta toward it also form a necessary con- cells which have lost some of their dition for the development of the light sensitivity. " Since our aim subcomponents (by means of genetical is not to validate either of these selection), i.e. for the differentia- views we refer to the discussion at tion of "alpha" cells, because in- the end of this article, which, in a vertebrate eyes which lack this turn- sense, deals precisely with the ques- ing may often also have retina com- tion of what is gained with that posed of non-tuned, double-peaked re- "loss." ceptors, such as the honeybee.

Obviously, even momentary bi- Based on the above , what would polar function needs a base or a happen if light could be conducted neutral point. The "hinge" in turn directly toward or through the fun- apparently has its counterpart out- dus, otherwise bypassing the eye (in side of the eye; being alternately vivo) . In other words, what would some predictable level of light in- be the sensation when the receptors tensity or adaptation, contrast, col- of the retina were forced to rely or, or the like. So the question for their orientation exclusively arises whether all bipolar functions on the light reaching them from the of the present, and the phylogenetic background of the eye. If perception long-term development of the past of something actually exists, what is could be content with a common base the direction, shape, and color of as a point of support. A search for the perceived object, and do indivi- such a structural property led to dual differences exist? the assumption that such a property could be an innate characteristic of the background of the eye in cooperation with its own reflectance

124 METHOD sign pertaining to color weakness or blindness in an Ishihara

The light was directed through test (given by the nurse) . La- the fundus of the eye via the mouth, ter a replication group were A round, 12 mm, acrylic plastic rod collected from a total group of was used, bent at one end about 90° 115 boys (all from the top form,

in order to direct the light upwards. mean age between 14-15) . From The other end of the rod was fixed to this group were selected 18 boys a can which concealed the light of whom 9 were evidently color source, a normal 45 W incandescent blind, 4 evidently non-color lamp. The rod was painted with blind and the rest unilateral strongly reflecting white paint, and or other problematic cases. then with black paint to hinder light penetration. Normal The subject was asked to hold the pipe in his mouth so that its This group was composed of 12 open end was directed upwards towards subjects with no sign of color the palate, the mouth closed as tight- defect. The majority of this ly as possible. He was instructed group were students or other to push a button if a change was per- persons working at the labora- ceived. At the beginning of the test tory. Two African students were a 5-10 minute dark adaptation time intentionally included in the was required but on retesting only group. four minutes were necessary."^ The light was switched on at irregular intervals. A positive reaction was RESULTS recorded when the subject pushed the button promptly after the flash was Blind Subjects switched on, in a sequence of twelve stimuli.^ The subject was then asked Diagnostic description of sub- to describe verbally the quality of jects and their reactions to the the change he perceived. If he said "light pipe" test are summarized in he saw a light he was further asked Table 1. to tell about the quality of the light and about the direction from which it The term blind is naturally a seemed to come. legal term conditioned by local circumstances. It is, however, a common observation that only a fraction of SUBJECTS the group are totally blind; estima- tions varying between 10 and 30 per- Three groups were used as sub- cent depending dominantly on the cri-

jects : teria applied in determination of the defect. In this orientation trial we were not anxious to keep the group Blind homogeneous with respect to the degree or severity of the blindness or Fourteen subjects diagnosed as with respect to other relevant char- blind or as having severe visual acteristics, such as etiology, dura- impairment were chosen randomly.^ tion, age at onset, or quality of the The search was originally focused blindness. Hence the picture obtained on subjects born blind but having is somewhat inconsistent. With re- presumably intact retina. The spect to qualitv, however, we hoped extreme rarity of such cases to have included subjects with some forced us to give up before any- sort of otherwise verified neural de- one was found. fect, who could give us a hint as to the practical usefulness of the device as a diagnostic instrument. Color Blind Along the structural continuum, The first group was comprised of lens-retina-neural net, the apparatus a sample of 12 boys from a nurse's seems to be most effective as a de- list of 24 pupils in elementary tector of expressly retinal malforma- school. The list embraced all tions or lesions. In general, subjects pupils having any kind of deviant who sense some light from ahead in

125 — ,

TABLE 1

Blind Siabjects

N Age Sex Status Test Reaction

Totally Blind

19 F Visual acuity in both eyes None

none . Congenital cataract. Retrolental fibroplasia.

19 F Visual acuity in both eyes-- Sees light coming from the none. front of the left eye, white Congenital cataract. light, no changes on the Chronic uveitis. right side.

Poor Visual Acuity in Both Eyes

32 F Right eye 0.1, left eye f.c. 2 m. None, Severe proliferative changes, in the left eye large white mass down. Diabetic proliferating retinopathy. Complicated cataract.

24 M Right eye 0.5. , left eye 0.1. None. Numerous hemorrhages of all sizes; numerous small exudates and swelled veins. Diabetic retinopathy.

63 F Right eye 0.1, left eye f.c. After extended adaptation, 10 cm. Complicated cataract. sees light coming from the Chronic uveitis. back of the right eye. No changes on the left side.

44 M Right eye 0.4, left eye 0.4. Sees faint dawn of light Optic atrophy. Keratitis dominantly on the right profunda. Lues. side.

20 F Right eye light projection, Sees faint light on the left eye 0.05. Both disks right side, weaker on the pale and atrophical Cranio- left side. pharyngeoma. Blindness after cranial operation.

10 31 M Right eye f.c. 1 m, left eye Sees light on the left-eye f.c. 4m. In the right eye side, light is coming from chronic postoperative uveitis; the front above the left eye, in the left eye a very large sometimes brighter, like ris- disciform macular degenera- ing sun, reddish border. No tion. Retinal detachment in changes on the right side. the right eye and myopic degeneration in the left eye.

126 . . ,

TABLE 1 (Continued)

N Age Sex Status Test Reaction

11 20 M Right eye 0.5, left eye f.c. Sees white blots located 4 m. Partial atrophy of straight ahead of both

both disks and nystagmus. eyes .

12 16 F Right eye f.c. 3 m, left eye Sees yellow light located

f.c. 1 m. Corneal degenera- above the eyes , sees the tion and chronic uveitis in light in both eyes but both eyes after transplanta- brighter on the left side.

tion .

13 18 M Right eye f.c. 1 m, left eye Sees yellowish light in f.c. 5 m. Severe nystagmus. both eyes, brighter in Congenital cataract. Chronic the right eye. uveitis

14 24 M Right eye f.c. 2 m, left eye Sees light in both eyes f.c. 2 m. Congenital retinal brighter on the right side, degeneration in both eyes upper part brighter, lower part reddish.

Poor Visual Acuity in One Eye

4 48 F Right eye 1.1, left eye f.c. Sees light on the right 2 m. Central chorioretinitis. side, reddish light, no changes perceived on the left side.

5 68 F Right eye 1.0, left eye no Sees light on the right light projection. Optic side coming from above atrophy, blindness after the right eye, no changes cranial operation. perceived on the left side,

127 . ^

their surroundings also sense light There is another trend also incoming from behind; that is 70 per- dicative of the critical position cent of those who saw at least some- the truly retinal characteristics thing from the normal direction re- have in connection with this test. acted to the light coming from be- Both unilateral cases and bilaterally

hind, whereas only 20 percent of those poor-sighted cases (Nos . 4-7 and totally blind reacted to the light 11-14 respectively) showed certain from behind. A drastic exception to lateral differences in their reac- this rule were retinopathic eyes, of tions to lightened background, while which only 17 percent reacted to the there were also differences between light coming from behind. This un- the eyes with respect to normally expected result may lead to a re- directed light. Between these asym- valuation of the relevant factors metries there seems to be a trend listed as causes of blindness. which implies that reactions to inside light stand in some proportion The syndrome is , in its most ex- to reactions to outside lights (dis-

treme form, called proliferative crimination of stimuli) . If we may retinopathy (Beetham, 1963; Caird & overlook some mixed cases and exag- Garrett, 1963; Safir & Rogers, 1970). gerate some slight differences, the A characteristic of this syndrome is trend seems to be the reverse in the the development of a pathogenetic qualitative subgroups of blindness tissue and a detachment of the retin- listed. If the defect is focused al layer from the sclera or from the just in the retina or "after" it, background of the eye (fundus oculi) there seems to be an expected rela- On the basis of the results there tion between the asymmetries so that seems reason to assume that espe- the more poor-sighted eye probably cially in connection with diabetic lacks the inside reaction or shows retinopathy the reduced optical per- damped reaction. On the other hand, ception or blindness is connected when the defect is located anterior with, if not primarily caused by, to the retina, as a rule, poor sight the reduced sensitivity of the re- correlates with brighter reaction, ceptors with regard to reflected light this being obviously due to the dim (from behind) which means literally tissue located between the lens and that the receptors have a reduced the retina (the dimness of the lens ability to receive that light. Ac- and vitreous humor) . cording to this assumption, the actual reducer of the receptor's sensi- This type of sounding naturally tivity is the pathogenetic tissue, a needs a replication which should cement- like nontransparent matter prove how far the handicap is actu- which intrudes between the pigment ally connected with the non-perema- layer and the sclera, blocking the bility of the tissue and not caused reflecting light from the receptors. solely by the displacing effect the To our knowledge, the question of growth of the tissue itself has. A how this malformation is related to replication focused on an analysis the disturbed metabolic functions of of the optical correlates (besides the diabetic is still open. Un- accuracy) associated with early signs touched, for the time being, has of retinopathy would also give hints also been the above-mentioned possi- with respect to the practicability bility that the effect of this tissue of the device. If the replication as a cause of blindness could actu- confirms the result obtained, its ally be based on its property as an theoretically most interesting im- obstacle--hindering the reflected plication is that a dim (non-perme- light from reaching the receptors. able) medium may impair visual accu- This displacement of the retina, caused racy even when it is located behind by the foreign tissue as such, has the retina. presumably hitherto been considered as a self-evident cause of blindness. This displacement may, however, be the necessary but not sufficient Color Blind Subjects condition of blindness. These more clinically loaded questions will, Diagnoses of color-blindness and however, be treated elsewhere (Miet- reactions to the light pipe test are tinen & Weckroth, in preparation). summarized in Table 2.

128 TABLE 2

Recordings of Reactions and Verbal Reports When Light was Directed Through the Fundus Oculi , and Index Showing Bilateral Asymmetry Between the Eyes in Ishihara Test

Light Test Number of Lateral Correct Asymmetry Sv±>ject Ishihara Diagnosis Reactions Verbal Report Index

1 Deuter anomaly Nothing 3

2 Deuter anomaly Nothing X*

3 Deuteranomaly 5 Sense vibration 9 as if eyes were opened, light color

4 Deuteranopic dichromat Nothing X

5 Strong deuteranomaly Nothing 4

6 Strong deuteranomaly Nothing 4

7 Protanopic dichromat 12 Dark shape on 5 lighted background

8 Deuteranomaly Nothing 4

9 Strong deuteranomaly 12 Yellowish flashes 1 in front above the eyes

10 Strong deuteranomaly 12 Round yellowish light on left side

11 Monochromat Vertigo, faint light behind the eyes

12 Color normal 12 Bright light more on right side, above the eyes

*Si±>jects marked with X could not be contacted for retest.

129 . ;

There was no reason to be suspect This time each eye was tested sepa- beforehand of the Ishihara test as a rately (subject covered the other criterion of color blindness. ^ Con- eye gently with his hand) fidence was based on the implicit assumption that the light test would be Ordinary coefficients of reliable to make distinctions in every ability were not computed because it case between protanopes and "others." soon transpired that the problem of Obviously, this was a false assump- reliability in a group of color-blind tion with respect to the critical si±)jects is distinctly different from class of color blindness. Control that in a normal group. Instead we measurement of color blindness car- restricted ourselves to recording for ried out after all subjects had been each individual a discordance index tested, revealed that the first sub- indicating the discrepancy between the ject who gave an entirely positive or eyes. Comparison of the m.ean index normal response was actually the only of the group with the corresponding protanope in the group. The control mean value of the control group dis- measurement also revealed that a sub- played that discrepancies between the ject showing no signs of colpr blind- eyes were manifold in the color-blind ness had mistakenly been included in group. The discordance index in the the group. observed group was 4.29 and in a

couple of control groups . 30 , N = 10 The attempt to analyze the reac- and 0.13, N = 40 . ^ The difference tions of color-blind subjects was between the means of the color-blind based on the assumption that explain- group and respective control groups able differences occur; if not within was highly significant (t = 5.1, the color-blind group, at least be- P < 0.01 and t = 14.7, P < 0.01 re- tween color blind subjects and normal spectively) . It is evident that the subjects. The trial was motivated reliability problem (in a sense of particularly by the likelihood that parallel-reliability which could be reactions may be lacking entirely, estimated on the basis of two such but that this could be explained by independent measurements) is entirely reference to the contracted spectral different from that of the normal sensitivity curve of the protanopes. group. Evidently the discrepancy so Thus, results obtained are paradoxi- common in this sample is not merely cal with regard to the expected re- a problem of decreased reliability. sults in several respects. First, It may be a central problem connected a remarkable proportion (70 percent with the structure of the deviancy it- of all subjects suffering from deuter- self, implying that quantitative dif- anomaly or deuteranopia) did not see ferences are not based on identical any change, that is did not perceive underlying dimensions in the anomaly any light. Second, there was a cate- group and the anopia group respective- gorial response set observable, as ly. From a strictly psychometric expected, but this ability to see point of view this implies that the all of the successive stimuli or none measurement of a color weakness or at all is annoyingly conspicuous also defect by means of such a test does in the groups of deuterans . In the not have the same meaning in the group of nine deuteran subjects color-blind group as in the normal there were two who reacted spon- group. This structural problem may taneously after each stimulus. reflect itself in connection with unilateral cases precisely in that the more deviant eye never fully develops to the degree genetically in- Actually, the one distinct extended, because of the interaction ception (No. 3) from the general rule between the eyes. The indefinite either to see or not to see forced us character of such cases is mentioned to inquire into the reliability of also by Graham and Hsia (1953, p. 282). the criterion measurement. On the basis of one well-known case of uni- Individual values of the dis- lateral color-blindness (Graham & crepancy index are listed in the last

Hsia, 1953; Graham & Hsia, 1954) , it column of Table 2. In harmony with seemed possible that the exceptional the expectations, the critical case intermediary case might be such a (No. 3) actually gets the largest unilateral case. Therefore a retest value. The conclusion seems to be was perfoirmed with the Ishihara test. warranted that the subject in question

130 is a unilateral case whose seemingly of anomalous trichromatism cannot be high reaction value in the light test predicted on the basis of the "sum can be explained by reference to the score" indicated in the fourth col- normal eye. umn. Actually the subjects included in the first two groups are, accord- This was an additional reason ing to Ishihara's conventional class- motivating the replication of the ex- ification, deuteranomalous trichromats periment. The replication group was or deuteranopic dichromats. Because treated differently in two respects the differentiation between protan from the beginning. First, the whole and deutan types is actually based group of top form pupils (N = 115) on only four charts with stimulus were tested bilaterally and the sam- number (plus a couple of charts ask- ple was picked on the basis of the ing the subject to trace a line) the results obtained separately for both whole determination of dichromacy or eyes. Second, the final score of the strong anomaly is probably not very test was arrived at by a somewhat un- reliable. But under these conditions conventional method in that every re- cases 1, 5, 7, 9, 12, and 14 are to sponse was registered separately for be considered as the most distinct each chart, and one point was given deuteranopics. If we take into ac- for each correct response. If the count the number of normal responses, correct response was identical with cases 1, 5, 9, 14, and 17 are the a response given as a rule by color- purest deuteranopic dichromats, while normal subject on that particular the rest are to be considered deuter- chart, it was designated normal, and anomalous trichromats of various de- correspondingly if it was identical grees of severity. Special attention with responses usually given by color- is paid later on to "unilateral" case. blind subjects it was designated as a No. 16, and to the rest of the middle color-blind response. Incorrect re- group. sponses were omitted. Again rather a considerable pro- On the basis of all the responses portion of color-blind subjects failed given, a particular aolor weakness totally (or almost totally) to per- index was computed for each subject ceive any change or any light whatso- by subtracting the color-blind re- ever directed through the background sponses from the normal responses. of the eye. The listed results con-

On the basis of this index the sam- verge with the prior results , includ- ple was divided into three subgroups: ing the annoying exception of a deutan color-blind, color-weak, and color- subject who actually was able to per- normal; one strongly normal case was ceive the light. In spite of this moved from the middle group to the exception, the difference is without last group. doubt significant between the means of the first group and two last groups The replication was expected to combined, or between the first or the answer to the probability that color- first two and the last alone. The blind subjects and expressly deuter- average number of correct reactions anopic subjects do not actually per- also increases systematically from ceive the light directed toward the first group to last. If a minimal background of the eye. This replica- number of correct reactions (0 or 1) tion could, at the same time, be taken had been used as a cut-off score we as a sort of cross-validation of the would have been in error if this test question to what extent could the had really been used as a criterion critical class of color-blind subjects of this type of color-blindness. be validly discriminated from the rest But the question of the validity of of the subjects by means of the ob- this procedure cannot be and was not served reactions in the light test. intended to be seriously treated here. Instead we may attribute decisive The results of the replication significance to the conceptual side are summarized in Table 3. The reader of the finding. This is, in our is asked not to mix the quantitative opinion, concealed in the possibility indication of color weakness with the of interpreting the lack of percep- classification of color blindness. tion as depending on insensibility As might be evident from the fore- to light and not merely to color. going discussion of the reliability We may then speak of light blindness problem, the conventional subtypes in the same sense as we did in

131 : .

TABLE 3

Original and Combined Scores in Ishihara Test, Index Showing Lateral Asymmetry Between the Eyes and Recordings of Reports When Light Was Directed Toward the Fundus Oculi

Ishihara Response Lateral Light Color- Sum Asymmetry Test Subject Normal Blind Score Index Reactions Verbal Report

Group I Color-blind

25 -23 12 Light from below, darker and brighter.

17 -10 No changes; claimed afterwards that he saw sm.all dots all the time.

5 1 23 -22 6 1 Once, very faint light.

6 5 17 -12 5 No changes, nothing.

7 ; '6 ^.. 20 -14 3 No changes, nothing.

9 25 -25 1 No changes, nothing.

12 ' 6 21 -15 6 No changes, nothing.

14 " -• 2 27 -25 3 No changes, nothing.

17 1 30 -29 3 1 Short flash on right side

Group II ' w ;..w .' Color--weak

4 16 10 -1- 6 7 3 Once, light from below.

13 16 11 + 5 5 3 Vertigo, flash of light.

16 20 14 -1- 6 11 7 Blots of yellowish light on left side.

13 10 + 3 14 11 Dots of white light, no direction reported.

Group III Color-normal

3 48 -1-48 11 White light from above.

8 48 -1-48 10 White light in the eyes and outside.

10 32 -1-26 11 Flashes of white light from below.

11 45 -1-45 12 Red flashes of light, direction indefinite.

19 41 -1-41 12 Red flashes of light in front.

132 . connection with, for example, the projected the image of light down- retinopathies in the first trial. wards , because he did not perceive This insensitivity may be transient the light until it reached the op- and it may depend on some external posite side of the eyeball after factor or other, but insofar as we having ineffectually penetrated the may accept the existence of the in- retina. By listing such underlying sensibility as such, it is a point factors we can hope that the continu- of view from which a aolor-blind ation of the study will- be able to subject can actually be considered solve them or take them into account. to be light-blind , not color-blind. Amonq the sources of error there is The external factor may be an accel- one which might already be treated a eration or inhibition of the develop- step further and that is the adapta- ment of some "opsin," or specific tion problem. From the point of view non-transparency or even color of of the theoretical argumentation ap- the background or other intervening plied here the adaptation problem is tissue. However, important in our essential. It may focus our atten- opinion is that insensible subjects tion because it may expose the possi- do not see shades of white or gray ble methodological power the unilateral in the presence of most probably red cases might have in this connection stimulus, but see nothing. (The pos- (and not only v/ith respect to the tulation that color-blind subjects adaptation limit) generally see shades of gray where color-normals see some color may be The question is v;hether insensi- erroneous altogether, but this is im- bility to light directed through the material for the present argument background of the eye is of a transi- (Weckroth, 1969, p. 182). ent nature or not. The attainment of the theoretical goal depends upon The approach is, of course, open whether the lack of perception of to criticism. Direction of light, light is due to a specific duration size and location of the patch on the of adaptation or not. From this retina, time and quality of adapta- point of view the unilateral cases tion, and duration of stimulus are are in a key position methodologically some of the most relevant character- because they can be considered as be- istics which determine actual per- ing their own control. In the second ception under normal conditions. Some sample there is one even clearer uni- of them, or those expressly listed, lateral case (Graham & Hsia, 1954; may be even more central when light Graham & Hsia, 1953, pp. 281-95). In is directed on the retina from an this case the eyes diverge from the entirely reversed direction. It had theoretical average value almost thus to be admitted that in addition equally as much, but in opposite di- to the loose frame of response, lack- rections. The subject has an almost ing an exact determination of the equally large "normal" score on the shape, brightness, hue and direction left side as he has "color-blind" on of light eventually perceived, there the right side. In the light test he was a group of factors possibly af- spontaneously reported sensing the fecting the outcome of the phenomenon light expressly and only on the left which were weakly controlled. Thus eye's side. The categorial tendency an incidental factor could account to "see or not to see" seems thus to for the main outcome itself and for be accentuated within one and the the exceptional presence of the series same individual, provided that the of correct reactions in particular. difference is great enough between the eyes. Therefore, the question Lack of fixation, for example, arises, if the time in the dark has could easily be referred to as a been enough for one eye , then why not source of error. Without rigid fixa- for the other? The possibility that tion there is no control of the direc- the response is absent on the right tion in which, or area where, the side because adaptation proceeds more light falls on the retina. Moreover, slowly in a color-blind eye does not even the closing of eyes and mouth as a matter of fact seem very prob- were not very rigidly controlled nor able. So much the less since we learn the position or movement of the test from Saugstadt (1959) quoting Uthoff tube. On the other hand, more sys- and Hecht, et al . , that adaptation to tematic inquiry into the verbal re- the dark of color-blind subjects is ports might have revealed, for exam- faster than the adaptation of color- ple, that the exceptional case normal subjects. The possibility

133 . that the color-blind eye has a pro- absence of pigment, as evidenced in longed adaptation due precisely to the studies of Rushton (1958) , Wald the unilaterality itself does not (1960) , and Weale (1966) . The re- seem very plausible either. If it sults obtained are not quite uniform is actually true, we are faced with and the loss-theory has also been an even more difficult question as criticized (Gregory, 1966, pp. 126- to the nature of such double adapta- 29). It is assumed that one factor tion. Worthy of consideration is the determining color weakness is the evidence presented by Rushton, show- lack of pigment in both deuteranopics ing ^° that the permeability of the and protanopics — at least in the lat- retina, in terms of absorption, va- ter, and Wald claims its absence in ries as a function of the level of the former as well. A variation of adaptation. So, with respect to this view is that some pathological where the response originates , the location of the pigment might explain adaptation phenomenon might be deci- blindness to both color and light sive. Inasmuch as we may assume that (Dartnall, 1960, p. 160). A further the observed effect is restricted possible explanation is associated and based on the immediate influence with observations of an especially the light has on the back of the general fall in sensitivity in color- retina, the problem may be left un- blind subjects (Hecht & Hsia, 1947; decided. Under these conditions it Wald, 1966, p. 1358). It would, of seems probable that the insensibility course, be simplest to explain that is not, at least in the bilateral blindness to light is a consequence cases, due to a prolonged adaptation of higher (than normal) absolute connected with the color-blindness. threshold values. Wald to some ex- It might be due to the possibility tent doubts any general rise in that the adaptation proceeds faster, threshold values , but arrives at but also in that case we may conclude results similarly indicating the that the insensibility (i.e. the ba- sensibility of deuteranopics as sis of the absence of reactions) is weaker, though the difference is not not a transient but a permanent char- significant. The fall in sensibility acteristic reflecting an insuffi- is observed precisely in connection ciency, malformation, or overdevelop- with dark adaptation. A third pos- ment of some fundamental component. sibly relevant structural factor here is the Stiles-Craford effect, the de- The unilaterally positive find- pendence of light effectivity upon ing also throws light upon direction- angle of incidence on the retina al and areal factors as sources of (Pirenne, 1967, pp. 47-51). The same error. Accordingly, it seems rather light is thus weaker in its effect unlikely that the existence or ab- upon the same retina, the further its sence of positive reactions is essen- angle of incidence diverges from the tially dependent on these dimensions optic axis. It is symptomatic that either. When a unilateral case reacts the angle gradient here is selective with one eye as color-normal peers do, in respect to color, and that the and reacts with the other eye as the loss of effectivity is greatest with majority of color-blind peers do, it red, diminishing gradually toward seems unreasonable to assume that the shorter wavelengths. If the Stiles- color-blind peers happened not to see Craford effect can be imagined to the light because they had their eyes extend to the other side of the retina in a different posture than the rest it might be thought to narrow (in the of the group. So all in all the uni- manner of a raised threshold) the lateral case strengthens the early im- conditions under which light reaches pression we had about the "all or the receptors and is sensed. The none" regularity the phenomenon seems angle gradient might thus constitute to obey a significant additional factor which, in combination with the above, would The impression was confirmed, of bring about an absence of reaction. course, by the conception of the cate- Its effect would not, on the other gorial nature of color blindness and hand, emerge in the normal eye pos- by the possibility, a priori, of re- sessed of both pigments or normal ferring to certain permanent struc- threshold. An orientating approach tural factors which might explain the is of course insufficient for a more absence of sensitivity to light. A precise definition of the channels of probable explanation is the complete influence. The theoretical significance

134 . . particularly of the latter factor unilateral. In the remaining series could be noted in regard to future of items, all belong to a minority research. It is associated with the again in showing occasionally normal fact that what is involved is a gradi- responses ent of angle, which in addition may be critical, especially with regard Thus the middle group in Table 3 to reflected light. It should be formed on the ground of dominantly noted also that it is a question of normal response sum in the Ishihara, gradient of an angle which is dis- is not intermediary in its color tinctively different in connection blindness in the unidimensional quan- with different colors or anatomically titative sense. Rather the group different receptors (rods and cones, would be located at the apex of a Pirenne, 1967, p. 49) triangle having as its base angles color-normal and color-blind, respec- The Ishihara test is not nor- tively. In actual fact, the middle mally carried out for each eye sepa- group forms a miniature triangle it- rately. In this connection we feel self, with the clearcut unilateral a number of comments are necessary. case as its apex. In the psychometric The critical unilateral case — or view it could be said that in place rather the concept of unilaterality-- of internal consistency there is in- is critical for the criterion. What ternal conflict. The reliability in the is the message of unilaterality , how problem proves exceptional contingent is such a case, and how middle group as well. The limited reliable are any conclusions drawn size of the trial for any actual on its basis? The center of empha- statistical conclusions must of course sis is the middle group, in the fol- be borne in mind, but its organiza- lowing analysis. tion suggests that the color blindness classified here is as a syndrome Sub-series may be distinguished at least two-dimensional, and that among the Ishihara tables according the poor predictability is consistent. to whether the item provokes only In other words, what prevails between color-normal, only color-blind, or the eyes correlates with what pre- both, to perceive the intended fig- vails within them. Be it noted that ure (Ishihara, 1960) . Predictability the conclusions concern the main (from eye to eye or from item to group of deutan-type of color blind- item) may be scrutinized starting ness . from the last mentioned series, which due to its discriminative capacity On this basis it may be envis- has been used as a screening set aged that there may also be, as Will- (Wald, 1966, p. 1349). On this ba- mer has proposed (Wald, 1966, p. 1349) sis the four cases in the middle two types of deutanopia. It is also group (Table 3) would be diagnosed justified to accept the comment of completely differently. Case 4 Ripps and Weale that the loss system would be diagnosed as completely is probably insufficient in itself as normal; Case 13 as totally deuteran explanation of color blindness anopic; Case 16 left eye normal, (Ripps & Weale, 1969; Gregory, 1966, right eye deuteranopic; and Case 18 p. 129) , and the observation of the genetics as mixed (perhaps tritanopic) . So, difficulties "that arise for it is observed that the members of when color anomalies are treated as this (statistically very small) group states intermediate between normality in fact differ strikingly from one and dichromacies (Ripps & Weale, 196 9, another but represent fairly pure p. 210)." The most significant in- subgroups. On the basis of these terpretation from the theoretical differences it cannot be predicted point of view is associated precisely that the members of the group respond with the poor predictability of re- in completely identical manner to the sponses; the difficulty of predict- the other following series , which as a rule pro- ing answers from one eye to vokes only the color-blind to perceive and from one series to the next. The a figure. They all prove normal. In group showing discrepancies between the next series, again, all of them the eyes also proved to be the ex- show occasional unipolar normal re- treme group in "internal discrepancies" sponses, usually two and, notably, within the eyes, because unipolar plus irregularly on one side or the other and minus answers are extremely rare except Case 16 who is most markedly in both the color-blind group proper

135 .

and in the normal group . One has the Normal Subjects impression that at this level color blindness is expressly an interaction There is no need for a table problem. The modal is seeking a state listing the observations of "normal" of peaceful coexistence between the subjects. All of them reacted cor- eyes. The purest unilateral case rectly after each stimulus flash, seems to have achieved the optimal as did the control cases in Table 3. situation, but here too not complete- We conclude this report of an appar- ly, as corresponding cases have previ- ently unknown phenomenon by a descrip- ously shown (Graham & Hsia, 195 3, tion of what was seen by our normal

1954) . From this standpoint one subjects. might claim that discrepancy between the eyes in fact reflects an internal Actually, half the cases claimed interaction problem of the eyes (per- the light to be reddish or pale where- haps specifically of one eye). In as the other half saw it as more or other words the discrepancy between less white. The shape seen resembled the eyes emerges, because color the form of a northern lights and blindness is (at least in respect of shadows , and may be due to the curved its component factors) an interaction bone below the eyes and the possible problem of fields or channels. fact that the light spread otherwise freely on the retina. The image is There is normally interaction usually projected upward in front of between the eyes. It is observed, the eyes. The hue or brightness seems for example, in the shades of "para- not to alter due to the length of ex- site colors" or color aftereffects posure or dark-adaptation; nor does produced by achromatic stimuli. The the flash seem to have substantial shades can be changed by changing the contractive effect on the pupil. In "inter-eye light intensity ." ^ * It the training period, some subjects may be assumed that the development revealed that some stimuli can be of a unilateral syndrome is based missed in the beginning due to the essentially upon the same interaction difficulty of distinguishing the ac- mechanism as instantaneous interac- tual stimulus from different possible tion. On this basis poor internal aftereffects. Since the correct predictability would be easy to ex- stimulus was detected the subject plain by reference to deficient or was able to react correctly each time. inflexible interaction between re- The same categorical trend was also ceptors and fundus, and between eyes. observed in connection with blind and The closing of the other eye, it may color-blind subjects. be assumed, fixates a certain balance level on which the patches of the The difference in hue of the pseudochromatic charts are distin- flash could be attributed to the po- guished or not, particularly as to sition of the eye (refer to the Stiles- brightness and not only to hue. Craford effect) . It could be mentioned, however, more curiously, a If the variegated intermediary phosphene caused by an electrical cases may be regarded as correlates current has been identified in two of certain phylogenetic phases of distinct hues, either yellow-red or development, sequel studies must pay blue-white depending on whether the particular attention to this area. current passes outward or inward For the moment it must suffice to through the retina (Duke-Elder, 1968, note that the number of intermediary p. 466) reactions in the light test seem to be associated with poor inter- or intra-eye predictability, but also CONCLUSION that the accumulation might be a co- incidence. It might also be due to An attempt was made to show that a characterological factor connected man normally sees light directed with the uncertainty of the observa- through the fundus of the eye. At tion. All in all, however, the cri- the same time, some conditions have terion characteristic has proved to been brought out in which this light be an unexpectedly complicated mat- perception is obstructed. According- ter. ly, one contributory factor in

136 . , blindness may be an obstacle or an determined by a combination of light opaque medium blocking the light, itself and its reflection from the even in cases where it is behind the fundus . retina. In examinations of color- blind subjects it was observed that a In the introduction the bipolarity certain type of deuteranopic dichrom- of certain sub-functions was referred asia might be a result of the insen- to as a general principle. The eye sibility of one receptor type, which seems to deal with light energy as the could be called light-blindness. It hand deals with an object by thumb and was not determined whether or not the other fingers operating in opposite dysfunction was caused by the light directions. Perception takes place coming from behind. The normal in- as a result of a clearly bipolar ef- dividual perceives light introduced fect in which the apparently centrally from behind either without or after determined efferent control plays an a very short period of adaptation essential part. Additional external The perception is based on irrita- symptoms of the do\±>le strata of the tion of the back of the retina, but structures include, above all, the might possibly be due to light scat- leap observed in connection with the tered throughout the eye. Purkinje phenomenon, the obviously dual dark-adaptation curve. At the same time, and independently, a device was developed, mainly The activity or dominance of the through experiments with cats, which strata is clearly a correlate of the made it possible to project light amount of light. Although the retina directly to the back of the retina here is assumed to be, for the most

(del Campo, 1971) . The idea was to part, permeable, the receptor and evolve a technique to assist a cer- pigment layers form a natural limit, tain group of blind subjects in whom which normally are located so that the macula lutea was intact in spite perception is based mainly on light of pathological developments in the coming from the front and partially preretinal regions. The principle on light coming from both front and of direction was the same, but re- back, A further correlate of the quired surgery. Nor can it be claimed amount of light is the double struc- that the capacity of the retina can ture of the main classes of receptors be preserved for long if it is di- rods and cones . For example , change rectly affected by light. It is evi- in permeability as a function of dent that the light tube herein de- dark-adaptation, the distinctively scribed could be developed into an different processes of the opsins as inexpensive appliance which a si±)ject well as the distinctive location of preretinally blind but retaining the the Stiles-Craford effect might be sight capacity of the retina, might regarded as factors promoting exhaus- use to facilitate limited vision. tive treatment of light energy on the basis of this kind of balance struc- The fact that man is capable of ture; the last-mentioned especially perceiving light directed through the in connection with reflected light. back of the eye does not of course justify the conclusion that light As internal factors supporting coming from such a direction is of the assumption of the central sig- functional significance in normal nificance of reflected light or pos- vision. If it is involved it must be sibly connected with it, mention noted that the light is by nature re- might be made first on the angle- flective. In principle the receptive gradient of the receptors. One argu- qualities of the eye may be different ment here would be that the angle- from what we know. Any conclusion as gradient seems to be associated ex- to the significance and effective plicitly with the function of the mechanism of light reflected from be- receptor type (red) , and which may hind the eye must be drawn with cau- be assumed to participate more action. However, one is forced to take tively in perception of light energy some sort of stand by the fact that coming from two directions and hence the phenomenon emerged on the basis be relatively greater. Furthermore, of theoretical reasoning. It seemed inasfar as we may assume the angle- that a nvmiber of light-perception and gradient of the receptors to function visual phenomena might be more simply in the manner of a polarized filter, explained if we assumed that in part it would function precisely as in the of our perception the outcome is case of reflected light. This would

137 change it into a real direction- is demonstrably capable of register- gradient. True, it must be noted ing only variations in quantitative that the restriction of the Stiles- frequency and amplitude. The essen- Craford effect to the cones has been tial question from the viewpoint of observed only in the case of light the evolution of the servosystem is coming from the normal direction. how the subsystem could have evolved The comment of Pirenne that the ef- from a more simple monogenetic state fect of the rods depends upon light- characterized by one receptor (di- adaptation may be of significance in vided in gaussian fashion as to its this context (Pirenne, 1967, p. 49). sensitivity) . It is sought here to A further internal structural argu- see the combined influence of the ment would be the evidence that the permeability of the retina and the role of the fundus oculi is essential reflectance of the fundus as factors in light scattered in the eye (Vos, forming a basis for both the develop-

1963) . In connection with the ob- ment of the servomechanism and the servation once reported by Dartnell balance of the bipolar components. that the opal media gather reflected light, the light reflected by the As part of the servosystem the fundus might also perform a special theoretical "primary receptor" must informative function, because light be assumed to possess a certain uni- is collected and is perhaps also di- versality. In cooperation with the rected selectively as a function of brain it must be a kind of Turing wavelength (Dartnall, 19 60, pp. 156- universal machine capable of solving

58) . all problems of reaction to environment arising and embracing all the The above connection of the variations subsisting in the later angle-gradient with the rods inde- differentiated developments of it. pendent of the level of adaptation The demonstrated power of the eye and to the cones dependent on it, may to receive light authentically from be precisely an indication of the the fundus , in our opinion allows the distinctive significance of reflect- possibility of seeing in a simpler ance with regard to the amount of fashion than hitherto, what is uni- light. Let us suppose that effec- versal in the receptor. The task of tivity changes, in the case of the the optic receptor is of course to cones, as a function of direction observe the environment. Its turn- more marJcedly because the perception ing may be understood simply as an was essentially affected by light re- early variation on this theme. Uni- flected from the background. From versal in the operation of the re- this point of view it may be extreme- ceptors is interaction. In turning ly relevant that the effectivity of to its background the receptor is red falls most sharply as obliqueness realizing its earliest mode. This increases. A number of other phe- turning may be seen to form the basis nomena could be listed which place for the tendency to interaction gen- red in a special position, phenomena erally characterizing the receptor which for their part verify the role function and manifested between the of the fundus, bearing in mind that fields of the retina, the types of the fundus is homogeneously red. Such receptor and, on the macroscale, also indicators include the adaptation between the eyes. In the treatment curve, perimetric color fields, the of information, the turning is, in sensitivity curve of protanopics, the principle, the hinge which opens the early learning of red and so on. In development toward a "comparative" outward physical reality, one might resonant "perception" in addition to note that red is perceived at both the rigid "detection-type" perception. ends of the spectrum, to say nothing Reversal away from the source of of the biological significance of this light is only ostensibly detrimental. color. It evidently forms a sufficient but also necessary condition for resonant Central to the theoretical argu- perception of a higher level than ri- ment adopted here is the consideration gid instinctual perception. It is, of the eye as a differentiated sub- as an outcome of intentional activity, system of the brain. The chief ques- in principle analogous to the "creation in instantaneous perception is tive perception" with which the bat hew the eye can distinguish and recog- and the dolphin sense the minutest nize colors when the neural process differences in energy in their

138 . — environment by producing sounds them- Gothlin, and for red-green that of selves which are then received as Hering as an explanation of the echoes Bezold-Brucke phenomenon with respect to its yellow component. Character- In his treatment of the kyber- istic of the split in spectral sensi- netic aspects of perception Klix has tivity, distribution would be a bal- recently stressed the central im- ance between the amount of light portance of apprehension of "propor- meeting the receptor from either tion" and mittelwertbildung . In this direction. Thus, hue would also be context he notes how inadequate the dependent upon the intensity of light. mere "measurement" of the amount of It may be supposed that from the radiational energy of an object would brightest point of light energy on- be (Klix, 1971, p. 243). In concord- wards, an ever greater proportion of ance it might be claimed that the the reaction of the eye would consti- change (or loss) involved in the "re- tute protective functions. Actually, versal" of the receptors, is concen- it is evident that the protective and trated specifically upon this very orientative response of the eye to possibility of measuring the amount amount of light could not subsist of energy or differences in it as without some basic level— the fundus such. At the same time it has of with which the varying amount of light course meant that the value of formal is comparable. It is undetermined stimuli (form) has in a sense given how far the photochemical process of place to that of relative stimuli a given pigment is essentially regu-

(hue) . On top of this one finds one- lative with respect to reflecting self apparently obliged to accept light. such inaccuracies as the Mach bands and the apparent colors of achromatic A structural detail of signifi- stimuli. These are, however, phe- cance from the achromatic standpoint, nomena brought out mainly in extreme- in principle involved in the split of ly precise laboratory conditions. spectral sensitivity distribution, Some of these symptoms may indeed comprises the horizontal neural con- derive from the fact that the fundus nections. These of course maintain is in reality multistratif ied because contact between the different groups of the pigments and thus endowed with of receptors. It is quite clear that a more complex functional role (Pir- their basic functions are of a kind enne , 1967, p. 181). The ultimate which depend upon interaction. There underlying aspect in all of this is is as yet no general agreement as to the absorption--reflection ratio of how they participate in various the original fundus , the constant spheres of activity. It has been quality which in a long process of suggested, for example, that the reg- genetic selection has been able to ulation of average perception, ex- split the (double-peaked) receptor plicitly in connection with achrom- distribution. atic stimuli, is based upon the function of the horizontal cells (Klix,

This imagined evolution is clear- 1971, p. 242) . It has also been pro- ly in keeping with the known spectral posed that antagonistic part proces- sensitivity distribution of the re- ses in the receptive fields might be ceptors. If, in addition, we take manifested at various levels of trans- into account the red emission spec- versal contact (Glezer, 1971, p. 43). trum of the fundus, the phylogenetic The essential point, however, is that (or ontogenetic) coding of color the interaction required by the sub- could scarcely derive from factors areas can, in general, only take of any simpler structure. Let us place against some background. The suppose that one of the receptor horizontal neural connections, which units — including the pigment--could possibly represent an evolutionary receive light from either direction. counterpart to the background of the It could then distinguish and iden- eye, and actually in respect to their tify light of the same color as the location do constitute a structural fundus solely on the basis of inten- counterpart on the other side of the sity because the loss of energy in retina, are also a functional counter- feedback of the same color would be part of the fundus. In this envis- minimal--red. The maximal loss, aged development, then, the status again, would be with the complement of the fundus is paradoxical. It af- of red--green. This would follow fords the only condition of function the pattern of balance proposed by for the cooperation of the receptors.

139 differentiated because of it. The is found to vary as a function of duality of the horizontal neural con- the effect of light (Rushton, 1952, tacts may thus be considered a kind p. 47) . In undertaking the experi- of redundancy which, for its part, ment we were aware of the ambiva- indicates the significance of the lence. It was not certain whether connection. With reference to the the absence of reaction was due analysis by Glezer (1971, p. 41), it to the fact that the light came should be emphasized here that co- from behind or the fact that it was operation is likewise not envisaged red. In the case of deuteranopics as based on scattered light, although the problem of sequel studies is this may be taken into account as one also a dual one. The reaction may possibility. be lacking because the light is not distinguished from the base light In an assessment of the so-called experienced by the closed eye, or degeneration or disintegration methods because the light does not penetrate Dekking once remarked, "even if it or affect the pigment of a certain does not give you a clear insight in- receptor unit. From the fact that to the structure you are studying, at the visibility curve of deutan sub- least it gives you a clear insight jects is characteristically higher into the deficiencies of your theory." than that of protan si±ijects, it The comment seems appropriate to our might be assumed that returning concluding discussion of an open ques- light exerts an inhibiting effect tion associated with the validity of on the reaction of the receptors. the evidence of color-blind subjects. It is also known that the spectrum On the basis of perception of light of a protanopic is contracted at coming from behind we feel it should the long wave end. Here one could be possible to create a coherent pic- use the term light-blindness in ture of the modal development of vi- the sense intended in this report. sion and the present situation such Let us suppose that the deutan re- as would agree with the known neuro- ceptors proved contracted in the physiological facts. In developing same way but with respect to re- the picture, stress has been laid turning light. The method applied especially on the position and loca- here does not allow of any conclu- tion of the elements as of prime sig- sion as to the precise role of nificance in interaction. It is red in the absence of reaction. claimed that the universality of the We may thus assume that the blind- centrally regulated peripheral ele- ness would concern at least the ments lies also in their positioning complement colors of the deutan with regard to each other. But to receptor. On this basis it would what extent is it adequate to regard be more natural to expect lack of the absence of reaction to light in light-reaction on account of the deuteranopics as an indication of the failure of returning light to development imagined, especially since penetrate or stimulate the deutan the suiagroup, which in actual fact receptors. (There is scarcely lacked the reaction, was contrary to more than a theoretical possi- expectations? bility that the blockage of reaction would derive from differ- The subgroups of color blindness ences in the pigment of the skin are clearly mutually divergent as to of the fundus.) A concrete idea their symptoms . Its favor as a veri- of how black red can be may be fication of theories derives precise- gained by examining a red surface ly from this. In theory it has been through a narrow band green (or justified to expect light reaction to blue) filter. be absent in color-blind subjects in general. Certain differences were to be expected in this area, for example, because the permeability of the retina

140 . . . . "

NOTES

There is a marked "jump" in The duration of the stimulus flash was not controlled nor adaptation curves , Purkinje "shift," "neutral points" ac- was the reaction time which was centuated in Bezold-Brucke accepted as a "prompt answer. phenomenon, all indicating The experimenter was asked to shifts of the reaction from push the switch equally long and one substructure to the other, it was found that s\±)ject's reac- or loss of the other, or that tion came in one or two seconds both are responsible for that or not at all. In this type of reaction. proof, more accurate control of the input and output were not Monogenetic is the simplest form considered necessary, but it is of theoretical assumption con- realized that the duration of cerning the start and develop- stimulus might be an essential ment of the retinal receptor criterion if the number of re- units from one single type of ceptor types responsible for the unit (receptor + pigment + elec- reaction is more than one. In trochemical properties + neural continuation of the study these connections) having originally questions will be treated furth- a certain sensitivity distribu- er. tion. The other possibility is naturally the polygenetic as- The majority of the blind sub- siamption, postulating that from jects were tested at the Tyyskyla the beginning there have been Rehabilitation Center for the two or more different receptor Blind during their participa- units or mechanisms tion in a rehabilitation program. The rest of the subjects Polyak (19 41) counts cornea and were tested at The Eye Clinic, retina as the transparent media University of Tampere, as a part of the eye, stating that the of a diagnostic work. first is entirely, the latter, in greater part of its thickness, This trend, if real, could natu- transparent. But this statement rally be utilized as a differen- is found in the introductory gen- tial diagnostic tool, together eral outline of The Retina and with the self-evident find that the property is not treated in if the eye does not see light the proper context. Nor is the from either of the directions, transparency or permeability of the defect is most probably neu- the retina or the reflectance ral, that is, a defect caused by characteristics of the background atrophy or surgical ablation. of the eye (Duke-Elder, 1968) These possibilities will be Rushton and Weale (Duke-Elder, treated in another paper. 1968, Vol. IV, p. 477) form an exception in this connection The test for color-blindness The method concerned is essen- (Ishihara, 1960) is given in tially based on transmission and Finland in some elementary absorption properties of the schools for vocational guidance retina, and will, therefore, be purposes treated later. 9 . Over one hundred subjects were It can be mentioned in this con- later gathered for a replica- nection that the third author tion of the results. Both eyes who served as a first subject were tested separately. The (outside the carrier of this pe- color-normal subjects formed culiar idea) was very soon able the control group. to perceive the light flash even in normal daylight illumination.

141 , . , ,

10. According to Rushton and Weale connection with achromatic figures (1968) a considerabl e amount of and backgrounds. An example of light coining through the lens such a color in connection with

passes the retina tw ice , the stationary figure-background amount being natural ly dependent conditions is given by Weckroth on the composition o f the light, (1969, p. 182). An example of but also on the leve 1 of adapta- the alteration of the hue of an tion. See Note 3 The method afterimage as a function of light is described in prin ciple by intensity see Weckroth (1964, pp.

Rushton (1958, p. 73 ). 67-74) . The interaction between the eyes has been overlooked as 11. We wish to replace the term sub- a possible general source of error jective color with the term in experiments where one eye is parasite color to pertain to used as a control of the other. the experience of colors in

REFERENCES

Beetham, W. P. Visual prognosis of of the retina. In G. V. Gersuni

proliferating diabetic retino- (Ed.) , Sensory processes at the pathy. British Journal of Oph- neuronal and behavioral levels.

thalmology , 1963, 17, 611-14. New York, London: Academic Press, 1971, pp. 19-46. Caird, F. I., & Garrett, C. J. Prognosis for vision in diabetic Graham, C. H., & Hsia, Y. Luminosity

retinopathy. Diabetes , 1963, curves for normal and dichromatic 12, 389. subjects including a case of unilateral color blindness. Science del Campo, A. Iconophoresis : A new 1954, 120, 780. technique for obtaining macular vision in some types of blindness. Graham, C. H., & Hsia, Y. Some visual In T. D. Sterling, E. A. Bering, functions of a unilaterally di- Jr., S. V. Pollack, & H. G. Vaughan, chromatic subject. In National

Jr. (Eds.), Visual -prosthesis : The Phys . Lab. (Ed.), Visual problems

interdisciplinary dialogue . New of colour. London: National Phys. York, London: Academic Press, 19 71, Lab., 1953, pp. 281-95. pp. 129-133. Gregory, R. L. Oga ooh hjdrna. Dartnall, H. J. A. Visual pigments Stockholm: Aldusuniversitetet, of colour vision. In Y. Galifret 1966. (Ed.) , Mechanisms of colour dis-

crimination. London: Pergamon Hartline , H. K. The receptive fields Press, 1960, pp. 147-61. of optic nerve fibers. American Journal of Physiology , 1940, 130, Duke-Elder, S. Systems of ophthalmol- 690-99.

ogy . London: Kimpton , 196 8. Hecht, S., & Hsia, Y. Colorblind Glezer, V. D., Bertulis, A. V., vision. I. Luminosity losses in

Ivanov, V. A., Kostelyanets , N. B. the spectrum for dichromats & Bodvigin, N. F. Functional or- Journal of General Physiology ganization of the receptive fields 1947, 31, 141-52.

142 ) . ,

Hubel, D. H., & Wiesel, T. N. Recep- and normal. National Phys. Lab. tive fields, binocular interaction (Ed.) Visual problems of colour. and functional architecture in the London: National Phys. Lab., 195 8, cat's visual cortex. Journal of pp. 73-105.

Physiology , 1962, 160, 106-54. Rushton, W. A. H. Electric records Ishihara, S. Test for colour-blind- from the vertebrate optic nerve. ness. Japan: Kanehara Shuppan Journal of Vhysiologij , 1952, 117, Co., 1960. 47. {Brit. Med. Bull., 1953, 9, 68-74.) Klix, Fr. Information und Verhalten Berlin: VEB Deutscher Verlag der Safir, A., & Rogers, S. H. The re- Wissenschaften, 19 71. lationship between duration of diabetes and presense of retino- Kuffler, S. W. Discharge patterns pathy. American Journal of Oph- and functional organization of thalmology , 1970, 69, 387-92. mammalian retina. Journal of Neurophysiology , 1953, 16, 37-68. Saugstad, P. A. The duplicity theory.

Advances in Ophthalmology , 1959, Miettinen, P., & Weckroth, J. Early 9, 1-51. detection and measurement of retinopathia changes in connection Vos , J. J. Contribution of the fun-

with diabetes mellitus . (in dus oculi to entoptic scatter. preparation. Journal of Opt. Soo. Am. , 1963, 53, 1449. Pirenne, M. H. Vision and the eye. London: Chapman and Hall, 19 67. Wald, G. Defective color vision and its inheritance. Proceedings of Polyak, S. The retina. Chicago, the National Academy of Science 111.: The University of Chicago 1966, 55, 1347-63. Press, 1941. Wasserman, G. Invertebrate color vi- Pribram, K. Languages of the brain. sion and the tuned- receptor para- Englewood Cliffs, N.J.: Prentice- digm. Science, 1973, 180(4083), Hall, 1971. 268-75.

Ratliff, F. Mach bands, quantitative Weale, R. A. The perception of col- studies on neural network in the our. Discovery Journal of Science, retina. San Francisco, Calif.: 1966, 27, 31. Holden-Day, Inc., 1965. Weckroth, J. Black and white colors.

Ripps , H. , & Weale, R. A. Color vi- In J. Jarvinen (Ed.), Contemporary sion. In P. H. Mussen, & M. R. research in psychology of percep- Rosenzweig (Eds.), Annual Review tion. Porvoo: W.S.O.Y., 1969,

of Psychology . Palo Alto, Calif.: 175-82. Annual Reviews, 1969, 20, 193-216. Weckroth, J. Studies in colour

Rushton, W. A. H. The cone pigment after-images . Helsinki: Suom. of the human fovea in colour blind Tiedeakat, 1964.

143

CURRENT RESEARCH NOTES

BLIND MOBILITY CURRENT RESEARCH PROGRAM

John Armstrong*

The research program is a wide- A number of new mobility aids, ranging investigation into the prob- including the Binaural Sensory Aid, lems experienced by blind people dur- the Sonar Cane, and the Swedish Laser ing independent mobility. Cane have already been evaluated using the assessment procedure. The Each project within the main relative merits of the individual program is aimed at gathering basic aids, with respect to safety, effi- information about the nature of spe- ciency, and reduction of psychologi- cific mobility problems and seeking cal stress have now been catalogued practical solutions to those prob- but none of them so far evaluated lems. In some cases, the results of show an overall superiority on all research have already been communi- of the performance measures. cated to mobility specialists and welfare workers so that changes in The research group contributes the provisions of mobility services to a sub-committee of the National throughout the country have already Academy of Sciences and much of the started to take place assessment procedure used here is shared by various research groups in Parallel with these practical the United States. developments, the research group continues to pursue more basic lines of investigation which are intended to LONG-TERM EVALUATION form the foundations for future de- OF MOBILITY AIDS velopment. The improvement in mobility performance which follows training in THE OBJECTIVE EVALUATION the use of a particular mobility aid OF MOBILITY PERFORMANCE may be transitory and some of the problems resulting from the use of The central problem in assess- that aid might not emerge until after ing the efficacy of mobility aids or long periods of continuous use. At mobility training programs has been the same time, mobility performance the lack of objective measures of mo- with certain aids might well improve bility performance. An objective as more experience with the device in procedure, which assesses mobility question is gained. performance in terms of safety, efficiency and level of psychological Although the evaluation proced- stress, has been worked out. ure provides information on improvement in performance at the time of training, it does not predict the effects of periods of free use with the aid. It is accepted that the advantages of a particular device to its *Director, Blind Mobility user might not necessarily be in Research Unit, University terms of mobility improvement but of Nottingham, England. could be reflected as a beneficial

145 change in life style. Accordingly, the objective measures related to measures of blind individuals' self- orientation. No comparable data is esteem (see "Measurement of Adjust- yet available for post-aural aids ment to Blindness") have been made or for deaf/blind subjects. prior to training with mobility aids and taken again at intervals during a three year period. MOBILITY MAPS

In addition to the detailed in- HEART RATE AS A formation about the immediate environ- MEASURE OF ment provided by conventional mobil- PSYCHOLOGICAL STRESS ity aids, the blind person also needs to know about the organization of a A series of experiments has been large segment of his surrounding en- carried out to examine a possible re- vironment so that appropriate progress lationship between instantaneous heart towards required destinations can be rate and specific environmental maintained. events. It was hoped that, if such a relationship could be established, in- Nearly 500 copies of our publica- stantaneous heart rate might be used tion which outlines work on the stand- as an objective measure of the effec- ardization of maps done here have been tiveness of individual aids in provid- distributed to mobility officers, ing appropriate information about com- teachers , and social workers and a plex and stressful mobility situations. follow-up study involving ten mobil- However, it has been found that it ity officers is being carried out in would be exceptionally difficult to order to evaluate its success in pro- control for work output over very viding the necessary information to short periods of time, therefore it enable complete amateurs to make mo- is concluded that instantaneous heart bility maps. rate is not a practically useful measure of psychological stress in mo- As a further step towards stand- bility situations. ardization, we will soon be marketing a kit of point and line symbols which will enable unskilled map makers to

' construct high quality master maps • THE MOBILITY OF , which THE DEAF/BLIND :. from thermoplastic copies can ^ be taken. Because of their inability to make effective use of sound sources Experimental maps of a large for orientation purposes , the deaf/ shopping complex and a complicated blind cannot normally be given ad- pedestrian subway system have been vanced mobility training. A survey produced and evaluated. Results show has been carried out among mobility that the maps, if used in conjunction specialists and many of the problems with additional information (specially encountered in attempting mobility provided new landmarks or compass instruction for the deaf/blind have bearings) can enable independent mo- been isolated. bility in areas which would normally be avoided by blind pedestrians. A series of experiments have been carried out to establish the accuracy of sound localization that different TRANSMISSION OF hearing aid systems would allow. The MOBILITY INFORMATION localization ability of normal hear- THROUGH THE SKIN ing subjects was measured under both monaural and binaural conditions to We have continued to monitor the provide a baseline against which per- progress of the research group at the formance with different hearing aid Smith-Kettlewell Institute of Visual configurations might be compared. The Sciences in San Francisco where an effect of head movement on accuracy attempt is being made to use patterned of localization was also examined. tactual stimulation for the purpose of transmitting environmental informa- It is concluded that the use of tion by non-visual means. We have two chest-mounted aids leads to an also kept in touch with research on overall raising of mobility perform- the Optacon reading aid developed by ance and a particular improvement in a team at the Stanford Research

146 .

Institute. Our own experiments have currently available for the visually determined the two-point thresholds handicapped so that the efficacy of using the same type of stimulation these programs might be established. used by the two devices (vibrating stimuli at 60 and 250 Hz) . Results show conclusively that differences in PROBLEMS OF both two-point thresholds of the two LOW-VISION BLIND surfaces alone will account for the comparative success of the Optacon Attempts are being made to es- This suggests that the back, or any tablish a working relationship with other part of the torso, would not one of the eye hospitals in our local be a suitable site for the tactual area with the aim of evolving a com- display of a visual prosthesis in- bined clinical/performance assessment tended for the transmission of de- procedure for the low-vision blind. tailed information. Information is also being collated on the various low-vision training programs which have been set up in other DEVELOPMENT OF parts of the world. AN ULTRASONIC DETECTOR OBSTACLE THE DEVELOPMENT OF A small ultrasonic obstacle de- AN INEXPENSIVE VACUUM tector, which signals target range FORMING MACHINE by means of a very simple auditory display, has now been developed. We Although our publication on the anticipate that the obstacle detector design and production of maps for the is likely to be of considerable use visually handicapped has done much to as a secondary aid for long cane users encourage amateur map makers , the lack and guide dog owners. It should be of availability of vacuum forming fa- especially valuable as a means of lo- cilities, required for the production cating landmarks positioned some dis- of plastic tactual maps, tends to tance from the main travel path. The slow down their progress. partially sighted and low-vision blind, many of whom are unable to Because commercially available discriminate solid obstacles from vacuum forming machines are usually shadows under adverse lighting con- very expensive (£150-1,300) , we are the possibility of making ditions , might also rely on the de- examining vice to establish a clear path. the designs for a "do-it-yourself" machine available to interested par- ties. An effective prototype machine, MEASUREMENT OF which should cost less than £50 to ADJUSTMENT TO build, has been produced in our work- BLINDNESS shops. Guide dog owners commonly claim that guide dog training brings about THE PROVISION OF beneficial changes in life style and IN-SERVICE TRAINING a feeling of improved status within the community. Periodically we organize and run in-service training sessions for mo- We have attempted to confirm bility officers. In the past, courses these claims by monitoring the change on the use and training of the sonic in self-esteem of fifty blind people aid have been provided and in the near resulting from guide dog training and future a series of courses on the de- subsequent independent mobility with sign and production of maps , and on the dog. how to teach their use, are being organized. Although this study has been concerned with the effects of guide dog At the same time, individual mem- training, the measure of self-esteem bers of the research group contribute is independent of the form of train- to the various technical training pro- ing or rehabilitation which the sub- grams offered by the North Regional ject undergoes. We are now applying Association for the Blind and the the same measurement techniques to Southern Regional Association for the the social rehabilitation programs Blind.

147 . ,

PUBLICATIONS Gazely, D. J. Communication and mobility training with hearing im- Because our current research paired clients. New Beacon, 1973, contract has been running for only LVII, 160-75. eighteen months much of the work carried out so far has not yet been Gill, J. M. & James, G. A. A study on published. The following list of the discriminability of tactual pi±ilications is divided into those point symbols. Research Bulletin papers which have appeared during the No. 26, American Foundation for course of the contract, and those the Blind, 19-34, 1974a. which have been submitted or are in preparation. Gill, J. M. & James, G. A. Mobility

maps : The choice of symbols . New Beacon, 1974b, LVIII, 35-38.

Heyes , A. D. A versatile ultrasonic Papers published since 1971: mobility aid for research purposes. Proceedings of Leonard Conference on Research into Visual Handicaps. Armstrong, J. D. Evaluation of 3-5 January 1972. Southern Re- mobility aids. Proceedings gional Association for the Blind, of a Conference on the Evalua- 1972. tion of Sensory Aids for the Visually Handicapped (Ed. James, G. A. Problems in the stand- symbol- Guttman) , pp. 101-04. National ardization of designs and Academy of Sciences, Washington, ization in tactile maps for the 19 72a. blind. New Beacon, 1972, LVI 87-91.

Armstrong, J. D. The adding of James, G. A., Armstrong, J. D., & visual information to previ- Campbell, D. W. Verbal instruc- ously recorded video-tapes: A tions used by mobility teachers functioning system. Quarterly to give navigational directions Journal of Experimental Psy- to their clients. New Beacon,

chology , 1972b, 24, 361-63. 1973, LVII, 86-91.

James, G. A., & Armstrong, J. D. Armstrong, J. D. The design and An evaluation of the "Silva" production of maps for the braille compass: 1. Static tests. visually handicapped. Mobility New Beacon, 1974, LVIII, 225-29. Monograph No. 1. Department of Psychology, University of James, G. A., Campbell, D. W. , & Nottingham, 50 pp. 1973. Armstrong, J. D. An evaluation of the "Silva" braille compass: 2. Field trials. New Beacon, Armstrong, J. D. Mobility problems: 1974, LVIII, 253-55. A review of methods and aids. Report on European Conference Leonard, J, A. Studies in blind mo- for Technical Aids for the bility Applied Ergonomics , 1972, Visually Handicapped. Handi- 37-46.

kappinstitutet , Stockholm, 1974. Leonard, J. A. The evaluation of blind mobility. Research Bulle- tin No. 26. American Foundation for the Blind, 13-76, 1973.

148 . ,

AN ELECTROMECHANICAL NUMERIC BRAILLE DISPLAY*

A Familiar Tactile Representation of Electrically Encoded Digital Signals

George F. Dairymple **

Summary: A simple electromech- The blind who want to work in anical numeric braille display has modern technological based jobs in- been developed and implemented. One stead of the more traditional jobs such system is now in use at a broad- have still another disadvantage; cast station. It enables an other- namely, much of the information wise qualified blind person to meet which they must use in their work the restrictions placed on his first is available only either as digital class radio telephone license by the encoded or analog electrical sig- Federal Communication Commission. nals.

This device can be interfaced How can the blind use these with almost any electronic instrument electrical signals or make the re- with digital binary coded decimal quired measurements? The moving- (BCD) outputs. It can make electronic pointer analog meter can be tactually instrumentation with visual digital read if the glass face is removed, a displays also available to the blind. braille scale added, and the user has a very delicate touch so as not to This concept can be extended to damage the meter. ^ This technique modules that display the full braille is not usable if the meter is deli- cell. Such modules can be used in cate (as is most generally the case) systems to give the blind almost if dangerous voltages exist in the equal access to the data presented meter, or if the signal is in digital displayed on video communication form. Another analog meter available terminals as enjoyed by the sighted. to the blind is the Servo Meter^ (American Foundation for the Blind) This device has a braille marked six inch (diameter) dial that the user reads. The dial is attached to a potentiometer and it is driven by a motor. Internal circuitry measures the differenc2 between the input voltage and ^he potentiometer voltage *Appeared originally in R. Foulds & which is proportional to shaft rota- Proceedings the B. Lund (Eds.), of The Servo voltmeter uses this Engineering Devices tion. 1974 Conference on signal to drive the motor in such a in Rehabilitation. Boston, Mass.: way as to reduce this difference to New England Medical Center Hospital, zero. This meter solves the diffi- 1974. culties of the moving pointer meters except that only an analog voltage **Acting Director, Massachusetts can be used. Institute of Technology, Sensory Aids Evaluation and The trend at the present time in Center. Development engineering is towards digital data

149 , with analog indicators primarily re- is that metal surfaces instead of served for certain applications where paper surfaces are used. they are superior; e.g. matching two values, etc. The braille display is composed of modules, each containing two cells. There are many ways to present It is necessary to make dual cell digital data with tactile representa- modules, as the smallest solenoids tion. Several different solenoid- with both adequate force character- operated single braille cells have istics and sufficiently high reli- been developed. All of those known ability are 7/16 inch in diameter. to the author contain either one or This allows the solenoids to be

two cells or have much larger than mounted on 1/2 inch centers , twice standard dimensions.^' ^ If addi- the pitch of standard braille cells. tional braille cells are required, Smaller solenoids could be used but then they must be in widely separated those available have not demonstrated groups of one or two cells. sufficiently high reliability for this use. These modules can then A line of ten pneumatically op- be stacked to form a line of elec- erated braille cells has also been trically controlled braille numbers demonstrated,^ Each cell is a com- of any desired length. plete replaceable assembly and includes a solenoid operated latch. The module, shown in Figure 1, When the latch is energized, plungers consists of a head or pin guide which for all dots in that cell are free to positions the movable pins in the move, according to the pressure in braille matrix. Each pin protrudes the pneumatic control line for that through the head when its corres- dot. There are six pneumatic control ponding solenoid is energized such valves that cause either a slight that the pin can be felt. The head

over pressure or a slight vacuum to .,. structure also contains springs that

be on its associated control line ,, hold down the individual pins below when the data in a selected cell is the reading surface when the associ- to be changed. The demonstrator in- ated solenoid is not energized. The cluding a keyboard was packaged in an head is attached to a solenoid frame attache case and was powered by in- or main support. Each pin is con- ternal rechargeable batteries. nected to or controlled by a lever that serves as a connecting link with There have been other tactile its appropriate solenoid. These lev- digital displays developed. One con- ers are arranged in two vertical lay-

sists of a matrix of solenoid actu- ., ers to prevent mechanical interfer- ated pins arranged in a matrix of ence. four rows and as many columns as digits 6 f The data is presented in The head can also contain a latch Binary Coded Decimal (BCD) form. The arrangement for holding the braille top row represents ones, the second pattern without expending power in row represents twos, the third row the main solenoids supporting the pins. represents fours and the fourth row The latch can be controlled by either represents eights. Thus to read the two solenoids; i.e. one solenoid for display, each pin must be sensed and each cell, or a single solenoid for the necessary addition to determine both cells of the module. The latch the value of the digit displayed must has not been demonstrated. be performed by the user. The display can be operated di- The display presented here is a rectly through suitable drivers and braille display. Each braille cell code translators from the BCD outputs corresponds to a digit and conven- or most digital electronic instru- tional braille patterns are used to ments. The first display used 4 SSI

represent the numbers. Since on]y (Small Scale Integration) , TTL (Tran- the four dots per cell are required sistor-Transistor Logic) integrated to give the braille number patterns circuits and 4 Darlington transistors only four are implemented. The to drive each cell. The BCD data braille cell dimensions used are source, an Analogic Digital Panel those used in the Perkins braille- Meter, was incorporated into the writer. The only difference between package both to keep the lead lengths this display and conventional braille short and to make the braille meter a complete unit.

150 Figure 1. Output Module, Including Solenoid Drivers

The next numeric braille dis- down position the mode switch causes play, shown in Figure 2, was made to the meter to hold its present reading be used with the metering equipment and the braille cells to display that of a Continental Electronics 5 KW value. The mode switch is a center broadcast transmitter at an Engle- off, locking up and momentary down wood, Colorado station. The display switch. To hold the system the mode is located on the top of the cabinet switch must be held down. at the left of Figure 2. The controls, on the front of the cabinet, It was not feasible to mount the are a mode switch and the power braille display within the metering switch. When the mode switch is up, rack of the transmitter, but the dis- the braille display follows the vi- play is such that it could be set on sual display of the meter. In the a shelf on the front of the rack.

Figure 2. Complete System Including the Numeric Braille Display, Multiplexer Encoder, and Digital Panel Meter.

151 .

A multiplexer system was built both The latest unit is very similar to reduce to six the number of wires to its immediate predecessor. A third necessary between the display and the output module is used such that four meter from the 19 wires that would digits and a decimal point can be dis- otherwise be necessary. With the played. Since all digits are the multiplexer four line drivers are same, the decimal point location is needed instead of the seventeen that determined by the multiplexer encoder. would be needed with direct connec- Logic in the encoder places the deci- tions. The multiplexer is the card mal point in its proper location in on the right of Figure 2. the data stream.

The multiplexer has a further This type of digital braille disadvantage in that the decimal point play can be used for any numeric ap- can be inserted at the correct loca- plication. For example, it is planned tion and moved by a simple logic sig- to connect five of these modules (10 nal. This feature is not in the unit braille cells) to an 8-digit four- at the broadcast station. The en- function calculator. This calcula- gineer had to devise a braille table tion will have both a visual seven- listing the location of the decimal segment display and a tactile four-dot point and the electrical units mea- braille display. sured versus the meter function select switch position. This braille module concept can be extended to a system with a full There are two other peculiarities braille cell; i.e. 6 dots per cell. with this particular unit. Since the A line of alphanumeric braille cells meter is a 3 1/2 digit meter, the can then be interfaced with a "soft most significant digit is either a copy"; i.e. video terminal, and pro- zero or one. Further the meter mea- vide the blind user essentially the sures both negative and positive same access to the data that a sighted voltages . To indicate a negative person has. Since only one or two input signal the multiplexer adds lines of most of the video display two to the most significant digit. can be displayed tactually at a time, If the tactile display indicates 2abc a selector switch must be included so or 3xyz the correct value is -Oabc or that the user can determine simply -Ixyz and rapidly which lines are available to him. The meter also has an over-range indication. Visually it is shown by a +1 or -1 with the trailing digits ACKNOWLEDGEMENT blanked. The multiplexer adds four to the most significant digit when this condition is encountered. The This work was partially supported resulting display when the input is by a grant from the E. Matilda Ziegler out of the meter's range is either Foundation and by the Social and Re- 5000 or 7000 depending on whether the habilitation Service, Department of signal is positive or negative. Health, Education and Welfare, Grant No. 23 P 55854/1-01.

REFERENCES

Wexler, A. Experimental Science Fieldgate, I. O. "Reading Machine for the Blind. New York: Pergamon for the Blind," U .S. Patent Press Inc. , 1961. 3, 395,247, July 30, 1968. "Servo-Voltmeter," Aids and Appli- Sutherland, N. B. "Development of ances. American Foundation for the a Dynamic Braille Display," MITRE Blind, New York, 18th Ed., p. 39. Corporation, MTR- 1951, Sept. 1970. Mick, D. E., Hittinger, G. N., "An Cranmer, T. "Braille Display for Inexpensive, Universal Braille Electronic Calcul ators," Patent Ap- (Unibrl) Output Device," Air Force plication 1974.

Weapons Laboratoiry , Albuquerque, "Position Feelout System," Elm Sys-

New Mexico, July 19 69. terns. Inc., Ameri can Machinist , June 1973, 1. 49. 152 .

THE AUTO-COM AT KENNEDY MEMORIAL HOSPITAL ^

Rapid and Accurate Communication by a Non-Verbal Multi -Handicapped Student

George F. Dairymple^ Ann Bullock ^ Janice M. Danca"*

The Auto-Com is a device to en- The Auto-Com was developed at able those with severe motor impair- the Cerebral Palsy Communication ment, e.g. cerebral palsy, to communi- Group of the University of Wisconsin- cate with others effectively. The Madison. The Auto-Com can be visu- device was developed by the Cerebral alized as an extension of the conven- Palsy Communication group of the Uni- tional letter board. With the letter versity of Wisconsin. A student at board a user points to a letter until that the Kennedy Memorial Hospital Day , a monitoring person understands School is now using the Auto-Com in the letter under the pointer is the her classroom. desired letter and records it. The Auto-Com replaces the monitoring person by a display and control elec-

tronics . ,., INTRODUCTION

An Auto-Com is being used by a DESCRIPTION OF AUTO-COM ten year old Kennedy Memorial Hospi- tal Day School student, Laura, who The board consists of 84 brightly has severe cerebral palsy. Although colored squares on a plastic covered Laura is unable to speak, the Auto- working surface. Each of the squares Com provides her with meaningful com- has a character painted on it. The munication. The Auto-Com has in- user moves a pointer to the desired creased her communication rate and character and holds the pointer over accuracy over that obtained using a the square. After a short, though vertically-mounted guarded keyboard adjustable period of time, the Auto- electric typewriter. Com responds and sends the selected character to the output display device. The output device on the unit ^ Paper originally appeared in used at Kennedy is a video terminal R. Foulds, & B. Lund (Eds.) Proceed- employing a commercial character gen- ings of the 1974 Conference on En- erator and a TV monitor. gineering devices in Rehabilitation Boston, Mass.: New England Medical reed switches are used Center Hospital, 1974. Magnetic as the sensors The pointer contains ^Acting Director, MIT, Sensory Aids a small magnet that operates the reed Evaluation and Development Center. switches. The most common pointer has been a piexiglass slider with a ^OTR, Kennedy Memorial Hospital, molded handpie ce and a magnet mounted Occupational Therapy Department. on it. A velc ro strap is used to ''Teacher, Kennedy Memorial Hospital secure the sli der to the hand of the Day School. user who is in capable of grasping

153 — the slider. The magnet is posi- encourage success and independence in tioned sufficiently far away from the the areas of cognitive skills, recep- grip so that the selected square is tive and expressive language, and vi- visible through the slider to the sual perceptual skills. user. Small strips of felt are se- cured to the underside to reduce the Unfortunately, the successful effort needed to move the slider. function of these devices and methods Laura cannot use a slider due to are often impeded by the interference severe spasticity in her upper ex- of motor incoordination, awkward po- tremities. She operates the Auto- sitioning of the device and a confu- Com with a felt covered magnet placed sion of the user's exact response. on the tip of her head pointer. As a result, you only have a "feeling" or vague idea that the child is The Auto-Com monitors the posi- alert and intelligent but you have no tion of the magnet at all times but effective means of knowing the degree does not respond until the magnet is and level of understanding by the held within the general proximity of child. Most important is that the a character for a short, adjustable children have a limited means of length of time. Thus the user has communication, expression, and inter- time to move the magnet if it is action with their environment. over the wrong character before the Auto-Com responds. A red light at the top of the board is illuminated TRAINING when a sensor switch is closed and the time delay is operating. When The Auto-Com has helped solve the character transfer occurs, a most of these problems. Initially, green light is turned on and a click the Auto-Com was located in a small is sounded. room for training individual children. It had three purposes: 1) training There are several control char- for nonverbal severely involved cere- acters available: carriage return; bral palsy children; 2) diagnostic space; cursor up, down, left or right; evaluation of traumatic physically home; and clear. The "home" charac- involved patients; and 3) motivation ter returns the cursor to the upper and training of children with learn- left hand position of the screen. ing disabilities. Initially, two The "clear" character erases the cerebral palsied children were trained. screen and returns the cursor to the One used the hand slider and the other home position. There is a signifi- the head pointer. The second child, cantly longer time delay associated Laura, was chosen to use the Auto-Com with clear character to prevent acci- in the school situation because of dental clearing of the screen. The the severe degree of her handicap, red switch-closed light also flashes her high motivation to learn and her during the clear time delay previous speed in mastering the equipment. to clearing as a warning that you are about to erase the screen. Previously Laura had used a vertically mounted guarded keyboard electric typewriter. Her training on the DESCRIPTION OF Auto-Com progressed in the following THE PROBLEM order: location of letters —numbers her name — carriage return (CR) --names Working and interacting with of her brothers and sisters —correc- nonverbal severely physically handi- tion via back spacing— copying words capped children has revealed a need (days of the week) — answering yes/no for an objective tinderstanding and or multiple choice questions — inde- evaluation of their knowledge and pendent words--speed tests — reading thought processes. Until now com- tests and mechanics of addition with munication by nonverbal cerebral correct signs. palsy students has involved a variety of methods--f rom a simple yes/no re- The correction capabilities of sponse to a pointing response by hand, video display of the Auto-Com has foot, or head while using a communica- helped determine the cause of many tion board to typing with a modified of the errors which she made while typewriter. Through these methods we using the typewriter. Her limited have attempted to evaluate and control abilities had prevented her

154 . .

from correcting all the mistakes she this potential. However, it did made. Now she has the ability to present many problems: 1) it tended correct any mistake she recognizes to isolate Laura from the rest of the as a mistake. Hence any errors re- class. She would always be working maining when she has completed writ- on a different task than her peers ing using the Auto-Com is a genuine as she was unable to work on paper- error, not one caused by her limited oriented tasks. 2) The typewriter motor capabilities. was non-corrective which increased her level of frustration. 3) The After she had learned to use all keys were too small and spaced too of the functions of the Auto-Com closely to each other. There was a while using it in a structured set- high level of error due to Laura's ting, it was decided that she should spasticity use it in a more purposeful functional setting — the classroom. When the Auto-Com was introduced, it opened a wide diversity of academic and social activities CLASSROOM PERFORMANCE for Laura. We were able to document how her excellent visual and auditory Laura has attended Kennedy receptive memory had given her a Memorial Hopsital for Children Day large sight vocabulary (mostly three- School for the past four academic letter nouns, color words, days of years. She is an alert and curious the week, number words, and a few

child who enjoys group situations descriptive adjectives) . From these and is well accepted by her peers. words she has been able to build and Her prior training with a head point- is now at the level of structuring er allows her to be involved in simple sentences. pleasurable activities such as paint- ing and coloring; yet this did not Laura has become a more involved address her real problem— communica- member of the class with the Auto-Com. tion. Her severe physical limita- She gives the children simple greet- tions present difficult and at times, ings in the morning since she quickly impossible situations for her per- learned how to write each child's formance in academic tasks and so- name cial activities; e.g., many other children with severe motor problems When an activity involves a can develop a pencil grasp or alterna- worksheet, Laura can now participate. tive methods. It is no longer necessary to provide a different activity for her as class With the sensitivity of a child, worksheets are adapted for her use. Laura began to feel the impact of Also, her work can be checked from a these limitations and demonstrated distance by simply reading the screen. a low level of frustration. This The Auto-Com has assisted the reading was particularly evident when she clinician in covering a greater amount was unable to make a response (other of material since Laura can respond than yes/no) to a question or to ex- spontaneously to questions asked. A press a feeling to her classmates and phonics program has been started for teachers. This child is now ten years Laura with excellent results. old and any frustration that is now evident will certainly increase with As her dependence on teacher and the approach of adolescence. While classmates has lessened with the use Laura always appeared to be an alert of the Auto-Com, Laura has become a child, it wasn't possible to deter- more self-confident child. She can mine her real potential in the areas be given a task and work on it inde- of reading, arithmetic and language pendently without interruption and skills. When training on the type- increasingly without assistance from writer began, however, her therapists the teacher. This has generally given and teachers began to suspect that Laura a better self-image which has Laura had a high retention level and very obviously affected her total intellectual potential. personality. The Auto-Com has been the most helpful in giving us substantial evidence of her abilities. The electric typewriter which With this information, we have been was used with the head pointer seemed able to set some realistic academic the most productive solution to pursue

155 , objectives and precise goals for Wisconsin, the Madison Public School Laura. The teacher and clinicians System, the Robert J. Ritger Memorial now have a clearer understanding Fund, the Dane County Chapter of of her capabilities and comprehension United Cerebral Palsy and the Bacon of all academic skills. Foundation. Additional development is being supported by NSF under grant EC-40316. ACKNOWLEDGEMENT The work reported here was performed at Kennedy Memorial Hospital The Auto-Com was developed at for Children and M.I.T. and is par- the University of Wisconsin-Madison tially supported by a grant from the with support of the University of Social and Rehabilitative Service, Department of Health, Education and Welfare under grant #23 P 55854/1-01.

REFERENCES

Vanderheiden , G. C. , Lamers, D. communication aid for the severely Volk, A. M., & Geisler, C. D. handicapped. Camahan Conference A new means of oommuniaation for on Electronic Prosthetics, Univer- the severely handiaapped: The sity of Kentucky, Lexington, Ken- auto-monitoring oommuniaation tucky, 1973, pp. 47-51.

board (Auto-Com) . (Abstract)

Fourth Annual Meeting BioMedical Vanderheiden, G. C. , Volk, A. M. , & Engineering Society, January 29- Geisler, C. D. An alternate inter- 30, 19 73. computers the physical- _ , face to for ly handiaapped: The auto-monitoring

Vanderheiden, G. C. , Volk, A. M. , & communication board (Auto-Com) , Geisler, C. D. The auto-monitor- 19 7 3 National Computer Conference. ing technique and its application (In press.) in the auto -monitoring communica-

tion board (Auto-Com) : A new

156 ,

REPORT OF THE PERCEPTUAL ALTERNATIVES LABORATORY FOR THE PERIOD JULY 1973 — JUNE 1974

Emerson Foulke*

Perception is the process of ac- the first three months of the current quiring information by searching for fiscal year were included within the and interpreting sensory stimulation. period for which support was provided When the functioning of a sensory, by funds from the Clark Foundation. cognitive, or motor component of a During the last nine months of the perceptual system is impaired, the current fiscal year, the laboratory's affected individual must find another program of research has received sup- way of acquiring functionally equiva- port from a grant of $75,000, awarded lent information; he must develop a by the Grant Foundation. Support perceptual alternative. On June 25, from this grant will continue through- 1969, by an action of the Board of out the fiscal year ending June 30 Trustees of the University of Louis- 1975. The Perceptual Alternatives ville, the Perceptual Alternatives Laboratory, in collaboration with Laboratory was established, under the Ada-Max, an audiovisual engineering auspices of the Graduate School of firm in Jackson, Michigan, is inves- the University of Louisville, for the tigating techniques for the retrieval purpose of investigating perceptual of information from tape recorded alternatives. At its present stage texts , and this investigation is re- of development, the research program ceiving support from a grant of of the Perceptual Alternatives Lab- $15,5 82.96, awarded to the Perceptual oratory is concerned primarily with Alternatives Laboratory by the Seeing the investigation of perceptual al- Eye Foundation, Incorporated. ternatives for individuals with visual impairment. However, perceptual I have applied to the National alternatives for visually unimpaired Institutes of Health for a grant of children with severe reading problems $66,911 to support a program of re- and for children with profound motor search in which the objectives are to disabilities are also receiving at- determine, through basic perceptual tention, and the long-range objective research, the requirements that must of the laboratory is to investigate be satisfied by devices intended to the entire spectrum of useful per- aid low vision, to use those require- ceptual alternatives. ments as guidelines in designing and constructing a family of low vision The laboratory's program of re- aids, and to evaluate the effective- search has received support from ness of the aids that are constructed three external sources during the in field tests involving members of year covered in this report. On the population for which they are in- October 1, 19 72, the laboratory re- tended. If this application is suc- ceived a grant of $25,000 from the cessful, it will provide support for Robert Sterling Clark Foundation, and a three year period beginning Janu- ary 1, 1975. The proposed research *Director, Perceptual Alternatives will be xindertaken in collaboration the Laboratory, University of Louisville, with Dr. Paul Jones, a member of Louisville, Kentucky. Psychology Department at the

157 University of Louisville, and Mr. and as a consequence, tape-recorded J. Malvern Benjamin, President of texts are less effective than their Bionic Instruments, Incorporated, inkprint analogs. Bala Cynwyd, Pennsylvania. The retrieval problem may be reduced by recording, at appropriate ACQUIRING INFORMATION locations in the track on which text BY LISTENING is recorded, tones so low in frequency that they are nearly inaudible. Be- The information that is ordi- cause they are nearly inaudible when narily acquired by reading the printed the tape is reproduced at the normal page can also be acquired by listen- playback speed, their presence does ing to the recorded oral reading of not interfere with the perception of another person. Though reading by speech that has been recorded on the listening is not as effective as the same track. When the tape is played visual reading of a skilled reader, back in the fast forward mode on a there are circumstances under which tape recorder that has been modified it may be a desirable alternative to so that the tape remains in contact visual reading. The busy physician with the playback head during fast or executive may make better use of forward operation, the tones are in- the time he spends in his car by lis- creased in frequency by an amount tening to recorded oral reading as he that is proportional to the increase drives. The skilled worker may be in tape speed, and they are heard as aided in carrying out a complicated clearly audible beeps, displayed assembly requiring continuous visual against the background of high-pitched monitoring by listening to a sequence chatter that results from reproducing of recorded oral instructions. The speech at the fast-forward tape speed. child with normal vision who does This system of indexing is currently not learn to read effectively, and employed by Recording for the Blind who is not aided by efforts of re- in its preparation of recorded text- mediation, may find reading by lis- books . A rudimentary code with two tening the most viable alternative. characters is used. One character, Reading by listening is an obvious a single beep, signifies the begin- alternative for those who have little ning of a new page. The other char- or no vision, and under many circum- acter, two consecutive beeps, signi- stances, it offers a better alterna- fies the beginning of a new chapter. tive than braille or large print. The An announcement at the beginning of investigation of reading by listening each track indicates the pages cov- is a major component of the labora- ered on that track, and if a new tory's program of research. This in- chapter starts on the track, this vestigation includes research intended fact is also announced. With this to explicate the processes on which information, the reader can, by in- reading by listening depends , and re- terpreting code characters that are search to evaluate a variety of meth- manifest during fast forward opera- ods and instruments intended to fa- tion, locate pages and chapters in cilitate reading by listening. which he is interested.

An index code with a larger num- The Comparative Evaluation ber of code characters should permit of Simple Index-Codes for a more detailed search of a tape- the Identification of Locations recorded text and the result should in Tape-Recorded Texts be more efficient retrieval. Accord- ingly, a project was initiated in A serious problem experienced by collaboration with the American Print- those who read by listening to tape- ing House for the Blind to investi- recorded texts is the difficulty of gate the feasibility of an expanded finding specific locations, such as index code. The objectives of this chapter headings, paragraph headings, project are to develop a code that or the locations at which new pages appears to be suitable for indexing begin in the inkprint analogs of such purposes, to demonstrate that it can texts. Because these locations can- be learned and used for retrieval of not be found easily in tape-recorded information by blind students, and to texts, the retrieval of desired in- conduct more extended field tests in formation is an inefficient process. which blind students are given the

158 . . opportunity to use tape-recorded texts character, and four seconds later, that have been indexed with the char- its name was pronounced. After an- acters in the code other two seconds, he heard the second character and so on, until all

The selection of a code . Tones six characters had been presented. produced by an audio- frequency oscil- On the second trial, the subject was lator were used to compose code char- instructed to guess the name of each acters because their generation is a character during the interval between simple matter, because they can be its presentation and the pronunciation recorded satisfactorily, and because of its name. Learning trials were ad- they can be varied in several dimen- ministered to each subject until he sions in which human observers demon- met a criterion of two consecutive strate good discrimination. Six ele- errorless trials. ments for use in composing code char- ^ acters were realized by varying the The results of this experiment signal produced by the oscillator in are summarized in Fig. 1. This fig- three dimensions. This signal could ure is a graph with five curves, one have either of two durations, either for each of the five codes. The of two frequencies , and it could ex- scale values on the x-axis are trial hibit either of two portamento ef- numbers. The scale on the y-axis fects — a tone that increases in fre- indicates the number of subjects to quency during its course, and a tone whom a given trial was administered. that decreases in frequency during The height of a curve above any given its course. Using these elements, trial nximber on the x-axis is a func- five codes were constructed and sub- tion of the number of subjects to whom jected to comparative evaluation. that trial was administered. Since Each code had six characters, and all subjects received the first trial, each character was composed with all five curves start at 20 on the either one or two elements. The ele- y-axis. However, as the subjects ments used in composing the charac- learning a given code reached cri- ters in the first code were long and terion, its curve gradually falls, short tones, or dots and dashes. The and reaches the x-axis when all 20 elements used in composing the char- subjects have met criterion. By com- acters in the second code were low paring the five curves , the impression and high-pitched tones. The two is strong that the first code was portamento effects were used to com- much more easily learned than the re- pose the characters in the third code. maining four codes, and this impres- The dimensions used in the first two sion was confirmed by an analysis of codes were combined to form the char- the variance of code learning scores. acters in the fourth code . The dimen- Consequently, the first code, the one sions used in the first three codes in which elements of two durations were combined to form the characters were used for the composition of code in the fifth code. When only one di- characters , was selected for further mension was used in composing char- evaluation acters, a total of six characters could be formed. However, combining dimensions increased the supply of Search and retrieval . An experi- possible characters , and so the char- ment has been planned to evaluate the acters used in the fourth and fifth efficiency with which blind school codes were chosen by drawing six char- children can find locations in re- acters at random from the supply of corded tapes that are indexed by code possible characters. characters. A junior high school history book was chosen for use in The five codes identified in this the experiment, and an outline was manner were taught to five comparable prepared which reflects the organiza- groups of subjects drawn from Intro- tion of that book . The book that was ductory Psychology classes at the Uni- chosen contains units that are divided versity of Louisville. Each group into chapters. Each chapter is further contained 20 members. Numbers were divided into subsections. Each unit used to name code characters , and a includes a section called "Unit Acti- paired-associates method was employed vities," and each chapter includes a for code learning. A learning trial section called "Reviewing the Essen- consisted of one randomly permuted tials." The locations at which these presentation of the six-code charac- five text components begin, and the ters. A subject heard a code locations at which new pages begin

159 20

5 - o IE

DURATION

• PORTAMENTO DURATION a PITCH PITCH a PORTAMENTO DURATION, PITCH 8 PORTAMENTO -r —r- 10 15 20 25 35 40 TRIAL NUMBER

Figure 1. Ease of Learning Five Index Codes are the six locations that are marked the appropriate code characters were by the six characters in the code. mixed and recorded on a final master We prepared two outlines in which tape. Cassette duplicates of master , these components are identified. tapes were then prepared for use in Each outline includes a couple of the experiment. These casettes will units and several chapters. The be reproduced on the cassette record- announcements of outline components ers sold by the American Printing were then recorded on tape at 20- House for the Blind. These recorders second intervals. The code charac- have been modified so that the play- ters that identity the announcements back head is not retracted from the of text components were generated by cassette during fast-forward and fast- an oscillator, switched on and off rewind operation, thus allowing it to by an electronic switch controlled scan the tape during these modes of by a battery of decade interval tim- operation. When a cassette contain- ers. The oscillator was set for a ing the indexed outline is played on frequency of 50 Hz . The timers were one of these recorders at the normal adjusted so that the duration of playing speed of 1-7/8 inches-per- short tones (dots) was 0.5 seconds, second, the announcements of text the duration of long tones (dashes) components are heard at 2 second in- was two seconds, and the intervals tervals. Of course, the code signals separating elements within a charac- that identify these announcements are ter were two seconds. Using the also reproduced, but since their fre- source tape and the apparatus for quency is quite low, and since the generating code characters , the an- low- frequency response of the cas- nouncements of text components and sette recorders on which they are

160 reproduced is not very good, they are that the use of the index code en- nearly inaudible. When a cassette is ables effective and efficient retriev- reproduced at the fast forward tape al, it will be used to index tape- speed (although this tape speed va- recorded texts. These texts will be ries, depending upon the amount of read by subjects who are given tasks tape that has been wound from the requiring search and retrieval, and supply hub onto the takeup hub, it the efficiency with which they per- is in the neighborhood of 30 inches- form those tasks will b^ compared per-second) code characters are clear- with the efficiency that character- ly audible against the background of izes the performance of such tasks high-pitched chatter that results when conventionally recorded texts from reproducing the announcements of are used. text components at the fast forward speed. At this tape speed, the frequency of the tones used in composing Other indexing schemes . Two al- code characters is approximately ternative indexing schemes are also 800 Hz. Short tones have a duration under consideration. In one approach, of approximately 30 milliseconds, and instead of using code signals, the long tones have a duration of approx- actual announcements of text compon- imately 125 milliseconds. The inter- ents will be recorded at 30 inches- vals separating elements within char- per-second, reproduced at 1-7/8 inches- acters are also approximately 125 mil- per-second, and imposed on the full liseconds . text recording at the appropriate lo-

cations . When a tape prepared in this Two of the code characters used manner is reproduced at the playback in preparing these cassettes, the one speed that is proper for the full text with a short tone followed by a long recording, the announcements will be tone and the one with a long tone heard as very low-pitched background followed by a short tone, are not noise. When the tape is reproduced symmetrical, and they are therefore in the fast- forward mode of operation, altered by reproduction in the fast- the announcements will be heard at rewind mode of operation. Preliminary approximately the correct pitch and experience strongly indicated the de- displayed against the background of sirability of code characters that high-pitched chatter that results would be reproduced properly in either from reproducing the full text in the fast-forward or fast-rewind mode the fast-forward mode of operation. of operation. Accordingly, two sym- In the other alternative approach, metrical characters with three ele- the feasibility of the Zimdex method ments -per- character (short- long-short of locating text components will be and long-short-long) were s\±istituted examined. According to this method, for the two asymmetrical characters, full text is recorded on the first and new tapes were prepared. These track of the cassette. On the sec- tapes appear to be quite suitable for ond track, consecutive numbers are use in the experiment that is planned. pronounced at a constant rate. The numbers are pronounced in reverse In this experiment, students in order, and a starting number is chosen grades 4 through 12 at the Kentucky that will permit the pronouncer to School for the Blind will learn the arrive at the number "1" just as the index code. The procedure will be second track is coming to an end. the same as the one employed in the Numbers are read in reverse order so experiment that was conducted in or- that number "1" will be adjacent to der to select the code, but code char- the beginning of track one. Track acters will be associated with the one is then played, and at each lo- announcements of text components in- cation that is to be indexed, the stead of with numbers. They will then cassette is turned over and the Zim- receive training in the use of code dex number recorded on track two at characters to find locations in one that point is determined. If the lo- of the recorded and indexed outlines. cation to be indexed falls between Upon the completion of training, the two Zimdex numbers, the higher of the other recorded and indexed outline two numbers is used. A written chart will be employed in a test of the is prepared in which a Zimdex number speed and accuracy with which subjects is assigned to each of the locations can locate text components. If the that is to be indexed. The person results of this experiment indicate using this method consults the chart

161 , to determine the Zimdex number asso- these dyads is small, readily avail- ciated with the location he wishes to able, and relatively inexpensive. find. He then searches track two of The electronic logic required to the cassette until he finds that num- interface this apparatus with de- ber, and by turning the cassette over vices that display their results at that point, he is assured of being digitally is well known and easily very close to the location for which provided. However, in addition to he is searching. its feasibility from an engineering point of view, a display must also be The method just described is perceptually feasible. This means, suitable for use on any conventional in the present case, that it must be cassette recorder. If a cassette re- possible for human observers to make corder is modified by replacing the absolute identifications of the dyads half-track playback head with a in the touch tone code with sufficient quarter-track stereo head and a two speed and accuracy to warrant their position switch for choosing the use as code signals in situations in quarter-track that is to be repro- which there is little tolerance for duced, a variant of the method be- error. comes possible that may prove to be more convenient. The full text is The results of the experiment recorded on one-quarter track, and that was conducted to evaluate this the Zimdex numbers , now pronounced possibility were disappointing. Num- in forward order, are recorded on the bers were used as names for the dyads parallel quarter-track. As before, and subjects were given trials in a chart showing the assignment of which they were to learn to name the

Zimdex numbers to text locations is dyads . The intention was to admini- prepared. Using the appropriate ster trials until a criterion of two switch setting, the user searches consecutive errorless trials was at- the second quarter-track until he tained. However, it soon became ap- finds the Zimdex number that marks parent that this criterion was un- the text component for which he is attainable, and the training of each searching. Then, simply by operating subject was therefore discontinued the track selector switch, and with- when the experimenter was convinced out even having to stop the cassette that no further improvement would transport, he finds himself listening occur. to the full text track at a point that is very near the location for The touch tone code was probably which he is searching. difficult to learn because both of the tones in a dyad may vary from Both the Zimdex method and the signal to signal. If only one tone method in which actual announcements varied from signal to signal, and the are used to mark text components are other tone served as a fixed refer- under investigation at the present ence against which the variable tone time. We expect that all of the in- could be judged, a dyad code might be dexing schemes described in this sec- easier to learn. Accordingly, a section will prove to be useful, and our ond experiment was conducted in which intention is to define the character- subjects attempted to learn a code istics under which each is most use- with 11 dyads as signals. The lower ful. tone in each dyad was always the same, while the higher tone varied from signal to signal. The reference tone The Absolute Identification in each dyad was the bottom note in of Tonal Dyads a major scale, while the 11 higher tones were the remaining 11 notes in In last year's report, an account that scale. As before, numbers were was given of an experiment that was used as names for the dyads, and sub- conducted to evaluate the feasibility jects were administered learning tri- of the dyads in the touch tone code als in which their objective was to used by the telephone company as in- learn the names of the dyads. dicators of the values registered on digital displays, such as those found The results of this experiment, on electronic calculators. The appa- and the results of the earlier ex- ratus required for the generation of periment in which names were learned

162 .

for touch tone dyads , are shown in decided to discontinue this line of Fig. 2. The two curves in this fig- research. ure are learning curves for the two codes . The curves were determined by computing mean error scores for The Use of Concurrent consecutive blocks of five trials, Abstract for Search and by plotting these scores against and Retrieval in trials. The graph indicates that al- Tape-Recorded Texts though the dyads with fixed reference tones were learned more easily This project has been undertaken than the touch tone dyads, they were in collaboration with Maxwell A. not learned well enough to warrant Kerr, an electrical engineer who is their use as code signals. If the the president of Ada-Max, an audio- tones in each dyad had been presented visual engineering firm in Jackson, consecutively instead of simultane- Michigan. Kerr contributed the con- ously, there is little doubt that cept of the concurrent abstract. He subjects would have found their iden- has made significajit progress in de- tifications easier to learn. However, veloping a procedure for subjecting since the original objective was the full text to a kind of scrutiny that practical objective of determining permits the rapid abstraction of se- the suitability of a readily avail- quences of crucial words and phrases able device for displaying the values which, with only minor editing, form registered on digital displays, and meaningful statements. When these since consecutive presentation of the statements are read in sequence, the tones in dyads would require the de- result is an abstract which exactly velopment of a new device, I have parallels the full text, but which

SINGLE TONE CODE" 12 ELEMENTS

"- i TONAL DYAD CODE II ELEMENTS

5- TRIAL BLOCKS

Figure 2 Ease of Learning Two Dyad Codes

163 is reduced to approximately one-eighth track of the master tape on parallel of the length of the full text. Al- tracks of a cassette. This cassette though the procedure for creating the is reproduced on a cassette player abstract is not yet fully developed, that has been further modified by Kerr's results suggest that little replacing the playback head with a training will be required for its stereo playback head, and by the pro- mastery, and that people without vision of a switch for choosing be- background in the text that is to be tween the two tracks scanned by this processed can produce satisfactory head. When the cassette is repro- abstracts of technical reports. duced at the normal tape speed, the switch is set for the full text track, When a text has been abstracted and the speech recorded on it is heard in this manner, both the text and without interference. The first ap- the abstract are typed, in adjacent proach requires less modification of columns, on rolls of chart paper. equipment and permits greater tape The transport mechanism of a chart economy. The second approach results recorder is used to wind the typed in a more satisfactory listening ex- chart paper from its supply spool on- perience. The choice between the ap- to a take-up spool. The rate at which proaches will depend upon information both the full text and the abstract acquired by consulting the experience are read orally is determined by ad- of listeners. justing the speed of the motor that drives the take-up spool. The oral Rapid scanning is an important reading of the text is recorded at a ability of the visual reader that is tape speed of 1-7/8 inches-per-second not shared by the aural reader. This on one track of a tape. The oral ability permits the visual reader to reading of the abstract is recorded gain a general impression of the con- at 15 inches-per-second on another tents of a book and to locate sec- track on the same tape. If this tape tions of particular interest to him is then reproduced at 1-7/8 inches-- for more careful reading. When he is per-second, the signals obtained from reviewing a book that he has already the two tracks can be mixed and re- studied, the ability to scan rapidly corded on one track of a cassette. permits him to spend his reviewing This cassette is reproduced on a time efficiently. With cassettes pre- cassette player that has been modi- pared in the manner just described, fied so that both the normal tape scanning can become an ability of the speed of 1-7/8 inches-per-second and aural reader, as well. Using a cas- a tape speed of 15 inches-per-second sette with the proper format, and a are available. When the cassette is cassette player with the necessary played at the normal tape speed, the modifications, the aural reader might reading of the full text is heard start by listening, at the fast tape against the background noise that re- speed, to the concurrent abstract, sults from reproducing the abstract which would suffice to give him a recording at one-eighth of its re- good idea of the contents of the full cording speed. At this speed, the text. When he encountered a section abstract recording is so low in fre- he wished to study more carefully, he quency that it is unintelligible, and would reduce his tape speed and listen nearly inaudible because of the lim- to the full text. ited frequency response of the cassette player. When the cassette is repro- The project just described is duced at 15 inches-per-second, the ab- funded by a grant of $15,582.96, stract recording is heard against the awarded to the University of Louis- background of high-pitched chatter ville by the Seeing Eye Foundation, that results from reproducing the full Incorporated. The project is behind text recording at eight times its re- schedule because of difficulties en- cording speed. At 15 inches-per-sec- countered by Kerr in his efforts to ond, the full text recording is unin- provide the transport mechanism needed telligible, but it adds a considerable for handling chart paper, the tape re- amount of noise to the signal in which corder needed for making the full text the listener is interested and this and abstract recordings, and the modi- noise may constitute unacceptable in- fied cassette players on which cas- terference. This problem can be solved settes are to be reproduced. However, by copying the signals obtained from the transport mechanism has now been the full text track and the abstract delivered, and Kerr claims that the

164 .

remaining equipment will follow short- for the Blind include the modifica- ly. Donald Ferguson, a Research As- tions required for the proper resistant employed with project funds, production of dictionary cassettes. has gained experience in the prepara- When the dictionary cassette is tion of abstracts and the taped need- played on the modified recorder at ed for completion of his project 15/16 of an inch-per-second, the full should be available before long. When text recording is reproduced properly. these tapes are available, evaluation The index recording is reproduced at experiments will be conducted to de- a speed that is so much slower than termine the extent to which the avail- the speed at which it was recorded ability of concurrent abstracts en- that it is nearly inaudible. Only ables the aural reader to realize the an occasionally low-pitched rumble advantages of rapid scanning. is heard, and it is unobtrusive. When the cassette player is placed in the fast- forward mode of opera- The Talking Dictionary tion, the index recording is intel- ligibly reproduced and displayed The objective of this project is against the background of high-pitched to record, on tape, the oral reading chatter that results from reproducing of a dictionary in a manner that will the full text recording at the fast- permit rapid search of the recorded forward tape speed. Since, at this format and efficient retrieval. The speed, the full text recording is un- dictionary is presented on cassette. intelligible, it adds only noise to To prepare a cassette, two tape re- the signal that is being processed cordings are needed. One tape, re- by the listener and the interference ferred to hereafter as the text tape, it contributes is tolerable. As the contains the full text of the dic- listener scans at the fast-forward tionary. The other tape, referred speed, he must make slight adjustments to hereafter as the index tape , con- in motor speed in order to maintain tains only the pronunciations of the the index recording at the proper words that are pronounced, spelled, pitch. Each word in the index re- and defined on the text tape. The cording is so located on the tape index tape is recorded at 15 inches- that its termination and the termina- per-second, which is an approximation tion of the definition preceding the of the fast-forward tape speed of the definition to which the index word modified cassette players sold by the refers occur at the same point on the American Printing House for the tape. Each cassette contains a Blind, when they are adjusted for a braille label and a large print la- normal playback speed of 15/16 of an bel indicating the first and last inch-per-second. In order to prepare word recorded on each track. By a dictionary cassette, the text tape referring to these labels, the lis- is reproduced at the speed at which tener selects the cassette and the it was recorded, and the index tape track containing the word in which is reproduced at 15/16 of an inch-per- he is interested, and plays the cas- second. The signals from these tapes sette at the fast-forward speed until are mixed and recorded on one track he hears the word for which he is of a cassette at 15/16 of an inch-per- searching. He then changes to the second. To reproduce this cassette slow-playback speed and listens to properly, two modifications of the the spelling and definition of that conventional cassette recorder are word required. Its motor speed must be continuously variable through a range A somewhat different implementa- that centers on a playback speed of tion of the talking dictionary has 15/16 of an inch-per-second, and the been made possible by an additional playback head must not be retracted modification recently provided on the from the cassette when the recorder cassette recorders sold by the Ameri- is placed in the fast-forward mode of can Printing House for the Blind. operation. This second modification These recorders are now available permits the tape to be scanned by the with stereo heads. A switch is pro- playback head when the cassette player vided that permits selection of either is running at the fast-forward speed. one of the two tracks scanned by the stereo head. This modification has The cassette players now supplied to House blind and physically handicapped been provided by the Printing to permit the proper reproduction of readers by the American Printing House cassettes with four tracks recorded

165 ,

on them. We can take advantage of tape with respect to the text seg- this modification by preparing dic- ments to which they refer. This de- tionary cassettes with the text re- vice is functioning satisfactorily, cording on one of the tracks scanned and an efficient procedure for pro- by the stereo head and the index reducing final master tapes now appears cording on the other track. In addi- to be at hand. tion to choosing the proper tape speed for the recording in which he This project is being undertaken is interested, the listener will have in collaboration with the American to use the selector switch to select Printing House for the Blind. Ac- the proper track, but by so doing, cording to the understanding that was he will be able to hear the recording reached in planning this collabora- in which he is interested without in- tion, the Perceptual Alternatives terference from the other recording. Laboratory is to develop the techniques The first approach has the advantage and procedures for producing a talking of requiring less modification of dictionary. The Printing House will equipment, and it permits greater ' review these techniques and procedures tape economy. The second approach and if they seem suitable for large results in a more satisfactory lis- scale production, the Printing House tening experience. The choice be- will participate with the Perceptual tween these approaches will depend Alternatives Laboratory in the conduct ' upon the preferences of experienced of field tests to determine the ac- listeners. ceptability and utility of such a dictionary. If these field tests are The time required by a listener affirmative, the Printing House will to consult this dictionary is consid- consider production of the dictionary. erably greater than the time required The techniques and procedures described by the normal visual reader to con- in the preceding paragraphs , and in sult a printed dictionary. However, last year's report, have been devel- if the alternative is a braille dic- oped to a point at which their evalua- tionary, the advantages of the talk- tion by the Printing House is appro- ing dictionary are more apparent. priate, and further progress on this The talking dictionary will require project awaits the outcome of this considerably less space and be sig- evaluation. nificantly cheaper than a braille dictionary. Though a superior braille In the meantime, an alternative reader may consume less time in con- procedure for the production of master sulting a braille dictionary than he tapes for the talking dictionary has would in consulting the talking dic- been developed for consideration by tionary, the difference is not so the Printing House. Under this pro- marked as in the case of a visual cedure, the tape players used to re- reader, and it may disappear alto- produce the text tape and the index gether in the case of a poor braille tape are operated by a technician, in- reader. Furthermore, most of the stead of being controlled automatic- blind people who read by listening do ally. A technique is employed that not read braille at all, and for them, permits very accurate timing of seg- the talking dictionary may provide ments on the text tape and the index the only alternative to dependence on tape, and with this information, the the assistance of a visual reader in technician can combine segments from consulting a dictionary. the two tapes with a high degree of precision in constructing the master This project was also described tape. Furthermore, he can monitor in last year's report, and that re- the process in which he is engaged, port included an account of the de- so that errors can be detected and velopment of an improved device for corrected as they are made. The first automatic control of the tape players method requires less of the time and on which the text tape and the index effort of human operators. The sec- tape are reproduced in order to pre- ond method permits more efficient de- pare the final master tape from which tection and correction of errors, and cassettes are duplicated. By sensing because it permits greater precision

control tones, this device starts and in the combination of segments , there stops both tape players at the appro- is less wasted time in the final mas- priate times , so that index words are ter tape. The choice between these correctly located on the final master methods will depend upon their

166 comparative evaluation by the Print- When these prototype s are avail- ing House. able, reading matter will be prepared for reproduction on them, A few aural readers will be given training in

The Development of an the use of these instrume nts , and Aural Reading Instrument their experience will be taken into account in designing more formal ex- In one line of investigation periments to evaluate tl)e ir contribu- carried on in the laboratory for tion to the reading proce ss, and in several years, an effort has been refinement of their funct ional capa- made to identify, by analyzing the bility. reading behavior of visual and aural readers, those operations that account for the relative efficiency of The Development of an visual reading, and that are not or- Aural Testing Instrument dinarily available to the aural reader. This analysis has suggested the Acquiring test information by functional requirements of a satis- listening to the recorded oral read- factory aural reading instrument. ing of a test form is an obvious alternative for those who, for whatever There is no piece of equipment reason, cannot read print. However, on the commercial market that em- if such a recording is reproduced on bodies all of the required functions. conventional equipment--record player, However, an office dictation machine open reel tape player, or cassette manufactured in Switzerland, and dis- player--the aural reader is handi- tributed in this country by the Dicta- capped. The person who acquires his phone Corporation can, with extensive test information by reading a printed modification, be made to incorporate test form can easily reread any test enough of these functions to enable item if he is not sure of its meaning. an evaluation of the improvement in He will typically read a few items reading effectiveness and efficiency once, many items twice, and a few that might be realized through the items many times . Furthermore he can use of a functionally adequate read- maximize the number of test items he ing instrument. Two of the dicta- is able to consider by passing over tion machines distributed by Dicta- items that seem difficult to him on phone have been obtained. The Per- first reading, responding to the items ceptual Alternatives Laboratory is he finds easy, and returning to diffi- now negotiating a contract with Bionic cult items for more thoughtful consid- Instruments, Incorporated, a bio- eration. However, if the recorded medical engineering firm in the Phila- oral reading of a test is reproduced delphia area. Under the terms of this on conventional equipment, the per- contract, Bionic Instruments will ex- formance of these operations becomes tensively modify the dictation ma- so inefficient that the advantages chines so that they include enough gained by resorting to them are large- of the functional requirements of an ly cancelled by the time that must be ideal aural reading instrument to spent in performing them. Returning qualify them as prototypes of such an to a precise location on a conven- instrument. These prototypes will tionally recorded tape is a task that record information on an oxide-coated must be accomplished by a trial and plastic sheet instead of tape, because error process of successive approxima- of the relative ease with which in- tions. It consumes so much time and formation recorded on that sheet can becomes so aggravating, particularly be accessed. A magnetic medium is when it must be carried out under the preferred since its use will permit time pressure associated with the tak- the construction of an instrument with ing of most tests, that the aural the recording capability of a conven- reader will usually elect to forego tional tape recorder, and this capa- the advantages he might realize by bility should be given high priority retracing. If the test score earned in the design of an aural reading in- by the aural reader is depressed be- strument. These prototypes will make cause he has not had access to opera- provision for flexible review and tions which contribute to successful preview, rapid search, and precise test performance, and which are ac- retrieval. They will be battery op- complished with ease by visual read- erated and portable. ers, his score will not be a valid

167 indicator of the ability the test This device performs the fxonc- purports to measure. As described tions for which it was designed. How- in last year's report, arrangements ever, experience in using it rfevealed were made for the design and con- that it was inadequate in terms of struction, by Electronic Systems human engineering. There were too Development Corporation (known at many switches to operate, and the that time as Biotronics , Incorporated), necessity of operating switches at of a device to be used as an acces- two different locations — the cassette sory to the cassette recorder sold by recorder and the remote device — in the American Printing House for the order to control a single process Blind. With the device connected to proved to be undesirable. In an ef- his cassette recorder, the aural test fort to solve this problem, the com- taker listens to a cassette on which ponents of the remote device were test items have been recorded. At installed in the cassette recorder, the end of each test item, a brief itself. To do so, it was necessary 60-cycle tone, recorded on the same to sacrifice the battery compartment. track as the test item, is sensed by The power supply that permits opera- the device and the tape recorder is tion of the cassette recorder from automatically switched off. The de- the 117-volt ac line proved to be vice incorporates a high-pass filter marginally adequate in terms of fil- which protects it from the signal tering and current capacity, and the generated by the speech recorded on reliability of the control circuitry that track. If the operator wishes was not satisfactory. In addition, to proceed to the next item, he further consideration of this problem presses a momentary switch on the suggested that convenience of opera- device, which starts the cassette tion would be improved if the control recorder again. If he wishes to hear circuitry were built into a cassette the item a second time, he must press recorder that offers the review func- the momentary switch on the device tion. A recorder with this function and then the rewind button on the re- has an additional control key. With corder. He releases the rewind but- the recorder in the playback mode of ton when he hears a high-pitched tone. operation, pressing this key causes This tone is the 60-cycle tone that the cassette to be rewound. When the marks the boundaries between test key is released, the recorder is re- items, but since the tape to which he turned directly to the playback mode. is listening is being played at ap- Since the recorder sold by the Print- proximately 16 times the normal playing House does not offer the review back speed, the tone he hears has a function, it was necessary to search frequency of approximately 1,000 Hz. for another recorder with this func- Upon hearing the high-pitched tone, tion that would be satisfactory for the operator stops the recorder, and our purposes. A model manufactured the tape in the cassette is properly by Sony has been identified, and one positioned for hearing the test item of these recorders has been purchased. again. Of course, at the conclusion The control circuitry will be installed of the second hearing of the test in this recorder. In addition, its item, the cassette recorder is auto- playback head will be replaced with a matically turned off as before, and stereo playback head and a switch for the operator has the same options as choosing between the two tracks scanned before. He may repeat these opera- by this head will be installed. Test tions as many times as he wishes. He items will be recorded on one track may return to any desired test item and control tones on the other track. by holding the rewind button down, Since the second track will contain and by counting high-pitched tones only control tones , the circuitry that until he arrives at the tone marking senses those tones will not have to the location of the item in which he include a high-pass filter, and the is interested. If he loses count, he high-pitched tones that are repro- need only release his rewind button duced when the recorder is in the fast when he hears the next tone , and lis- forward or fast rewind mode will also ten to the number of the test item operate the control circuitry. Con- indexed by that tone to find his place sequently, when the review key is de- again. If he wishes to skip ahead, he pressed, the cassette will be rewound can press the fast forward button on only until a control tone is encoun- the cassette recorder and count high- tered. It will then be stopped, and pitched tones until he arrives at the the tape will be in the proper posi- location in which he is interested. tion for the rereading of a test item.

16 8 If these modifications are suc- A subject's score was the num- cessful, several recorders will be ber of words identified correctly. purchased and modified in the same The four columns in Table 1 show way. These recorders will be used the scores earned by subjects under in experiments designed to evaluate the four experimental conditions. the difference that may be realized The mean score for each condition is by the use of these machines for the shown at the foot of the column in administration of tests. which the scores earned under that condition are entered. Comparison of these means suggests that subjects Testing the Articulation found the natural speech more intel- of the Votrax Speech Synthesizer ligible than the synthesized speech. It also appears that subjects found Vocal Interface, a division of the synthesized words more intelli- the Federal Screw Works, has devel- gible when they were the first words than they were oped a speech synthesizer to which to be identified when it has assigned the trade name "Vo- identified after having first identi- trax." The Votrax synthesizer is fied the natural words. The apparent roughly the size of an attache case. differences among these means were It can be operated by parallel or examined by analysis of variance of serial code obtained from conven- test scores, and found to be signifi- tional code sources, such as com- cant. puters, punched tape, etc., and it can be operated from a typewriter- We expected to find that natural like keyboard, on which phonemes are words would be more intelligible than assigned to keys instead of letters. synthesized words, and in view of the The quality of the speech synthesized fact that the intelligibility of sin- by the Votrax synthesizer is surpris- gle words can be degraded consider- ingly good, and since it can provide ably without impairing the intelli- a voice for computers and many other gibility of sentences formed with devices that ordinarily display re- those words, it is our judgment that sults visually, its potential for the intelligibility of the speech offering a perceptual alternative is synthesized by the Votrax synthesizer considerable. Our first step in as- is adequate for many of the applica- sessing this potential was to evalu- tions in which we are interested. ate the Votrax synthesizer as an Furthermore, it must be remembered articulator. With the assistance of that our subjects in this experiment Max Bishop, a sales representative of had no prior experience with the syn- Vocal Interface, the Votrax synthe- thesized speech, and a lifetime of sizer was used to synthesize two 50- prior experience with natural speech. word lists of phonetically balanced The order effect expressed in these words. Each word was appended to a results is puzzling, and its explana- common carrier phrase, also synthe- tion will require further research.* sized on the Votrax synthesizer. The output of the synthesizer was recorded TABLE 1 on tape. Using the same common car- A Comparison of the Articulation of a rier phrase, the same two PB lists Human Speaker and the Votrax Speech were recorded on tape by a profession- Synthesizer al announcer. Under my direction, Tony Ahl , a senior student at the Group I Group II University of Louisville, conducted Syn . Hum. Hum. Syn an experiment in which two comparable 50 94 98 25 groups of subjects were first given 60 95 95 42 practice in the identification of PB 63 97 98 43 words, and then asked to identify the 70 91 83 43 natural words and the synthesized 77 98 98 46 words in the test lists. One group 80 97 97 53 heard the natural rendition of List A and then the synthesized rendition of x=66.43 x=95.43 x=9 5.0 x== 42.5 List B. The other group first heard the natural rendition of List B and then the synthesized rendition of List A. *An account of this experiment is given in the Senior Thesis written by Ms. Ahl.

169 .

The modest experiment just de- listeners, who will be asked to re- scribed is, of course, preliminary produce, in writing, the sentences in character. Bishop and I are dis- they hear. Each sentence will be cussing the possibility of further scored for the niamber of words cor- research concerning the intelligi- rectly reproduced, and an analysis of bility of Votrax speech, and if suit- the errors represented in incorrectly able arrangements can be made, several reproduced words will be attempted. experiments involving Votrax speech In another approach, the relative will be conducted in the Perceptual preferences of listeners for the sig- Alternatives Laboratory. nals of the speech compressors under evaluation will be deteinnined. The listening experience upon which pref- A Comparative Evaluation erences are to be based will be proof the Signal Quality of vided by extended samples of fluent Different Speech Compressors speech prepared on these compressors. With information about intelligibil-

In the past few years , rapid ity, and information about the pref- progress has been made in the devel- erence of listeners , it should be opment of the technology upon which possible to make meaningful compari- the time compression and time expan- sons among the output signals of the sion of recorded speech depends, and commercially available speech com- several electronic and electromechan- pressors. The results of this study ical speech compressors are now com- should be useful to manufacturers and mercially available. The Center for to consumers. Rate-Controlled Recordings, a unit of the Perceptual Alternatives Labora- tory, has, for many years, played a The Role of Aural Reading in prominent role in the conduct of re- the Public School Program search concerning time-compressed and time-expanded speech. Because the Adequate functioning in the pro- Center has no commercial interests at gram of education offered by our stake, it has received requests from public schools is predicated on the several manufacturers, as well as ability to read. In large measure, many of their prospective customers, a child must obtain, from printed for an objective evaluation of the reading matter, the information he signals produced by different speech needs to succeed in the tasks set compressors for him by his teachers. If he cannot read well, his ability to gather In last year's report, a plan was this information is impaired, and described for gathering the data need- his performance is adversely affected. ed for the comparative evaluation of Public school educators are becoming speech compressors. Because of the increasingly aware of a large number imminent commercial availability of of children in pxiblic schools who, a couple of compressors still under in spite of persistent efforts at development, we have postponed the remediation, never learn to read ef- study in the hope that we can include fectively. They do not perform well samples from these compressors in our in school, and it is tempting to con- evaluation. However, our plan is to clude that their poor school perform- compose sentences which adequately ance and their poor reading ability sample the phonemic spectriim, and are both the consequences of a super- record them on tape. These tapes will ordinate learning disability. This be compressed and expanded by several may sometimes be the case , but experi- degrees, on each of the compressors ments performed in this laboratory under evaluation. The specimen pre- and elsewhere raise the possibility pared on each compressor will be sub- that some of the children who read mitted to the manufacturer of that ineffectively can listen effectively, compressor. If the manufacturer is and have the ability to succeed in not satisfied with its quality, the school tasks, if they are not required master tapes used in preparing the to depend upon visual reading for in- specimen will be made available to formation. him, and he will be invited to prepare a specimen that fairly represents Last year, in collaboration with the performance of his compressor. Marilyn Kreisle, the coordinator of When specimens have been collected, the Learning Disabilities Program in they will be presented to groups of the Louisville Public School System,

170 a proposal to study the efficacy of reading because its maintenance does aural reading for learning disabled not demand overt behavior from the children was made to the Louisville aural reader. In order to read, the Public School System. Because of a visual reader must continuously pro- shortage of funds , the support reduce the behavior required for visual quested by the proposal could not be search of the printed page. If, be- provided, and our investigation of cause his attention wanders, or be- aural reading was temporarily sus- cause he falls asleep, he fails to pended. However, we are now prepar- produce this behavior, the reading ing another proposal in an effort to process comes to a halt. The aural obtain funding for an investigation reader is not required to engage in that we hope to initiate in time to any specific overt behavior in order get materials ready for use during to maintain the reading process in the Fall Semester of 1975. which he is engaged. His attention may wander, or he may fall asleep, Through testing and observation, but the tape recorder or record play- it has been determined that there are er to which he is listening will con- a significant number of children in tinue to deliver its recorded mes- the Louisville public schools, with sage. better than average intelligence, who have seriously inadequate reading If this analysis is sound, it ability. They perform poorly in should be possible to increase the school, and it is apparent that they effectiveness of aural reading by do so because they cannot acquire by finding a way to make continuation reading the information they need for of the aural reading process contin- success in school tasks. We are pro- gent upon behavior produced by the posing a study that will disclose the aural reader. An experiment to extent to which these children may evaluate this possibility is in benefit from the opportunity to read progress, and its description fol- by listening. Parallel forms of a lows . test of listening comprehension, one printed and one recorded on tape, will A reading selection with a level be administered to those children of interest and difficulty suitable whose poor scholastic achievement ap- for use with college students has pears to be a consequence of reading been chosen. A multiple-choice test disability. Those children who, on covering the facts and implications the basis of test performance, can be of this selection has been developed. described as poor visual readers but Its coefficient of reliability is adequate aural readers, will be di- 0.88. The selection has been re- vided into an experimental and a con- corded on track one of a stereo tape. trol group. Members of the experi- Control tones have been recorded on mental group will be given the oppor- track two of this tape at sentence tunity to do all of their studying by boxindaries. Two head sets are con- listening to recorded textbooks. The nected to channel one of the recorder effectiveness of aural reading will on which this tape is played. One be evaluated by comparing the experi- set is worn by the active listener, mental group with the control group and one by the passive listener. on such indices as pretreatment and Channel two is connected to an ampli- posttreatment scores on tests of fier, the output of which is rectified scholastic achievement, school grades and applied to the coil of a relay. earned before and after service in The tape recorder used in the experi- the experiment, and the subjective ment is one on which provision has appraisals of teachers. been made for the remote control of the start-stop function, and the remote control terminals are connected Active -Vers us -Passive together through a pair of normally Learning closed terminals on the relay. When a control tone is sensed, the relay It is generally true that learn- closes, the normally closed terminals ing proceeds more efficiently when open, and the tape recorder stops. the learning task is designed in such The relay is latched in this state a way that behavior on the part of the so that the tape recorder remains learner is required to carry forward stopped until the latch is broken by the learning process. Aural reading pressing a momentary switch. Since suffers by comparison with visual the control tones are recorded at

171 sentence boundaries, the tape recorder tion of each reader was recorded, and stops at the end of every sentence. samples lasting one minute were ob- The active listener must press the tained from these recordings . By momentary switch in order to obtain pairing each sample with every other each new sentence. His instructions sample, 66 pairs of samples were suggest that he can operate the mo- formed and rerecorded in a format mentary switch immediately, or take that placed the two samples in each advantage of the stopped tape in or- pairing on concurrent tape tracks. der to think about the sentence he During playback, each of the samples has just heard before going on to the in a pairing could be reproduced se- next one. The passive subject hears lectively by operation of a switch. what the active subject hears, but he The subject's task was to compare the has no control over the delivery of samples in each pairing, and in each sentences. The listening selection case, to express a preference for one is also presented to a control group of them. of subjects in the conventional manner. Upon conclusion of the listen- As was expected, preliminary ing selection, all subjects are tested analysis of the data indicated that for comprehension. When the experi- professional readers were, with one ment is finished, there will be three exception, preferred over experienced

distributions of test scores , one for amateur readers, and that experienced the group of active listeners, one amateur readers were, without excep- for the group of passive listeners, tion, preferred over inexperienced and one for the control group. Only readers. Since manipulation of oral a few sv±)jects have been run at this reading ability produced the expected writing, and the data are too meager variation in listener preference, we to suggest the outcome of the experi- seemed to have the kind of data that ment. would be suitable for a dimensional analysis of reading ability. Accord- ingly, the data were punched on Hole- A Dimensional Analysis rith cards , and a computer was used of the Productions to carry out an analysis, the objec- of Oral Readers tive of which was to disclose the multidimensional preference structure

Aural readers express clear ' expressed by the subjects who served preferences with respect to oral in the experiment. The analysis in- readers. In some cases, the bases dicated that there is no single di- for these preferences are evident. mension which will account for the However, in some cases, it is not expressed preferences. However, the easy to identify the factors that analysis that could be accomplished are evaluated by a listener in form- with the available computer program ing his preference. Even a superfi- could sort preferences into only four cial analysis suggests that several dimensions, and there was no distribu- dimensions must be taken into account tion of preferences among four dimen- in evaluating the production of an sions that would account for the ob- oral reader, but it is not yet pos- tained results . A more complex analy- sible to specify these dimensions, sis will apparently be required be- or the way in which the dimensional fore these data can be explained values represented in the production satisfactorily, and no decision re- of a given oral reader combine to garding the appropriate analysis has form the product that is preferred yet been made. by the listener.

In last year's report, an ex- The Effect of Oral Reading periment was described in which sub- Skill on the Comprehension jects were asked to evaluate all of of Time-Compressed Speech the possible pairings of 12 oral readers — four professional readers, Recorded oral readings of the four experienced amateur readers, eight selections in the comprehension and four inexperienced readers drawn subtest of the Nelson-Denny Reading from the student body at the Univer- Test were obtained from the 12 read- sity of Louisville, whose audition ers — four professional readers, four tapes were rejected by the panel of experienced amateur readers , and four judges that selects volunteers for inexperienced readers —who were em- Recording for the Blind. The produc- ployed in the experiment just

172 . , . described. The recordings were used 55 characters which stand for letters in an experiment conducted in col- or small letter groups in the braille laboration with Dr. Normal Lass, a code. To do this, they employed a member of the faculty of the School tachistotactometer. This is an in- of Medicine, Division of Otolaryngol- strument which controls the time of ogy, at West Virginia University. The exposure for dot patterns. Patterns recorded selections were presented at of dots rise above the display sur- four different word rates--175, 225, face of the instrument to the height 275, and 325 words-per-minute--to 480 of a braille dot, remain in position subjects drawn from classes at West for a preset time during which they Virginia University. The 12 rendi- press against the fingertip of the tions of the reading selections and observer, and then recede below the the four word rates at which they display surface again. A psychophysi- were presented combined factorially cal procedure is followed in deter- to produce 4 8 experimental treat- mining the minimum time of exposure ments, and each experimental treat- that is required for identification ment was administered to 10 subjects. of the dot pattern, and the time thus Their scores on tests covering the determined is regarded as its thres- listening selections indicated that hold of identification. A possible the renditions of skilled oral read- difficulty with this method for de- ers were more comprehensible than the termining thresholds is that it only renditions of unskilled oral readers, minimally realizes an essential con- and at three of the four word rates dition of cutaneous stimulation, a the data suggested that the rendi- condition that is satisfactorily met tions of male oral readers were more when braille is read in the conven- comprehensible than the renditions tional manner. The excitation of of female oral readers. However, cutaneous receptors requires movement this suggestion was not confirmed by of the tissue in which they are em- an analysis of the variance of test bedded. When a subject is stimulated scores. A full report of this ex- by the dot patterns displayed on a periment has been prepared and sub- tachistotactometer, tissue is moved mitted for publication to an appro- by the dots used to form patterns as priate professional journal. they rise above the display surface, and again as they recede below that surface, but not while they are sta- ACQUIRING INFORMA- tionary during the interval of ex- TION BY TOUCH posure. On the other hand, when braille is read in the conventional Investigation of the ability to manner, there is continuous movement gather information by touch is a ma- of the cutaneous tissue in contact jor activity of the laboratory. In- with the page as the fingertip passes cluded in this category are experi- over the dot patterns in the line of ments concerning man's perceptual writing, and hence there is continu- capacity with respect to dot patterns ous excitation of cutaneous receptors. of the sort represented in the braille Thresholds of identification obtained code, experiments to evaluate various under stimulation that more closely schemes for expanding the braille approximates the conditions of stimu- code, experiments concerning factors lation realized during the reading of that define the effectiveness of tact- braille may prove to be more useful ographic displays, and experiments in elaborating an account of the per- concerning the haptic perception of ceptual basis for braille reading than form. Experiments in visual percep- thresholds based upon the relatively tion are also occasionally conducted static stimulation provided by the in order to acquire data for compara- tachistotactometer tive purposes In collaboration with John Kil- patrick, a graduate student who has Identification Thresholds conducted research in the laboratory, for Moving Braille Characters and with the cooperation of Fred Gis-

soni , Director of the Kentucky Indus- In an effort to understand the tries and Rehabilitation Center for perceptual basis for braille reading, the Blind, who arranged the schedules Nolan and Kederis have determined of clients at the center so that they thresholds of identification for the could volunteer as subjects, an

173 experiment was conducted in which of 500, 333, 250, 200, 167, 143, 125, braille readers were required to 111, 100, 91, 83, 77, and 71 words- identify braille characters and words, per-minute. Braille reading rates embossed on a paper tape that passed were also determined for the 10 sub- beneath their fingertips. This tape jects who served in the experiment. is handled by a special tape transport, designed and constructed in the The outcome of this experiment laboratory for this purpose. The tape was rather complicated, and no effort is transferred from a supply reel, will be made here to report a thorough across a display surface, and onto a presentation and analysis of the data. take-up reel. The subject under ex- However, one of our findings is par- amination places his fingertip on the ticularly deserving of comment. The moving tape, and is stimulated by the finding of Nolan and Kederis that the dot patterns embossed on that tape as time needed to identify words is fre- they pass beneath his fingertip. Tape quently greater than the sum of the speed is well regulated, and can be times needed to identify the letters varied through a wide range. with which that word is composed suggests that the reading of braille de- Thresholds of identification were pends upon serial perception. The determined for braille characters, and braille reader must identify each of for words composed with those charac- the characters in a word, and then ters. Thresholds were determined for integrate these perceptions to achieve words because Nolan and Kederis had a word percept. In the present case, found that, in many cases, the time comparison of the times needed for the needed for identification of a word identification of words with the sums exceeds the sum of the identification of the times needed for the identi- times for the characters of which fication of the characters with which that word is composed, and we wanted those words are composed indicated to know whether or not this finding that, in many instances, whole words would be replicated, using our method were identified in significantly less of determining thresholds. To deter- time than the siams of the identifica- mine thresholds for braille charac- tion times for the characters in those ters, the characters of interest are words. This result may be inconsis- embossed on the paper page. Each tent with the result reported by Nolan character is separated from each other and Kederis, although further experi- character by an amount of tape that mentation and analysis will be re- will result in a satisfactory inter- quired to clarify the issue. Our character interval when the tape is result was observed in the perform- transported at the speeds to be em- ance of the faster braille readers in ployed in the experiment. The trans- the experiment. It is reasonable to port is initially adjusted for a tape assume that faster braille readers speed that is too fast to permit iden- find reading a more rewarding ex- tification of any of the characters perience and that they have , as a under test. As the examination pro- consequence, done a good deal more ceeds, tape speed is gradually rereading. Their greater backgro\ind of duced until all of the dot patterns experience in reading may have enabled have been identified. The tape speed them to predict some of the letters in at which a dot pattern is first iden- words with greater success than the tified is taken as an index of the slow readers. To the extent that suc- threshold of identification for that cessfully predicted letters do not character. have to be identified, the words in which they occur may be identified In the present case, thresholds more quickly. However, there is an- of identification were first deter- other possible explanation for our mined for 35 of the characters in the result. With experience, some braille braille code. Then, the procedure readers may learn a kind of perceptu- was repeated, using 36 familiar words al "chunking" that permits them, in

composed with these characters. Char- many cases , to treat whole words as acters were presented to subjects at unitary patterns. This perceptual tape speeds that were calculated to ability should significantly reduce produce exposure times of 0.02, 0.03, the time needed for the identifica-

0.04, . . . 0.14 of a second. Using tion of chunked words, with rapid these tape speeds for the presenta- braille reading as a consequence. tion of words resulted in word rates This is an important line of research

174 to pursue, because if by chunking exceptionally fast braille readers, some braille readers can read faster and to formulate and test hypotheses than others because they are able to concerning the perceptual process im- reduce the number of patterns they plied by this behavior. must identify, and if a way can be found to teach this perceptual ability Last year's report contained the to others, it may be possible to bring description of a plan for the inves- about a general improvement in the tigation of rapid braille reading. rate at which braille is read. This plan called for the identification of exceptionally fast braille readers for intensive examination. The Reading Behavior of A variety of measures that might be Exceptional Braille Readers related to braille reading ability, such as intelligence, cutaneous acu- Much of the research concerning ity, reaction time to cutaneous stim- the perceptual basis for the reading uli, age at which braille was learned, of braille has had as its objective and so forth, would be obtained. In the elucidation of the processes order to proceed with this investiga- xinderlying typical braille reading tion, it would be necessary to develop behavior. The process depicted by a means of closely observing the be- this research is one in which letter havior of braille readers, and the percepts, acquired serially, are in- design of an instriament intended to tegrated to form word percepts. Re- enable such observation was described. liance on this process results in the This instrument was constructed, but slow reading rate (in the neighbor- its functioning was marginal, and it hood of 100 words-per-minute for could not be made to function satis- adult braille readers) that is typic- factorily with the adjustments per- ally observed. However, there are a mitted by its design. few braille readers who read so rapidly (200 words-per-minute or faster) With the assistance of Stewart that their reading performance is Cooper, a photographer on the staff not convincingly explained by a proc- of the Department of Ophthalmology, ess involving the integration of seri- the construction of a redesigned in- ally perceived letters. In the course strument is now nearly accomplished. of collecting data for the experiment As before, braille, written on a described under the preceding heading, transparent sheet of plastic, is two subjects with exceptionally fast placed on a plate glass surface. An reading rates were encountered. In electric stopclock is also placed, the phase of that experiment in which face down, on this surface. A large, whole words were presented on the mov- first surface mirror is positioned ing tape, one of these s\±>jects made beneath the plate glass display sur- accurate identifications at the fast- face at an angle of approximately 45°. est tape speed employed in the experi- A Super-8 motion picture camera is so ment. At the slower tape speeds, her positioned that it views the image identification times increased, and reflected from the surface of the she began to make errors. She report- mirror, and the field of view is iled spontaneously that, at the slower luminated by two 30 0-watt photoflood tape speeds, word patterns began to lamps. To minimize problems that disintegrate, and she was compelled might be caused by glare, the lens to fall back on a strategy of identi- of the camera is fitted with a polar- fying letters one at a time. The izing filter. report of this subject, together with her reading performance, suggests that With this instrument, the hands patterns consisting of whole words may of a reader can be photographed as he frequently be the perceptual units of reads. The res ult is a continuous the process in which she engages as film record of the reading behavior of she reads. the subject and of the time registered on the face of the timer. The hope is Observation of this braille read- that a frame by frame analysis of this er has revived an earlier interest of film will permi t an accurate and de- mine in an approach to the study of tailed analysis of the different be- braille reading in which an effort is haviors employe d by the braille read- made to obtain a detailed and accurate er, and of the distribution of these description of the behavior of behaviors in th e time dimension. If

175 this approach is fruitful, it will Kederis determined the thresholds of lead to an understanding of the per- identification for some of the char- ceptual processes upon which rapid acters in the braille code. To do braille reading depends. With this this, they employed a tachistotactom- understanding, it may be possible to eter, an instrument that provides for formulate training procedures that temporal control of the intervals dur- will make these processes available ing which dot patterns are made avail- to slow braille readers. able for tactual examination. The dot patterns to be examined are embossed on a sheet of paper mounted on Increasing the Brai lie a platform beneath the display sur- Reading Rate by face. This paper is coated with Presenting Braille plastic, and it will support dots of Characters at greater height than the dots which Controlled Rates can be embossed on conventional braille paper. In order to present a pattern, Last year's report included the the platform is elevated by solenoids, description of a plan to explore the and the dots embossed on the sheet possibility of increasing the braille mounted on it are pressed through

reading rate by giving subjects prac- holes in the display surface , a thin tice in reading connected prose, pre- membrane of tightly stretched metal sented on moving tape at a speed that with holes in it at locations corres- is gradually increased as practice ponding to the dot positions in proceeds . According to a current braille cells. The dots embossed on hypothesis, there is a kind of dynamic the sheet beneath the display surface patterning involving whole words and protrude above that surface to the

phrases that is realized when the mo- , height of braille dots when the plat- tion between the fingertips of the form is elevated, and recede again braille reader and the line of writing when the platform is returned to its is maintained at a fast and constant resting position. Up to 30 charac- rate. The ability to identify such ters, or an entire line of braille

patterns enables the exceptional : writing, can be displayed at one time. braille reader to read braille at a The interval during which dot patterns rate that greatly exceeds the typical are available for examination is con- braille reading rate, and that com- trolled by an interval timer, con- pares favorably with the silent visual nected in series with the power sup- reading rate. If patterning of the ply of the instrxmient. sort required by this hypothesis is a reality, the ability to control the Nolan and Kederis defined thres- rate at which braille characters move hold as the minimum time of exposure across the fingertips employed for needed for the identification of a reading may make it possible to estab- character, and used a modified method lish the conditions under which read- of limits to estimate thresholds. Be- ers can learn to become aware of and cause of the time consumed in changing to interpret this patterning. the sheets on which characters are embossed for presentation by the tachis-

The tape transport used in deter- totactometer , they found it expedient mining thresholds of identification to emboss several characters on a for moving braille characters can be sheet, and to deteimiine thresholds used in this study as well. Because for those characters before replacing of a shortage of research personnel, it with a new sheet. Since a si±iject this study has not yet been initiated, could quickly memorize the characters but it will probably be conducted at embossed on a single sheet, he could the Kentucky Industries and Rehabili- eliminate most of the characters in tation Center for the Blind, where the code from his consideration, and clients of the center will be paid to thereby greatly reduce his uncertainty serve as subjects in the experiment. regarding the identity of any character presented to him for examination. To solve this problem, it was neces- Identification Thresholds and sary to eliminate those series of Identification Times for the stimulus administrations ordinarily Characters in the Braille Code employed in the method of limits in initially above In their investigation of the which the stimuli are and to retain only those perception of braille, Nolan and threshold,

176 series of stimulus administrations braille code can be brought under in which the stimuli are initially examination at one time. below threshold. Estimates of threshold obtained by this attenuated pro- The new instrument can also be cedure are apt to be unduly influ- used to determine the times needed enced by errors of habituation. for the identification of characters in the braille code. In this mode Due to design limitations of of operation, the experimenter se- the tachistotactometer , stimulus lects a character for presentation presentations could not be made brief by operating a keyboard. Pressing a enough to permit the administration key switch on the appropriate sole- of complete series of stimuli. The noids for the dot pattern controlled mass of the platform that had to be by that key, and their circuits are moved from one position to another completed by a final switch closure, and back again in order to present which operates a relay, connected in a stimulus , was so great that its the common return of the power sup- movement could not be accomplished ply. When this relay closes, it also in the time allowed for very brief starts an electric stopclock. The presentations , and for some of the relay latches when it closes, so that characters in the braille code, it the character continues to be avail- was not possible to make the time of able for examination after its key exposure brief enough to preclude is released. When the subject makes their identification. Consequently, a vocal identification of the charac- thresholds for these characters could ter under examination, he generates not be determined. a microphone signal, which is ampli- fied, rectified, and used to operate More precise estimates of thres- a second relay, connected in the holds of identification of the char- latching circuit of the first relay. acters in the braille code may be When the second relay closes , the useful in further analysis of the first relay opens, the character un- perceptual basis for reading of der examination disappears, and the braille. Consequently, a study has timer. stops. The subject can then been planned that seeks to eliminate be scored for both time and accuracy the sources of error in the study re- of identification. ported by Nolan and Kederis. An instrument has been constructed that The apparatus just described permits the presentation, one at a will be used by Beth Challman, a time, of all of the characters in the member of the laboratory's staff, in braille code. Dot patterns are formed the experiment she reports in her by pins that rise through holes in master's thesis. In this experiment, the display surface. Each pin is she will determine identification driven by its own solenoid. The in- thresholds and identification times terval during which a character is for the characters in the braille available for examination is deter- code. It will, of course, be useful mined by an interval timer, connected to have more precise estimates of in the common return of the power the thresholds of identification for supply that energizes the solenoids. the characters in the braille code, Since the mass of the pins that must and the availability of identifica- be moved from one position to another tion thresholds and identification and back again in order to present a times will permit their comparative dot pattern for an interval of con- evaluation as predictors of such as- trolled duration is smaller than the pects of braille reading performance mass of the platform used for the same as the identification times for words, purpose on the tachistotactometer, and and the braille reading rate. since the distance traveled by pins is much less than the distance traveled by that platform, much briefer stimu- The Necessary Conditions of lus presentations are possible with Stimulation for the Efficient the new instrument, and stimulus val- Reading of Braille ues below threshold can easily be produced. Since any character in the It has already been pointed out braille code is immediately available in an earlier section of this report for presentation on this instrument, that movement of the tissue in which the entire set of characters in the cutaneous receptors are embedded is

177 necessary' for their excitation, and Development of a that more of this movement is pro- Braille Page Embosser duced when braille is read in the conventional manner than when braille Efforts are now underway at many characters are presented by pressing locations in this country and else- patterns of dots into a stationary where to take advantage of the rapid fingertip. In order to explore this expansion of computer technology in variable more thoroughly, an experi- implementing services for blind per-

ment is planned in which braille sons . These services could include reading rates will be determined un- the production of hard copy in braille der several conditions of stimula- at the user's terminal, but the tion. In the first condition, braille braille page embossers currently characters will be displayed, one at available that would be suitable for a time, on the device used for deter- use as terminal devices are so expen- mining identification thresholds and sive that, in most cases, individual identification times, described in ownership is not feasible. To solve the preceding section. The device this problem, an effort has been will be controlled by a punched tape undertaken in collaboration with reader. The rate at which this tape Electronic Systems Development Cor- reader is stepped determines the rate poration to develop a braille page at which characters are presented. embosser that will operate at tele- In the next condition , the tape type speeds and that can be manufac- reader will be used to operate a tured at a low enough price (under device in which characters are formed $1,000) to make individual ownership by patterns of vibrating pins. These feasible. patterns will be presented one at a time to a stationary fingertip, but In last year's report, I indi- the increased movement of tissue pro- cated that a first generation proto- duced by their vibration may result type of the embossing mechanism had in an increased supply of receptor been constructed. Since then, with excitation from which the reader can funds provided by the Grant Founda- acquire the information that speci- tion, information gained by experi- fies characters. The design and con- ence with this prototype has been struction of this instrument has been used in constructing a second gen- undertaken by Electronic Systems De- eration prototype. This mechanism velopment Corporation, a biomedical appears to be functioning satisfac- engineering firm in the Louisville torily, and the carriage needed to area, and it should be delivered to transport it is now being designed. us before long. In the third condi- We hope that a working model of the tion, the paper tape transport de- embosser will be ready for test in

scribed in an earlier section will a few months . Experience with this be used to present a continuous line model should provide the information of braille characters that moves needed to complete the design of a across the stationary fingertips of model suitable for production. How- the reader. There should be almost ever, before this final phase can be as much movement under this condition carried out, it will be necessary to of stimulation as under the normal find additional funding. condition of stimulation. Of course, the reader of a moving tape cannot preview or review as effectively as THE PERCEPTUAL BASIS the reader of a braille page. In the FOR MOBILITY final condition, the reading rate for braille presented in the conventional manner will be determined. The data The perceptual basis for the obtained with this experiment may skill that enables the blind pedes- lead to a better understanding of trian to reach his objective inde- the necessary conditions for effec- pendently, safely, comfortably, and tive cutaneous communication. Furth- gracefully, is one of the labora- ermore, the information provided by tory's continuing research interests, the experiment may be of use to de- Experiments have been conducted to signers in reaching decisions about evaluate the contributions of the the types of displays that are suit- physical characteristics of the cane able for various applications. employed by the blind pedestrian to

178 its effectiveness for gathering in- permits the user of the cane to ex- formation. Techniques for the produc- plore surface discontinuities more tion of tangible maps have been stud- effectively. The tip is formed from ied because tangible maps constitute nylon rod. The rod is bent to form cin important source for the informa- an elbow, and it is this elbow which tion needed by the blind pedestrian contacts the surface that is examined for orientation and mobility. Al- by the blind pedestrian as he walks. though a program of research concern- The portion of the rod extending being orientation and mobility has been yond the elbow prevents the tip from formulated for some time, active re- becoming lodged in cracks or other- search in this area has been limited wise impeded when abrupt surface dis- by the lack of funds. However, with continuities are encountered. the general laboratory support provided by the Grant Foundation, it has Clark Shingledecker, a graduate been possible to initiate programmatic research assistant employed in the research, and Clark Shingledecker , a laboratory, is currently conducting graduate student employed as a Re- an evaluation of this cane. As a search Assistant in the laboratory, preliminary step, canes embodying intends to make the investigation of the features just described have been orientation and mobility the basis given to experienced blind pedestrians for his doctoral dissertation. in the Louisville community. When these critics have had the opportunity to use their canes for some time, A Rudimentary Theory they will be interviewed by Shingle- of Mobility decker, and their criticisms will be taken into account in refining the A rudimentary theory of mobility design of the cane. has been formulated in an effort to account for the behavior of the blind With the assistance of David pedestrian. This theory seeks to Stegner, the mobility instructor at identify the operations upon which the Kentucky School for the Blind; the blind pedestrian depends in gath- Wanda Biggs, a graduate student in ering the information he needs to the Orientation and Mobility Program pursue a course of travel success- at Western Michigan University; Dr. fully. Although the theory has not Richard Smith, a member of the Depart- yet received much experimental ex- ment of Psychology at the University amination, in its current state of of Louisville who has experience in development, it provides guidelines the analysis of performance; and my- for a systematic research effort. self, Shingledecker has constructed a questionnaire designed to elicit information about the advantages and The Development of an disadvantages of canes, and about Improved Cane for Use how they might be imoroved to make by the Blind Pedestrian them more serviceable. This questionnaire will be used by Shingledecker Earlier research conducted in as an interview guide, when he inter- this laboratory has indicated the views a large number of blind pedes- relevance of weight, length, rigidity, trians exhibiting varying degrees of and form of tip as factors affecting proficiency in the use of a cane. the efficiency of the cane employed by the blind pedestrian to gather in- The information obtained by these formation. These factors have been interviews , and the information ob- taken into account in constructing a tained by consulting the experience cane that appears to be more informa- of blind pedestrians who have used tive than canes currently in use. the preliminary model of the cane, The aluminiam tubing selected for will guide the construction of a construction of the new cane makes a new generation of canes. The re- shaft that is a little lighter and designed canes will be subjected to more rigid than the shafts of canes more formal evaluation in field tests now in general use. It is equipped in which an effort will be made to with a grip that fits comfortably in develop measures of their effective- the hand and that permits the cane to ness , and to compare them with con- be held securely so that it can be ventional canes in terms of measured manipulated with greater precision. effectiveness. If the new canes sur- A new tip has been designed that vive these tests, arrangements will

179 be made for tJieir production and dis- A Communication Device tribution. Charles Cox, Director of for Children with Kentucky Industries for the Blind, Profound Motor Disability has expressed an interest in evaluating the feasibility of their pro- Last year's report included a duction in the workshop operated by description of a device that can be his agency. operated with the minimal movements remaining to persons with profovind motor disability. This device en- The Preservation of Cues ables the selective backlighting of to Distance and Direction 64 transparencies, arranged in eight in Tape Recordings Used columns and eight rows. Selection of for Mobility Training the transparency to be backlighted is accomplished by the operation of two In last year's report, there was switches. Repeated operation of one a description of a method for record- switch causes the backlighting to be ing environmental sounds that permits transferred from column to column their reproduction with startling re- across the display, while repeated alism. The reproduced effects are so operation of the other switch causes natural that it is difficult for the the backlighting to be transferred observer to believe that the sounds from row to row down the display. he hears are emanating from the ear- The switches are designed for opera- phones he is wearing. They appear to tion by poorly controlled arm move- be definitely localized in the space ments, because a girl at the Cerebral surrounding him. Palsy School in Louisville, who might benefit by this device, is capable of The vividness of this experience such movements. However, the switch- suggested a possible role for tape es can be designed for operation by recordings of this sort in the any pair of independent movements, courses of instruction in orientation regardless of the parts of the body and mobility that are provided for involved, or even by the myoelectric

blind persons . Some progress was potentials that are generated when made in constructing the apparatus muscles are contracted.

needed for such recordings, but the • graduate research assistant who was A child unable to speak because working on the project left the uni- of motor impairment could use the de- versity, and it has been necessary vice for simple communication, by to suspend work on this project backlighting a transparency with some temporarily. However, the project bearing on a current need. A picture will be resumed if a research as- of a glass of water might suggest to sistant can be found who has the others that the operator was thirsty. time and interest to pursue this A picture of a bed might signal the line of research. desire for sleep, and so forth.

In addition, the device might OTHER PERCEPTUAL enable steps leading to the ability ALTERNATIVES to read, by prompting an improved understanding of the relationship between words and the objects for Though the three areas of re- which they stand. Training with search on auditory perception, tactual the device might proceed in the fol- and haptic perception, and the percep- lowing manner. The student first tion on which mobility depends have learns to select a variety of objects received major emphasis in the lab- by operating the switches that con- oratory's research program, other trol backlighting. Words naming perceptual alternatives have also re- those objects are then added to the ceived some attention. Accounts of transparencies used to depict the these activities follow. objects. After some experience with objects and their names, the student may discover that it is possible to achieve the same communicative outcome by selecting a transparency which displays only a name. He may

180 then learn to compose more elaborate that might result from sexual mis- messages by displaying sequences of conduct by carefully supervised seg- words, and so on until he is engaging regation. As a result, the blind in reading behavior. child who spends much of his life in a residential school missed the op- Initial experience with the de- portunities to learn about sexual vice was quite encouraging. Children differences that are ordinarily af- seemed to enjoy using it, perhaps be- forded by growing up in a family with cause it enabled them to exercise brothers and sisters. Of course, the some control over their environment-- blind child becomes aware of the em- to make something happen. However, phasis on sexuality in our culture, the attempt to use the device for in- and this combination of curiosity and structional purposes revealed short- ignorance frequently leads to the comings. The prolonged switch clo- formation of bizarre theories con- sures produced by children with poor cerning sexual differences. muscular control caused damage to the electronics. The inability to trans- In recent years , there has been fer backlighting from right to left a growing conviction on the part of as well as left to right, and up as many educators that blind children, well as down, proved to be a serious more than most, need adequate sex inconvenience. Also, since this de- education. They argue that the vice was only a laboratory model, it blind child's lack of opportunity was much larger and heavier than nec- to acquire the experience that in- essary. However, our experience with forms his sighted peer, and the mis- the device has taught us how to im- conceptions resulting from this lack prove it. The grant awarded by the of experience, predispose him to Grant Foundation made specific provi- sexual maladjustment in adult life. sion for the support of further work On the strength of this argument, on this project and Electronic Sys- attempts have been made to provide tems Development Corporation has sex education in many residential nearly completed the construction of schools, although no satisfactory an improved prototype. If the new way has been found to show the blind device appears to be satisfactory, child what the sighted child can see. several of the devices will be built Although a convincing case can be and limited field testing will be ini- made for the provision of sex educa- tiated. tion for blind children, its validity depends, for the most part, on anec- dotal evidence. There has been none The Blind Child's Knowledge of the systematic observation that of the Anatomy and Functions would permit an accurate description of Sex of the state of ignorance of blind children concerning the anatomy and In the course of growing up, the functions of sex, and no search for child who sees has many opportunities the relationship between that ignor- to learn about sexual differences. ance and the consequences that pre- He continually observes differences sumably ensue. in body conformation, and unintended exposures give him occasional glimpses Two years ago, Thomas Uhde , a of normally concealed parts of the medical student at the University of body. Furthermore, anatomical dif- Louisville, visited the laboratory to ferences are explicitly shown in pic- learn about its program of research tures to which the sighted child has and to explore the possibility of easy access. participating in a research project involving blind children. He had The blind child may have little been working part-time at the Ken- opportunity to discover such differ- tucky School for the Blind, and his ences . He can explore his own body, experience included some counseling but he cannot s ee other bodies , and of adolescent boys. During his con- learning about them by touch is pro- versations with these boys, he had hibited. Many blind children receive occasionally encountered some curious their education in residential schools, deficits in their sex knowledge. He operated by the states in which they discussed his findings with a member live. In these schools, there has of the school faculty, who told him been a tendency to avoid the problem of the need for sex education in the

181 , school, and of the school's interest Computer Services in formulating a program of sex edu- for the Blind cation. The curiosity aroused in Uhde by his experience at the school, The ARTS system (ARTS is an together with the school's current acronym standing for Audio Response interest in sex education, seemed to Time Sharing) was conceived by Dr. provide a climate that would be con- Kenneth Ingham, and he has brought genial to an investigation of the the system to its present state of blind child's knowledge of the anatomy development. The ARTS system is a and functions of sex, and of the atti- configuration of computing machinery tudes of parents and educators concern- that is programmed to provide a va- ing sex education for blind children. riety of services that should be use- Accordingly, an investigation was ful to blind persons. These services planned that calls for Uhde to inter- are delivered by telephone, and since view blind adolescents, of both sex- it is a time sharing system, 16 users es , in order to discover what they can be accommodated at one time. The know about sexual anatomy and the input to the system is provided by a manner of its involvement in sexual typewriter-like keyboard, connected behavior, and to obtain expressions to the user's telephone. The output of their attitudes concerning sexual of the system is computer-compiled behavior and its regulation. The speech, heard over a loudspeaker, plan of the investigation also calls also connected to the user's tele- for the preparation of questionnaires phone. Services planned for the sys- to be administered to the parents tem include dictionary consultation, and the house parents and teachers composition of letters and other of blind children. These question- manuscripts, bookkeeping, filing, naires are designed to elicit ex- mathematical computation, transcrip- pressions of attitudes concerning sex tion from print to braille, computer- education, and opinions concerning aided instruction, computer program- the method and content that would be ming, and so forth. appropriate in a course of sex education designed for blind children. My report for the year before last included an account of the events A grant of $750 was obtained I initiated that led to the establish- from the Rauch Foundation to provide ment, by an action of the Kentucky a summer salary for Uhde so that he State Legislature, of Computer Ser- could work on this project. At this vices for the Blind (CSB) , a non- writing, he has completed over half profit public corporation charged of the interviews. With assistance with implementation of the ARTS sys- from members of the laboratory staff tem and with the development of other and members of the faculty at the computer-generated services for the Kentucky School for the Blind, he has blind citizens of Kentucky. The constructed a questionnaire, and has State provided $173,000 to meet CSB's sent over 2,000 copies to parents, costs during the first two years of schools, and other agencies. Ap- its operation. proximately 500 of these questionnaires have been returned, to date. The transactions of CSB are su- A preliminary report of the project pervised by a Board that includes the on which Uhde is working was published Director of the Division for the in the May issue of The New Outlook Blind, Bureau of Rehabilitation Ser- for the Blind. Mr. Uhde plans to vices; the Superintendent of the Ken- spend this summer completing the tucky School for the Blind; the Di- interviews and analyzing the ques- rector of Industries for the Blind; tionnaires that have been returned. and six members appointed by the However, efforts to find a source Governor of the State. I was one of for his summer salary have so far those appointed by the Governor, and not been successful. Since he needs I am currently serving as Chairman of an income to meet living expenses , he the Board. The Board appointed Glenn may have to seek other employment, Smith as the first Director of CSB, and in this case, it would be neces- and he served in that capacity xantil sary to suspend work on the project. recently. The University of Louis- ville has made available to CSB, without charge, the space it needs for computing machinery for the administration of its business.

182 .

Initially, we believed that the support was rejected, and we will development of the ARTS system was have to find einother source of sup- substantially complete. We expected port in order to continue the project to purchase a system which , after when current funds are expended. In correcting the difficulties routinely spite of the difficulties we have en- encountered during the installation countered, we remain convinced that of a complex system, could be put im- the ARTS system can provide the ser- mediately into service. However, as vices originally claimed for it, but our attempt at implementation pro- we now know that a developmental ef- ceeded, it became apparent that con- fort of considerable magnitude will siderable development of both hard- be required before full implementa- ware and software was still needed. tion can be achieved. The time re- As a result, we were unable to adhere quired for this effort will depend to the schedule set forth in our pro- upon the level of support we are able posal to the State of Kentucky. to obtain, but with adequate funding, it can be accomplished within three By the beginning of the current years fiscal year, the system was supposed to be delivering services and earning A first draft of a proposal has income to defray operating expenses. been written, and we are now search- Instead, it is not yet operational. ing for a funding agency that might The principal hardware components of be receptive to such a proposal. In the system have been delivered and the meantime, with remaining project

interconnected. However, the system, funds , CSB plans to use the applica- as it stands , does not have enough tions computer that is a part of the capacity to meet the anticipated ARTS system for the transcription of needs for the storage of user files. print to braille and, if time and The terminals provided with the sys- money permit, for the generation of tem are only minimally adequate , and tangible graphs and maps. In our a source for better terminals has not judgment, both types of material are yet been identified. The quality of needed by the blind persons served the speech produced by the system is by CSB, and with the addition of the not quite satisfactory, and the sys- appropriate accessories to the basic tem's vocabulary has not been se- system already owned by CSB, we can lected with adequate care. Although easily develop the capability to pro- some softv^are has been created, a ma- duce these materials. jor developmental effort is still needed to provide the software that will be required for full implementa- The Transcription of tion of the services that were ex- Print to Braille pected when the system was purchased. The system was delivered without ade- The NOVA 800 computer that is a quate documentation. CSB refused to component of our ARTS system is ade- accept the system without such docu- quate for the recording that is re- mentation, and many months have been quired for the transcription of print spent in acquiring it. A final ac- to braille. The print that is to be ceptance procedure has been negoti- transcribed into braille is typed on ated, and will be carried out shortly. a typewriter that creates a punched If the vendor of the system can meet paper tape. This tape provides in- the tests specified by this procedure, put to the computer. The output of responsibility for system maintenance the computer is used to operate an will be transferred to CSB. automatic braille page embosser. The system can be made more flexible if It is now clear to us that what provision is made for storing the we originally regarded as a service output of the computer so that it is project will have to be regarded not necessary to operate the page em- henceforth as a research and develop- bosser on-line, and so that the out- ment project. The State of Kentucky put of the computer can be preserved funded our efforts in the expectation for future use. In the system we that services would be delivered. We have planned, teletypewriters will have failed to do so, and the State be used to punch the paper tapes that of Kentucky is not interested in provide input. We have one teletype- funding a research and development writer, and another one is on order. effort. As a consequence, our ap- We have also ordered the automatic plication for continuation of project braille page embosser manufactured

183 ,

by Triformations . This machine em- reduction in the time required to fill bosses braille a line at a time, and orders for transcribed braille read- it can emboss 120 lines-per-minute. ing matter. We will be using cassettes to record the output of the computer, and these cassettes, when reproduced on a trans- The Production of port connected to the Triformations Tactographic Displays embosser, will control its operation. At present, there is no adequate In addition to the hardware com- source that can produce tactographic ponents of the system, software will displays of high quality and that can be needed to transcribe the input respond to the needs of individual code to an output code that will cause users. Orientation and mobility in- the Triformations machine to emboss structors are making increasing use Grade 2 braille. CSB will contract of tangible maps, and it is likely with Glenn Smith, the former director that many blind pedestrians would use of CSB, for the creation of the re- them regularly, if they could obtain coding program that is required. them. Blind students could use raised line drawings of the graphs and fig- The system we envision is one ures in printed books to their advan- that will depend heavily upon volun- tage, but there are often no provi- teer services. A system manager will sions for making these drawings avail- insure that the computing machinery able. The techniques currently used is functioning properly, and will as- for the production of tangible dis- sist volunteers in the use of that plays make heavy demands on patience, machinery. However, volunteers will skill, and time, and they are limited prepare input tapes, operate the with respect to the variety of sym- braille page embosser, and bind bols that can be displayed with ease. brailled pages into volumes. Because we are using a minicomputer, our prod- Dr. John Gill, a researcher at uct will not be perfect. There will Warwick University, has developed a be occasional violations of the rules method for the construction of tacto- governing the use of contractions in graphic displays that appears to have the Grade 2 braille code. However, many advantages over methods now in the occasional departures from per- use, and with the addition of acces- fect form will not have serious con- sories to the computing machinery al- sequences for the reader. In fact, ready owned by CSB, this system can it is likely that many of them will be implemented and used to produce be unnoticed by the average reader, the tangible graphs and maps requested who does not subject his copy to the by the blind persons served by CSB. scrutiny of a proof reader. Our primary objective is rapid response. Original composition of the tangi- We intend to satisfy the ongoing ble display is done at a computer needs of students, and of blind per- terminal that includes a visual dis- sons engaged in professions and oc- play unit, an analog control of the cupations. Under present management, joy stick type, and a teletypewriter. there are well developed facilities The information required for the for the volume production of braille, specification of symbols is stored but there is inadequate provision for in the computer. The supply of sym- responding to individual needs. A bols includes symbols for points blind person may request a group of such as dots, filled and unfilled volunteer transcribers to prepare a circles, filled and unfilled tri- book for his use, but with the meth- angles, stars, and so forth; symbols ods they currently employ, he will for lines, such as solid lines, have to wait months for the delivery dashed lines, broad and narrow lines, of that book. It is frequently the and so forth; and symbols for area experience of the person who needs a consisting of a variety of surface book to meet educational or occupa- textures. Teletype commands cause tional requirements that, by the time these symbols to be displayed on the the book can be delivered to him, the visual display unit, where they can need for it has passed. If the sys- be positioned by operating the joy tem contemplated by CSB functions stick control. Dot patterns corres- properly, it should be possible to ponding to the dot patterns in the bring about a very significant braille code can also be added to

184 the display by teletype commands, and interest. These services include positioned by the joy stick control. the dissemination of information, When a satisfactory graphic display the preparation of research materials, has been composed in this manner, the the development of equipment, and information that is needed to specify consultation with educators concern- the composition of the display is en- ing the utilization of perceptual al-

coded on punched paper tape. This ternatives . tape is used to operate a numerically controlled engraving machine that en- The Center for Rate-Controlled graves a negative of the graphic dis- Recordings is a unit of the Percep- play on a sheet of plastic. From tual Alternatives Laboratory. It this negati\'e, a rubber positive is was established to provide a source made that serves as a master for use for time-compressed and expanded on a vacuum forming machine with recorded speech of high quality, and which duplicates are made on plastic at a moderate cost, for use in re- sheets. The advantages of this sys- search and education, and to dis- tem include the relative ease with seminate information concerning rate- which graphis displays can be com- controlled recorded speech. During posed, the nearly inexhaustible sup- the current year, the Center has con-

ply of symbols for points , lines , and tinued to meet a steady demand for areas to which the composer has ac- rate-controlled recorded speech and cess, and the remarkably high quality has continued its publication of the

of the end result. The use of an CRCR Newsletter . Many requests for automatic engraver provides better information have been met through control over the formation of symbols correspondence, telephone consulta- than has heretofore been possible, tion, and consulting visits. and the reduced variability among symbols of the same type permits a The most serious impediment to more reliable and more finely graded the application of recorded speech differentiation among symbols of dif- that has been compressed or expanded ferent types. in time has been the lack of suitable equipment at a moderate price. If CSB implements this system, However, efforts to develop moder- it will be necessary to purchase the ately priced equipment have been visual display unit and the numeric- underway at several locations. ally controlled engraving machine. Several compressors have become com- The teletype machine that is needed mercially available in the last three

is already owned by CSB, and a vacuum years , and other compressors now un- forming machine is reliably accessible. der development will be available Although we regard the services that soon. Because of the prominent role can be provided by such a system as played by the laboratory in the de- potentially quite valuable, our lim- velopment of time-compressed and ex- ited resources have made it necessary panded speech, four manufacturers for us to establish priorities, and have donated compressors to the lab- we have given first priority to the oratory, and others have indicated implementation of the system for an intention to do so shortly. The transcribing print to braille. We laboratory is now able to provide will not make the decision to proceed speech that has been compressed or with the implementation of the system expanded in time by a computer as just described until we have made well. It is becoming a museum of enough progress in implementing the contemporary compression technology, transcription system to be confident and this should result in a further of our ability to complete its im- expansion of its role in the dissemina- plementation. tion of information about time- compressed and expanded speech.

SERVICES PROVroED BY THE LABORATORY Audio-Tutorial Instruction In audio-tutorial instruction, The laboratory's principal busi- the objective is to make available to ness is the conduct of research. How- groups of students the advantages ever, a continuing effort has also been ordinarily associated with individual made to provide useful services to edu- tutoring. This is accomplished by cators and researchers in its field of recording on tape the information

185 that would ordinarily be presented Discussions have been held with by a lecturer to a group of students, members of the faculty at the Ken- or by a tutor to an individual stu- tucky School for the Blind, in order dent. Careful attention is given to to acquaint them with the possibili- the organization of content and, in ties offered by audio-tutorial in- addition to the information that is struction. An offer has been made given, the recording gives direc- to assist the school in the training tions for learning activities in of teachers and the preparation of which the student is to engage. The audio-tutorial instructional materi- recording is accompanied by a set of als. Will Evans, who has recently written learning objectives, which assumed the superintendency of the specify the student behavior that Kentucky School for the Blind, has will constitute evidence of mastery. expressed a strong interest in the Mastery of each unit is required use of audio-tutorial instruction, before the student is allowed to and I am currently working with him proceed to the next unit. Each stu- toward the preparation of an applica- dent progresses through the course tion for the support that will be at his own rate, and every student needed to equip an audio-tutorial who completes the course is assured learning center, to train teachers of a passing grade. in the production of audio-tutorial instructional materials, and to Because of the laboratory's prepare the materials needed for a recording facilities, it is ideally trial course. suited for the preparation of audio- tutorial instructional materials. Over the past three years, these fa- Consultation cilities have been used to prepare the materials required for the audio- Since its inception, there has tutorial presentation of an intro- been a steadily growing awareness of ductory course in Psychology at the the Perceptual Alternatives Labora- University of Louisville. tory and of its research program. As a consequence, the director of The students in a typical class the laboratory is frequently invited of blind school children show extra- to serve as a consultant to research- ordinary diversity in their readiness ers and educators . During the cur- for the course of instruction that is rent year, the director of the lab- to be covered in the class. They show oratory has continued to serve as a more than usual variability in intel- regular consultant to the Research ligence, prerequisite learning, se- Department of the American Founda- verity of handicapping condition, tion for the Blind; and to the Edu- and so forth. As a result, no single cational Research Department of the instructional experience arranged by American Printing House for the a teacher can be expected to be ef- Blind; and, as in the past, the fective for more than a few class director has continued to fill re- members, and the individualization quests for consultation by State of instruction becomes a necessity. Departments of Education, school Audio-tutorial instruction offers an systems, and so forth. Consultation attractive solution to this problem, has covered the contributions of psy- particularly since blind school chil- chology, education and educational dren have already had considerable technology to the solution of prob- experience in reading by listening. lems engendered by visual and other perceptual impairments.

186 AUDITORY AND TACTUAL DISPLAYS FOR SENSORY AIDS FOR THE VISUALLY IMPAIRED

J. M. Gill*

The increasing use of digital published on the various devices displays, in particular the advent being developed. This survey is of inexpensive electronic calcu- based on replies to a questionnaire lators, has renewed interest in non- circulated in August 19 73; the visual digital displays for the vi- main features of these displays sually handicapped. However, rela- are summarized here. tively little information has been

1. Developer American Foundation for the Blind 15 West 16th Street New York, New York 10011, U.S.A. Application Output for electronic calculators, etc. State of development A prototype has been produced but a new version with braille output is being developed. Description Tactile display using binary coded decimal.

2. Developer L. Andersson Projekt AB Alea, Henriksbergsvagen 104, S-136 67 Handen, Sweden Application Output for electronic calculators, etc. State of development Two prototypes built. Description Each dot consists of a small container filled with paraffin and covered with a flexible "window." A resistor is embedded in the paraffin so that an electric current through the resistor causes the paraffin to melt and expand; the flexible window forms an embossed dot. It takes about three seconds to change state.

Price Circa $50 for 6-digit display. d'

*Department of Engineering, University of Warwick,

Coventry CV4 7AL , England.

187 '

TABLE 1

A Summary of the Main Features of the Displays

State of

Output Mode Output Applicat ion Deve1 lopment

o rH •H s -s .0 (0 I-I • •fi M 0) •H o +J U U] 1-1 o (0 ?* O +J (U (0 3 > u u (1) (U c u -o 10 0] 0) flj >i m H •H •H o S{ iH u i c •i u c H (0 u a a. iH flj 3 9 u M M (0 iH P C m 0) 0) O ^ u CD (0 la 04 H w e ft M-l +J Oi w a •H g i-i >i 0) >-i 3 (C tfl 01 >i >i U -H fi ra (0 H H u (d C £ +J M c Q) lO 4J M X -H rH iH (0 3 fl) u (0 W 3 3 •H >M o (U o a. +J •H (0 -H 0) x; a o o H qj +j M -P U) O -a U U •i e a >i +> H (0 4-> M P Qa U) •H (C (0 (0 3 0) 3 3 3 (1) O-H M o a Q H < ft CO s ^ Si o u « < T) Cm u <; u

1 + + + + 2 + + 6 + + + 22

- 3 + + + ^ . . + 11 4 + + + + +

-^ : 5 + + 25 + ; : + + 300 6 + + 12 + + + 250 7 + + 10 + + + 8 + + 10 + + + 490 9 + + 7 + + + 710 10 + + + + + 29 11 + + + + 12 + + + + 13 + + 8 + + + 14 + + + 4 + + + 140 15 + + + + + -' 16 + + + + + 450 17 + + 6 + •" '• + + + 18 + + 3 + + + 19 + + + + 20 + + + + 21 + + 12 + + + 325 22 + + + + + 430 23 + + + + + + 24 + + 8 + + + 110 25 + + + + + 1530 26 + + 12 + + + 770 27 + + + +

188 .

Developer L. Andersson Projekt AB Alea, Henriksbergsvagen 104, S-136 67

Handen , Sweden

Application Output for electronic calculators, etc.

State of development Proposal only.

Description An auditory output for presenting a half braille- digit at a time. A rotary switch selects the half of the digit which consists of two dots. The upper

one is presented as a high note , the lower one as a low note and both as a medium note.

Price Circa $24. •••-. r~

Developer Privat-Doz. Dr. H. Bebie, Institut fur theoretische Physik der Universitat, Sidlerstr. 5, CH 3012 Bern, Switzerland and Prof. Dr. F. Fankhauser, Universi- tats-Augenklinik, Inselspital, CH 3010 Bern, Switz- erland.

Application Output for electronic calculators, etc.

State of development Prototype built.

Description An auditory display which can produce the numbers to 9 as spoken words which are stored optically.

5. Developer Clarke and Smith Research Ltd. Melbourne House, Melbourne Road, Wallington, Surrey, England.

Application Braille output device.

State of development General availability September 1975.

Description A 12, 25, or 72 character 6-dot braille display with bit serial input and solenoic activiated dots,

Price Circa £300 for 25-character display.

189 Developer Dr. A. J. Croft and D. T. Smith The Clarendon Laboratory, Department of Physics University of Oxford, Parks Road, Oxford OXl 3PU, England

Application Output for electronic calculators, etc.

State of development A dozen are being built for further evaluation.

Description Serial audio output with two tones — short tone is a single unit and a long tone is five units. It has been made on five printed circuit boards as an integral part of a calculator.

Price Circa &250.

Developer K. Cummins School of Engineering University of California 401 Verano Place Irvine, California, 92664, U.S.A.

Application Output for electronic calculators, etc.

Description The working prototype consists of two units — the tactile output device and the main calculator unit. The output device has a 10-digit capability but displays one digit at a time; a rotary switch selects the digit to be displayed. After the output is read for each digit, the pins are reset by hand.

8. Developer G. F. Dalrymple Sensory Aids, Evaluation, and Development Center Massachusetts Institute of Technology Cambridge, Massachusetts 02139, U.S.A.

Application Braille analogue of "Nixie" tube or 7-segment numeric display.

State of development Building 10 dual cell numeric only modules for further evaluation.

Description The braille display (four dots in the prototype) is composed of modules each containing two cells. These modules can be stacked to form a line of braille of any desired length. Each pin is connected by a lever to a solenoid; these levers are arranged in layers to prevent mechanical interference. The head can also contain a latch arrangement for holding the braille pattern without expending power in the main solenoids.

Price Display and drive electronics are estimated at $100 per unit plus $200 for each two digits, i.e., a 4- digit display is approximately $500 or a 10-digit display $1100.

190 Developer The Electro Physics Company 9 30 3 North Major Avenue Morton Grove, Illinois 60053, U.S.A.

Application Output for electronic calculators, etc.

State of development In limited production.

Description A binary coded decimal display with 7-digit capability. Each cell consists of 4 solenoid-operated pins.

Price Circa $1600.

10. Developer W. A. N. Ellis Tudor Lodge Cottage Fairmile Park Road, Cobham, Surrey, England.

Application Braille display as an output for digitally stored information.

State of development First prototype produced. General availability in 1975,

Description Serial output to 6-dot braille cell operated by solenoids. The user can set the output rate.

Price $65.

11. Developer. Dr. R. M. Fish

' 1507 West Acre Road, ' -'. ' Joliet, Illinois 60435, U.S.A.

Application A mobility aid when used in conjunction with a television camera. A non-visual equivalent to the cathode ray oscilloscope.

State of development Prototype built.

Description A binaural two-dimensional display has been developed. The frequency of the tone corresponds to the

vertical position of the scan (200-7000 Hz) . The relative intensities of the tones in each earphone are functions of the left-right position; interaural intensity differences as high as 40 dB are used.

191 .

12. Developer A. P. Grunwald Argonne National Laboratory, 9 700 South Cass Avenue, Argonne, Illinois 60439, U.S.A.

Application Braille display as the output module of the Argonne system.

State of development First production prototype.

Description Recirculating plastic belt with variable speed drive, means for writing and erasing, start-stop and speed control, timing pulse generator, speed 4-20 characters-per-second (up to 6 dots each) U.S. Patent #3,624,772.

Price $500 - $1500.

13. Developer R. J. Hitschmann Department of Engineering, University of Warwick, Coventry CV4 7AL, England

Application Output for electronic calculators, etc.

State of development Prototype built.

Description A serial 4-dot solenoid-operated braille display which can be connected to a commercially available electronic calculator.

14. Developer K. Isakson Karlavagen 90, S-115 22 Stockholm, Sweden.

Application Output for electronic calculators, etc.

State of development Prototype being evaluated.

Description The tactile display is read by moving the fingers over the array of braille cells; an audible signal indicates the correct figure in each array. The reading speed is about four figures per 10 sec-

onds .

Price 1500 Sw. Crs.

192 15. Developer R. E. B. Makinson and L. I. Paterson School of Mathematics and Physics, Macquarie University, North Ryde 2113, Australia.

Application Output for electronic calculators, etc.

State of development Prototypes built.

Description A serial 6-dot braille solenoid-operated display which can be interfaced for seven segments of BCD input. A thumb-wheel switch selects digits

in turn . Supplemented by audible pip for a decimal point when the calculator display does not allot it a separate place.

Price Cost of materials A$ 15.

16 . Developer Mauch Laboratories Inc. 30 35 Dryden Road Dayton, Ohio 45439, U.S.A.

Application Reading aid.

State of development Commercially available.

Description The Stereotoner outputs 10 musical tones. The input is normally from, a camera held over normal ink print, so the user has to recognize the characteristic set

of tones for each letter. i

Price $1020.

17. Developer J. Penksa Bialobrzeska 53, Warsaw, Poland

Application Output for electronic calculators, etc.

Description The Discon 201 is a 6-dot braille solenoid-operated display. The output is parallel to seven digits

(polarity sign plus six digits) ; also includes six positions for the decimal point. The input is either BCD (1,2,4,8), BCD (1,2,2,4) or decimal. A 13-digit (sign plus 12 digits) display will soon be in production.

18. Developer Professor J. A. Phillips and Dr. P. N. Seligman Department of Electrical Engineering, Monash University, Clayton, Victoria, Australia. Application Numeric display for use in a braille digital multi-

meter . State of development Prototype built.

Description A 4-dot solenoid-driven braille display where the user presses a button to energize the 3-cell display,

19 3 . .

19. Developer Professor J. A. Phillips and Dr. P. N. Seligman Department of Electrical Engineering, Monash University, Clayton, Victoria, Australia.

Application A non-visual equivalent to the cathode ray oscillo-

scope .

State of development Prototype built.

Description The instrument makes use of a multi-dimensional auditory display in which frequency, amplitude and timbre are all utilized. The instrument, which incorporates a signal-sampling time-scaling device, can be connected to an essentially conventional cathode ray oscilloscope.

20. Developer T. Robinson University of Manchester Institute of Science and Technology, Manchester, England.

Application A non-visual equivalent to the cathode ray oscilloscope.

State of development Prototype built.

Description A voltage controlled oscillator gives an audio display in which the frequency depends on the y deflection

21, Developer Dipl. -Ing. K. -P Schonherr Arbeitsgemeinschaf t fur Rehabilitationstechnik an der

' . Universitat Stuttgart e.V., D-7000 Stuttgart, Schloss Solitude,

; Haus 3, West Germany.

Application Output for electronic calculators, etc.

State of development Prototype built.

Description The prototype consists of two units — the tactile output device and the commercial pocket calculator. The output device has a 12-digit braille display (5 dots from a 6-dot braille cell)

Price DM 20 00 total.

19 4 .

22. Developer Science for the Blind Products 221 Rock Hill Road, Bala-Cynwyd, Pennsylvania 19004, U.S.A.

Application Output for electronic calculators, etc.

State of development Production models have not yet been built.

Description The CALCU-TAC is a braille printout system for use

with electronic calculators . When printout is required a "print" button is pressed. The braille printer will print all the numbers in the display in proper sequence, including decimal. The braille printer will print out on narrow tape.

Price Circa $700 - $1000 but possibly $500 when in full production

23. Developer N. B. Sutherland The Mitre Corporation, Bedford, Massachusetts 01730, U.S.A.

Application Braille output device.

State of development Prototype built.

Description The six pins of a braille cell are operated pneumatically and locked in position by a solenoid.

24. Developer Tactile Display Systems, Inc. 911 South Owen Street, Mount Prospect, Illinois 60056, U.S.A.

Application Output for electronic calculators, etc.

State of development Working prototype.

Description An electronic calculator with a 9-digit braille solenoid-operated display. The braille is displayed when the momentary push button switch is pressed. This can be utilized at any point during calculation. The decimal point is presented by blanking one of the nine digits giving an eight digit display capability.

Price Circa $250 each for 1000 machines.

195 25. Developer Telesensory Systems, Inc. 1889 Page Mill Road, Palo Alto, California 94304, U.S.A.

Application Reading aid for print and many electronic calculators and cathode ray tubes.

State of development Commercially available.

Description The Optacon has a camera which displays an enlarged tactual version of ink print text on a 2 4 x 6 array of vibrating reeds. The display consists of highly efficient piezoelectric bimorph stimulators (250 Hz) Tactile stimulator format: rods spaced 45 mils vertically, 90 mils horizontally, with an overall area 1.1 X 0,55 inches. ~^ Price $3450.

26. Developer Telub AB Pack, S-351 01 Vaxjo, Sweden.

Application Output for electronic calculators, etc.

State of development General availability in Spring 19 74.

Description A 4-dot braille display for connection to a binary coded decimal output. A 12-digit version is intended for use with calculators; presentation includes sign and decimal point (fixed or float-

ing) . A 5-digit version is intended for use with digital voltmeters, counters and similar equipment.

Price The 12-digit display, including a Facit 1118 calculator, will cost approximately Skr. 8000.

27. Developer Brother G. Kane Physics Department, Manhattan College, Bronx, New York 104 71, U.S.A.

Application Output for electronic calculators, etc.

State of development Feasibility study completed. Patent pending.

Description Braille output is detected by electrocutaneous stimulation using concentric electrodes activated by constant current bipolar pulses.

196 . , ,

NON-VISUAL COMPUTER PERIPHERALS

J. M. Gill*

Non-visual computer peripherals A survey of devices in produc- have been developed for: tion, or under development, based on replies to a questionnaire circulated 1. Braille production, in May 1974, The main features of these devices are summarized in Ta- ble 1. 2 Information retrieval systems A selected bibliography on the use of computers by the visually im- 3. Outputs for reading paired, but excluding papers on speech machines synthesis and the employment and training of blind programmers , is at the 4. Blind programmers. end of the report.

1. Developer American Systems, Inc. 123 Water Street Watertown, Massachusetts, 02172, U.S.A.

Status Commercially available as part of the ARTS system.

Description Compiled speech output delivered over telephone lines. ASI speech and teleprocessing system Nucleus 3000 V/L with central processor and 8192 x 16 bit core memory, ASI 16-line multiplexer with receiver data sets or modems and speech storage disc capable of containing up to 500 seconds of speech definable as more than 2000 English spoken words.

Price $66 ,700

Information Dr. K. R. Ingham, June 1974.

*Department of Engineering, University of Warwick, Coventry, CV4 7AL, England.

197 TABLE 1

Summary of the Main Features of the Devices

Tactual Braille Braille Speed Device Speech-like Soft Copy Paper Tape Page Chars ./Sec.

1 ARTS 2 Anderson 3.3 3 Boiten 4 Boldt 5 Boldt 6 Brown + 10 7 Charlesworth + 8 Gee + 10 9 Grunwald + 10 Haskins 11 Heginbotham + 60 12 de Jong + 200 13 Kruger 14 Loeber + 16 15 Longini 16 MIT + 16 17 Myers - + 18 PAL + 10 19 Rahimi 20 Rubens te in + 10 21 Schonherr 30 22 Spanjersberg + 7 23 Tagg + 8 24 Telesensory 25 Thiel + 12 26 TNO + 4 27 Trask + 8192 28 BD-3 + 15 29 LED-120 + 15 30 LED + 120 31 Wienberger + 32 Zawistowski + 12 33 IBM + 250

198 .

2. Developers G. B. Anderson and D. W. Rogers Lawrence Radiation Laboratory University of California

Livermore , California, U.S.A.

Status Unknown

Description New print head for Model 33 teletype. Speed one third that of the normal teletype.

Information Research Bulletin, No. 22, Dec. 1970, pp. 111-117.

3. Developer Prof. Ir. R. G. Boiten Laboratorium voor Werktuigkundige meet-en Regeltechniek Technishe Hogeschool Stevinweg 1 De 1 f t , The Ne the r 1 ands Status Experimental prototype.

Description Braille printer, powered by electric motor, which prints on Kraft paper tape moving from right to left.

Information Research Bulletin, No. 26, June 1973, p. 219.

4. Developer Professor Dr. Werner Boldt, Pedagogische Hochschule Ruhr Dept. of Educ. and Rehab, of the Visually Handicapped 46 Dortmund Kreuzstr. 155 West Germany.

Status Commercially available.

Description Electronically-controlled system for programmed learning of the blind. Audio and braille output freely combined. Braille and multiple choice input according to the character of the program. Braille input electrically evaluated as well as multiple choice input. Branched programs- branching controlled by the input modes. All information (audio, braille, memory, coding) stored on magnetic tape

(cassette type) . During stop periods use as electronic braille-writer, with program cassette or additional material useful for most school subjects.

Price Circa $2,800.

Information Prof. Dr. W. Boldt, May 1974.

199 .

Developer Professor Dr. Werner Boldt Pedagogische Hochschule Ruhr Dept. for Educ, and Rehab, of the Visually Handicapped 46 Dortmund Kreuzstr. 155 West Germany.

Status Prototype development.

Description "Braillex" enables the blind user to reach prestored information with relatively short access. Information (verbal or braille) is stored and coded on magnetic tape cassettes. Verbal and braille output (for control) is possible. The information wanted is "called" by braille input and discriminated electronically. Examples: information from dictionaries, storing and use of private archives (telephone numbers, scientific notes, etc.), recalling of special parts in literature (pages, chapters, code words)

Information Prof. Dr. W. Boldt, May 1974.

"f^ '-,-;: 'or

Developers V. Brown and E. Stuckey Teletype Corporation Little Rock, Arkansas, U. S.A. in collaboration with: E. Knoch Arkansas Enterprises for the Blind, Inc. 2 811 Fair Park Boulevard Little Rock, Arkansas 72204, U.S.A.

Status One prototype built.

Description The device, which produces braille at 100 words-per-

minute , is in parallel with a standard model 32 or 33 printer so there is keyboard input, print output and braille output simultaneously. The unit consists of a modified punch which punches two rows of three dots each or two rows of four dots each, depending on the need, instead of the 8-bit ASCII.

Price The original price of duplicating this unit in quantity was estimated to be $1600. However, with the modern LSI circuits available, it is felt that this device could be duplicated for under $600.

Information E. Knoch, June, 1974.

200 7. Developer D. V. Charlesworth Clarke and Smith Industries, Ltd. Melbourne House Wellington, Surrey, England.

Status Prototype built, general availability mid 197 6.

Description 12, 24, or 72 character 6-point braille display with solenoid operated dots module and computer terminal/information retrieval unit. Fresh lines of information are obtained by pressing button at end of line.

Price fc300 - £1600 depending on application.

Information D. V. Charlesworth, May 19 74.

8. Developer Dr. M. J. Gee Scientific Systems Group Office of Computing Activities University of Dayton Dayton, Ohio, 45409, U.S.A.

Status Prototype

Description A paper embosser which employs an ASR 33 terminal, and produces a configuration of seven dots arranged in an expanded braille cell, with the addition of a seventh position below position six.

Information Association for Computing Machinery Newsletter for Blind Computer Programmers 4th Issue, Vol. 2, No. 1, page 5.

9 . Developer Dr. A. P. Grunwald Engineering and Technology Division Argonne National Laboratory 9 700 South Cass Avenue Argonne, Illinois 60439, U.S.A.

Status Production prototype.

Description Device converts symbols recorded on magnetic tape to raised braille dots on a plastic belt. The belt moves conveyor- like at an adjustable speed; the dots are "erased" by depressing them and new ones raised.

Information Research Bulletin, No. 23, June 19 71, page 10 6.

201 .

10. Developer Haskins Laboratories, Inc. 270 Crown Street New Haven, Connecticut 06510, U.S.A.

Status Prototype

Description Full text-to-speech processing system is now working in

the laboratories . The synthetic speech is not entirely natural although intelligible. Sufficient progress has been made to justify the construction of a pilot reading service center which would have a combined research and service function.

Price $350,000

Information P. W. Nye, June 19 74.

11. Developers Professor W. B. Heginbothan, D. W. Gatehouse, and D. G. Hassel University of Nottingham University Park Nottingham, NG7 2RD, England.

Status Design study.

Description A terminal to produce a 36-character-wide line of braille print at a speed of 60 characters-per-second operating on a line printer principle. Also, meant to be adaptable to desk calculating machine. The machine sets up the braille terminals by operating from a bank of seven electromagnetic devices.

Price Estimated &1000.

Information Prof. W. B. Heginbothan, May 1974.

12. Developers Professor D. de Jong and Jr. A. N. Westland Technological University Delft Leeghwaterstraat 3 Delft, The Netherlands

Status First prototype ready in July 1974

Description Braille lineprinter operating with papertape as input or on-line with a computer. Working on normal braille paper

(180 gsm) , printing five lines-per-second , each one consisting of maximum 40 braille cells.

Information Jr. A. N. Westland, May 1974.

202 13. Developer F. Kruger National Center for Deaf Blind Youths and Adults New Hyde Park, New York 110 40, U.S.A.

Status Prototype available.

Description A single-cell softcopy braille display which can be connected to a computer via an acoustic coupler.

Information Electronics, 7th February 1974, page 46.

14. Developer N. Loeber IBM Corporation P.O. Box 6 6

Los Gatos , California 9 50 30, U.S.A. . .

Status Experimental unit.

Description Page braille embosser, based on standard IBM terminal. The unit embosses from the rear, with the data appearing on the front side of the paper. A metal die is used to mate with the selected pins to provide positive control in forming the raised dots.

Information American Federation of Information Processing Societies, Vol. 39, 1971, pp. 79-87 and N. Loeber, June 1974. ..

15. Developer Professor R. L. Longini Medical Systems Engineering Laboratory Carnegie-Mellon University Pittsburgh, Pennsylvania 15213, U.S.A.

built. •. . . . Fourth generation device . Status .

Description An alphanumeric audio output where the sounds are voice-like enough so that 100 percent phonetic output can be learned as a dialect. Forty hours of training (high 10) permits 150 words-per-minute of English to be understood.

Price Circa $300 if built in quantity.

Information Professor R. L. Longini, J\me 1974.

16 . Developer Massachusetts Institute of Technology 7 7 Massachusetts Avenue Cambridge, Massachusetts 02139, U.S.A.

Status Commercially available but supply limited.

Description The Braillemboss is an automatic braille printer which can operate at speeds up to 16 characters/second.

Information Research Bulletin, No. 24, March 1972, page 161.

203 17. Developer F. H. Myers Bell Telephone Laboratories, Inc. 6 200 East Broad Street Columbus, Ohio 43213, U.S.A.

Status Laboratory prototype.

Description A braillewriter was modified for electronic control, Seven solenoids and seven contacts were added.

Information Research Bulletin, No. 24, March 1972, page 164.

18, Developer Perceptual Alternatives Laboratory 35 8 Life Sciences Building University of Louisville Louisville, Kentucky 40208, U.S.A. in collaboration with Electronic Systems Development Corporation.

Status Second prototype built.

Description Braille page embosser that will operate at teletype speeds,

Price Under $1000.

Information Annual report of Perceptual Alternatives Laboratory, July 1974.

19 . Developers Professors M. A. Rahimi and J. B. Eulenberg Department of Computer Science Michigan State University East Lansing, Michigan 48824, U.S.A.

Status Limited production.

Description The system uses the speech synthesis hardware/software package implemented on Michigan State University's main computers. The buffered output is normally operated at 300 Baud. The phonetic images of words and sentences are built up by concatenation of 8-bit symbols representing the phones of English in the buffer memory. The 8-bit code consists of six bits representing the phone and two bits representing one of four levels of intonation.

Price $4000 - $5000 per unit.

Information Professor M. A. Rah ami , July 1974.

204 . .

20. Developer R. Rubenstein University of California Irvine, California 92664, U.S.A.

Status One off-working model.

Description A teletype terminal was adapted to produce braille print-

out .

Information Research Bulletin, No. 24, March 1972, page 161.

21. Developer Dipl. -Ing. K.-P. Schonherr Arbeitsgemeinschaft fur Rehabilitationstechnik an der Universitat Stuttgart e.V. D-7000 Stuttgart 1 Schlob Solitude, Haus 3 Germany

Status Prototype

Description The device is an electromechanical modular display, the module being a 6-dot braille cell such that braille lines of any desired length can be made. Writing speed is 30 characters -per- second.

Information Dipl. -Ing. K.-P. Schonherr, June 1974.

22. Developer A. A. Spanjersberg '' Dr. Neher Laboratory ." ':':. -^ St. Paulusstraat 4

' - Leidschendam, The Netherlands. ,

Status Six printers have been built.

Description The seven bits of a braille code have to be presented in parallel to the input. The braille characters are embossed on the paper which is transported in the printing mechanism by pin feed. The speed is about seven braille characters- per-second.

Price D.Fls. 15,000. -'--'

Information A. A. Spanjersberg, May 19 74.

205 23. Developer Dr. W. Tagg Hatfield Polytechnic Hatfield, Herts, England. In conjvmction with the Royal National Institute for the Blind and Business Data Products, Ltd.

Status Prototype built and in daily use.

Description For Output - IBM Model D braille typewriter plus sighted typewriter. For input - specially designed keyboard. These three components are linked electronically to each other and to a standard Datel data transmission unit.

Information M. Jenkins, Hatfield Polytechnic, May 19 74.

24, Developer Telesensory Systems, Inc. 1889 Page Mill Road Palo Alto, California 94304, U.S.A.

Status Prototype being evaluated

Description Hewlett-Packard (Santa Clara) have developed an ASCII to Optacon interface bypassing the Optacon's camera to feed the signals directly into the circuitry.

Information Measure (Hewlett-Packard), May 19 74.

25. Developer Ing-Buro Thiel 6105 Ober-Ramstadt Grafengasse 2, West Germany.

Status Commercially available from stock.

Description 'Braille tape input-output attachment to enable blind person

to operate telex-teleprinter m.achines . Output speed is 12 braille characters-per-second.

Information International Catalog of Aids and Appliances for Blind and Visually Impaired Persons, 19 73.

26. Developer T.N.O. Delft, The Netherlands.

Status One off-working model.

Description Input is from eight hole punched paper tape which is fed to a modified Perkins brailler and automatic cutting apparatus. A code-translator has a maximum capacity of one braille plate embossing and six braille printing machines. The speed of production is four characters-per-second.

Information Research Bulletin, No. 25, January 1973, page 266.

206 . . .

27. Developer Trask Datasystem AB Stockholmsu, 34 182 74 Stocksund, Sweden.

Status Prototype being built.

Description The Zoltan Braille Embosser is a fully automatic system which can provide single copy documents at the rate of 10 seconds-per-sheet embossed on both sides (interpoint) A second sheet with the same text takes 1 second and all additional copies 1/4 second each. The quality will be equivalent to that obtained from rotary press using zinc plates. All the functions of the machine are controlled by an electronic digital controller. Text input is made with paper or magnetic tape or alternate as desired. For this reason the acquisition and storage costs of a text library are minimal.

The machine contains two embosser drums which print both sides of a sheet simultaneously. Each drum contains 6720 movable pins. The input text determines the position of each pin, which in turn embosses the paper to provide the corresponding braille text.

"^ Price Circa Skr. 200,000. " ^ ' Information Trask Datasystem AB, May 19 74.

28. Developer Triformation Systems, Inc. P.O. Box 127 Wall Street Station

New York, New York 10005, U.S.A. ' •

Status Commercially available, delivery 90 to 120 days.

Description The BD-3 is a portable unit which produces braille on paper tape, when used in conjunction with a keyboard, digital equipment, a computer, almost anything that uses coded information. Weight - 15 pounds. Speed - up to 15 cps Operating modes - EIA interface, TTY interface (other inter-

faces optional). Input code - ASCII, EBCD , BCD, Correspond- ence, Baudot, any code with eleven bits per character or less

Price $1850 ^

Information Triformation Systems, Inc., June 19 74.

207 . .

29. Developer Tri formation Systems, Inc. P.O. Box 127 Wall Street Station New York, New York 10005, U.S.A.

Status Under development.

Description The LED-1 is a stand-alone braille device which produces braille from a keyboard, from a computer, from almost any coded information on a page of paper. Speed up to 15 cps Operating modes - EIA interface, TTY interface (other interfaces optional). Input code - ASCII, EBCD, BCD, Cor- respondence Baudot, any code with eleven bits per character or less

Information Triformation Systems, Inc., June 1974.

30 . Developer Triformation Systems, Inc. P.O. Box 127 Wall Street Station New York, New York 10005, U.S.A.

Status Commercially available, delivery 90 to 120 days.

Description The LED-120 is a high-speed braille printer which can produce braille from a keyboard, from a computer, from magnetic cassettes or from almost any source of coded information. Speed up to 120 cps. Operating modes - EIA interface, TTY interface (other interfaces optional). Input code - ASCII, EBCD, BCD, Correspondence, Baudot, any code with eleven bits per character or less.

Price $9000.

Information Triformation Systems, Inc., June 1974.

31. Developer Z . Weinberger National Physical Laboratory of Israel Hebrew University Campus Jerusalem, Israel.

Status Unknown

Description A braille tape embosser produces braille cells on half- inch machine paper tape. By suitable interfacing, the brailler can accept input from computers, card readers. Telex tape and Monotype tape.

Information Research Bulletin, No. 21, August 1970, page 118.

208 .

32, Developer W. A. Zawistowski Computation Centre Polish Academy of Sciences P.O. Box 22-00901 Pkin, Warsaw, Poland.

Status Prototype.

Description The tape reader-writer unit (type WAZA) reads 8-hole punched paper tape and outputs braille embossed on paper tape. Writing speed is 12 characters-per-second

Price Circa $100

Information World Council for the Welfare of the Blind, Sao Paulo, August 197 4.

33. Developer IBM Corporation Armonk, New York 10504, U.S.A. - -

Status Commercially available.

Description The printing mechanism on an IBM 1403 line printer has been modified to produce braille at about 250 characters- per-second.

Price Dkr. 550,000.

'" Information J. Vinding, September 1974. ,

209 . .

REFERENCES

American Systems, Inc. Arts Equip- Computerised braille production.

ment and Services . 123 Water Copenhagen, September 19 74, to Street, Watertown, Massachu- be published. setts 02172, USA, Nov. 1973,

21 pp. Dinius , G. Beitrag zur bestimmung der dv- funktionen und ihrer Anderson, G. B. & Rogers, D. W. zuordnung zu funktionstragern An inexpensive braille terminal und darauf aufbaunde konzeption device. Communications of the zur ers chliebung von blinden- Association for Computing Ma- arbeitsplatzen in der edv. chinery, Vol. 11, June 196 8, Dissertation, Technischen 417, 418, 440. Hochschule Aachen, Germany, 1971, 266 pp. Association de Nederlandsche Blinden- bibliotheek. Provision of hraille Eickenscheidt, B. The braille reading-matter by means of punched translation program of Munster

tapes. Noordwall 7, The Hague, university . The Munster Work- Holland, 1973, 15 pp. shop, March 19 73.

Boyer, J. J. Brailletran : A compre- Forbes, J. Braille translation: The hensive braille trans cription computer aids the blind. Mitre program. Proceedings of the Matrix, 1972, 5(2), 22-31. Braille Research and Development Conference. Cambridge, Mass.: Foulke , E. Computer services for Massachusetts Institute of Tech- the blind. The Braille Monitor, nology, November 1966, 10-13. May 1972, 207-10.

• , , Boyle, J., Jacobs, W. , & Loeber, N. , Gerhart, W. R. Millen, J. K. & A braille code for interactive Sullivan, J. E. Dotsys III: A terminal use. American Founda- portable program for grade 2 tion for the Blind, Research braille translation. Mitre Bulletin No. 27, April 1974, Corporation, 19 71, 139 pp. 267-72. Glaser, E. L. Small computers and grade 2 braille. Proceedings of Branachk, H. & Brix, G. the Braille Research and Develop- Mashinens chreiben fur blinde und ment Conference, Cambridge, Mass.: sehschwache : Ein lehrbuah fur den Massachusetts Institute of Tech- gruppenunterricht und sum selbst- nology, November 1966, p. 14.

studien . 1966 Goldish, L, H. Braille in the United Clarke & Smith Industries. Clarke States: Its production^ distribu- & Smith braille system. Melbourne tion and use. New York: American House, Wallington, Surrey, England, Foundation for the Blind. February 1974, 11 pp. Grunwald, A. P. A braille reading Coleman, P. W. F. Computer terminals machine. American Foundation for for the blind. Electronics and the Blind, Research Bulletin Power, March 1972, 18, 84-86. No. 16, May 1968, pp. 73-78.

Dalrymple, G. F. Comparison of uni- Grunwald, G. M. Prospects for uti- versity of Dayton and MIT one-cell lization of compositors tape in braille codes. Sensory Aids the production of braille Evaluation and Development Center. Argonne National Laboratory, 19 71. Cambridge, Mass.: Massachusetts Institute of Technology, Report Final Report to John A. Hartford No. 49, September 19 70. Foundation. Development of a high-speed brailler system for Danish Association of the Blind. more rapid and extensive produc- Proceedings of the workshop. tion of informational material

210 . , , . , . .

fov the blind. Cambridge, Mass.: facture. Sensory Aids Evaluation Sensory Aids Evaluation and Devel- and Development Center, May 1967. opment Center, Massachusetts Insti- tute of Technology, September 19 70, Massachusetts Institute of Technology. reprinted 19 73, 54 pp. Proceedings of the conference on new processes for braille manu- Ingham, K. R. Spelled speech as an facture. Sensory Aids Evaluation output for computers and reading and Development Center, February machines for the blind. IEEE 1968. Transcript HFE-8, No. 3, Septem- ber 1967. Mick, D. E., & Hittinger, G. N. An inexpensive, universal braille Ingham, K. R. Braille, the language (UNIBRL) output device. Research its machine translation and dis- Bulletin No. 24, March 1972, play. IEEE Transcript MMS-10, pp. 158-160. No. 4, December 1969. Millen, J. K. Choice of cobol for Ingham, K. R. A speech read-only braille translation. Mitre memory store for blind programmers Corporation, December 1969, 19 pp. IEEE Transcript BME-17, No. 2, April 19 70. Millen, J. K. Dotsys II: Finite state syntax-directed braille transla-

Kamp , H. Gaining production rules tion. Mitre Corporation Report for a Markov braille translation No. 1829 , 1970, 43 pp. algorithm. The Munster Workshop, March 1973. Morrison, R. E. Braille embossing and transmission equipment. Leffler, L. C, & Prastein, S. M. American Foundation for the Blind, The Argonne braille translator Research Bulletin No. 22, December Argonne National Laboratory 1970, pp. 71-82. Technical Memorandum No. 180, May 1969. Nemeth, A. Recent developments in computerized mathematical braille

Lichtenstein , M. The Medcomp Braille Proceedings of the Conference on Program. Computer work for the New Processes for Braille Manu- blind. Cincinnati, Ohio: College facture, Massachusetts Institute of Medicine, University of Cincin- of Technology, February 1968, pp. nati, September 1963, pp. 23-27. 50-53.

Loeber, N. C. Proposed braille com- Parsons, J. L. Computer program to puter terminal offers expanded translate inkprint to grade 2 world to the blind. Conference braille. Department of Engineer- Proceedings of the American Fed- ing, University of Warwick, Eng- eration of Information Processing land, 1972. Societies, Vol. 39. Pearson, P. J. Tactile music score.

Mann, R. W. , & Gammill, R. C. Auto- Department of Engineering, Univer- mating braille translation and sity of Warwick, England, 19 74.

presentation . In: Evaluation report on work in progress on Schack, A. & J. Computer conversion sensory aids and prosthetics of compositors tapes to grade 2 Department of Mechanical Engineer- braille. Proceedings of the ing, Massachusetts Institute of Braille Research and Development Technology, Report No. 9211-2, Conference, Massachusetts Insti- April 1964. tute of Technology, November 1966, pp. 44-74. Massachusetts Institute of Technology. Proceedings of braille research Schack, A. & J. Studies in the auto- and development conference mation of braille mathematics and Sensory Aids Evaluation and De- music. Final Report to the Depart- velopment Center, November 19 66. ment of Health, Education and Wel- fare, The American Printing House Massachusetts Institute of Technology. for the Blind, June 1969, pp. 8- Proceedings of the conference on 78. new processes for braille manu-

211 . .

Schack, A. S., & Mertz, R. T. Vinding, J. Braille printing program Braille translation system for system description. Copenhagen: the IM 704. New York: IBM Data Dansk Blindesamfund, Computerised Systems Division, Mathematics and Braille Production, 19 74. Applications Department, 1961, 6 6 pp Vinding, J. Braille translation program system description . Copen-

Siems , J. R. Report of new braille hagen: Dansk Blindensamfund, Com- translation program at APR. Pro- puterised Braille Production, 1974. ceedings of the Conference on New Processes for Braille Manufacture, Werner, H. , Dost, W. , & Seibt, P. Massachusetts Institute of Tech- Automatic translation of inkprint nology, February 1968, pp. 5-10. to braille by electronic data processing systems. American Founda- Slaby, W. A. A universal braille tion for the Blind, Research Bul- translator. The Munster Workshop, letin No. 14, March 1967, pp. 99- March 19 73. 108.

Smith, G. W. The ARTS system. Braille Werner, H. The historical develop- Forum, November 1972, 11(3), 14-17. ment of automatic braille production in Germany. The Munster Sullivan, J. E. DOTSYS III: A port- Workshop, March 19 73. able braille translator. Proceed- ings of the Association for Com- Windebank, C. L. Computer-produced puting Machinery, 1973, pp. 398- braille. Data Systems, November -'' 40 3. 1968, pp. 18-21.

Sutherland, N. B. Development of a Wolfson Industrial Automation Group. : A dynamic braille display . Mitre Braille computer terminal feasi- Corporation, Report MTR 19 51, bility study. Department of Pro- September 1970. duction Engineering and Production Management, University of Notting- Thomhill, D. E. Translation from ham, England, 1974, 13 pp. monotype tape to grade 2 braille. American Foundation for the Blind, Wolfson, R. G. System study of elec- Research Bulletin No. 5, July tronic braille reproducer : Final 1964. report to Veterans Administration prosthetic and sensory aids service University of Munster, Proceedings of June 1961. New York: Adelphi Re- the Workshop. Towards the aommun- search Center, 19 61, ARC Report ality of algorithms among braille No. 101. transcription systems for multi- lingual usage. Munster, Germany, Woodcock, R. W. Development of an March 1973, to be published. electromechanical braille transcribing and reproduction system. George Valery, N. Computer tutor for short- Peabody College for Teachers, Nash- May 196 5. hand braille. New Scientist , ville, Tennessee, 10, 19 72, 319. February Woodcock, R. W. Braille research at George Peabody College. Proceedings Vinding, J. Modern methods of trans- of the Braille Research and Develop- cribing and printing braille . ment Conference, MIT, November 1966, European Braille Conference, Oslo, pp. 20-28. September 1973, pp. 10-12. Woodcock, R. W. An electromechanical brailling system. Research Bulle- tin No. 21, August 1970, pp. 101- 107.

212 ,

COMPUTER BRAILLE SYSTEM

Don Keeping

Editor's Note. The following with respect to tab settings, iden- description of the Computer Braille tification of pictures, tables, and Service, Computer Center, The Uni- diagram handling. A preliminary re- versity of Manitoba, Winnipeg, Mani- port is made and after meeting with toba, Canada R3T 2N2 —which prepares other members of the group and the texts in braille for the Manitoba assistant supervisor to discuss length Department of Education —was written of volumes, format, etc., the report for us by Professor Donald Keeping, is completed. In addition to format, the supervisor of the service. The the report might contain proposed Center and Professor Keeping are also braille volume length (average number currently engaged in the development of cards per braille volume is ap- of algorithms for handling multi- proximately 850 cards, or from 5-3/4

language transcription of ink print to 6—1/4 inches) . Volume size de- into braille. pends to a large extent on content.

The analysis is referred to at every stage in the preparation of the STAFF ORGANIZATION book including keypunching, proofreading and correcting. If it becomes 1. Supervisor of project necessary during this operation to deviate from the rules set out in the 2. Assistant Supervisor analysis, another group consultation is held, and the analysis retyped. 3. Text analysts (part-time and full-time) divided into groups Reports of the various stages in of three or four persons each. the preparation of each volume are Each person in a group is re- made at the end of each working ses- sponsible for transcription of sion. one book and is assisted in various stages of transcription by other members of the group. PRODUCTION OF BRAILLE VOLUME 4. One part-time person to help to run the brailler, to do proof- Key punching is done by any mem- reading in braille, tearing and ber of the group. Data is keypunched collating of pages. on cards and from the viewpoint of the keypunch operator (text analyst) 5. One part-time person to document, is to be considered as one continu- maintain, and improve the sys- ous stream; that is, column 80 of one tem's programs. card is continuous with column 1 of the following card.

TRANSCRIPTION OF Preparation for braille trans- BRAILLE BOOKS cription necessitates modifications in the text which are m.ade in two

Upon arrival, a book is assigned ways : to a text analyst who will analyze it

213 .

1. Modifications in the text itself corrections inserted in a clear and by the keypunch operator. legible manner. Format commands are extremely important. A clear under- 2. Modifications carried out by the standing of the rules of modification DOTSYS program under the direc- is absolutely necessary before any at- tion of commands beginning with tempt at proofreading is made. a dollar sign interspersed throughout the text. These $ The following notes to the proof- commands must be inserted by the reader are part of a short training keypunch operator. course given to each of the text analysts before they attempt work on any The keypunch operator is also book responsible for setting up proper braille volume lengths according to "Do not assume that a keypunch the specifications mentioned above. operator knows the rules better Volume headings and footings should than you do. be inserted. Do not leave errors for future proofreaders to correct. From Cards to Disk If you do not understand an item From here on, the system is op- in the listing, question the As- erated through TSO (Time Sharing Op- sistant Supervisor or the Super- tion) . visor.

When keypunching is completed, Do not leave a 'known' error in cards are put into an OS data set. the text." The data is then run through a program "Cutdown," the function of which is to cut down the length of each re- First Corrections cord from 80 columns to approximately 50-55 colxomns. It also eliminates Corrections are made through TSO words which run over more than one at a computer terminal. The indivi- card, and renumbers the data set and dual Dotsys text files are made avail- outputs it on the printer. able through the Edit Command. This Command allows changing of text, in- sertions, deletion of text, and re- Backup placing of incorrect lines.

All of the OS files containing Dotsys text are backed up on tape. Second Proofreading This is a single-file tape containing and Corrections a copy of all the current Dotsys text files. It is updated once a week, or The procedure for second proof- when any major change is made in the reading and corrections is a repeti- current Dotsys text such as several tion, although a very important one, new volumes being added. When it is of the first proofreading and correc- is backed ascertained that the volume tions . up, the original cards are deposited in a retaining drawer for one week, At the end of each session of at the end of which time they may be keypunching, proofreading, or correc- destroyed. In case of loss, a Dotsys tions, an entry is made on the current text file may be retrieved from the book report for this book giving in- tape. The Copy program inserts all formation as to the current stage of

current files together into a single the work on the volume . This log is OS data set. The lost file may then extremely important and is always be isolated and retrieved with the kept up to date. Edit Command of TSO.

Final Look First Proofreading After the second corrections the The listing is proofread veiry Dotsys text file is run against a pro- carefully. Errors are noted and the gram "Look." The function of this

214 program is to check for errors in $ STORING DOTSYS TEXT commands, format, etc. The output of this program is a list of errors found When the volume has been brailled, in the text, along with their line the Dotsys text file is transferred to numbers. If there are any errors, an- a magnetic tape. These tape files ac- other correction session will be nec- cumulate and become the backup for the essary. system. They are the sources from which copies are made, and are also used as the basis for corrections if BRAILLE TRANSACTION the student finds errors in his text. Every two weeks a duplicate of the At this point, the text should magnetic tape files is made, and kept be ready for brailling. The trans- in the office. cription is done through two computer programs, "Dotsys" and "Return." The final output from these programs is CATALOGUING THE BOOKS an OS data set which contains the text in braille format, that is, the When a book is accepted for trans- lines are 39 columns long and 2 5 cription into braille, an entry is made lines-per-page and contain braille in a text data set to include the fol- contractions. This file is then run lowing information: author, title, pub- through the Triformation LED-120 or lisher, date of publication, number of the MIT Braillemboss . The braille braille volumes, subject, and a code pages are burst and the perforated indicating that this is a braille book edges removed. They are then sent done at the University of Manitoba. to the University of Manitoba Central The output from the program is a list Printing and Duplicating for binding. of the current books, as well as those The braille book is then labeled in already transcribed, sorted either by ink print and braille, and delivered title or author. Copies are neatly to the student. bound and are available to anyone.

215

. "

AIDS FOR THE SPEECH -IMPAIRED: INTERNATIONALLY COORDINATED DEVELOPMENT WORK

Birger Roos

Editor's Note. The following private firms interested in description of a forthcoming inter- this field, please let us national cooperative research program know . in development of aids to the speech- impaired seems important to us as a We concur with these sentiments, model for such undertakings, quite and we are sure that our readers will apart from its intrinsic importance; also. Inquiries about the project hence our inclusion of the details in might be best directed immediately the Bulletin. In a covering letter to Birger Roos at the Handikappinsti- Sweden. written by Karl Montan , Director of tutet, Pack, S-161 3 Bromma 3, the International Society for the Re- habilitation of the Disabled's Com- mittee on Technical Aids, Housing, and Transportation, it is noted that: BACKGROUND

"It is of extreme importance Persons who for different reasons that ICTA is able to build up have their speech capacity impaired, models for international co- have presently very limited access to operation in different acti- technical aids that may compensate vities concerning technical for their handicap and make the indi- aids for the disabled. NVR vidual less dependent upon the assis- in the Netherlands has con- tance of other persons. The present tributed to such a model as application for financial support in far as building norms are con- this area of technical aids aims at cerned. trying to remedy this. Considering the fact that the speech-impaired have "Handikappinstitutet in Sweden considerably different needs, depend- is now offering another possi- ing upon the nature of the impairment, bility to create an interna- the number of persons in each group tional model for research and in need of a certain type of aid is development in near coopera- relatively small. These circumstances tion with ICTA. This case make production of new products diffi- concerns a study of technical cult. The Swedish Institute for the aids for persons with speech Handicapped, therefore, wishes to in- impairment (see attached pro- vestigate the field of aids for the ject plan) speech-impaired on an international basis. This intention has previously "Very early in the study we been declared in a letter of Febru- will have to find out what ary 13, 1974 to the Swedish Board for is going on in this field in Technical Development. The appropri- other countries. For this ateness of international activity in purpose we will later on pre- this field has also been mentioned in pare a systematic inquiry but "Memorandum Concerning the Activities

we are already now eager to of ICTA" (1972-02-08) . hear your reaction to the project. If you know of any The project is calculated to run institutions, scientists or for three years at a total cost of

217 30 Swedish in 4,000 crowns, addition Delayed development of speech . to which there will be costs for ex- Due to the retardation or deafness, ternal investigations of 50,000 crowns the development of speech may be at most. delayed. Persons with speech difficulties due to this reason form a very heterogeneous group and the PROJECT DESCRIPTION situation for aids is not clear. This group in particular requires Handicap Groups suitable training aids. Certain aids have been developed at the Department Speech-impaired persons form a of Speech Communication, Royal Insti- heterogeneous handicapped group that tute of Technology, Stockholm. is still unknown both in number and in type of impairment. The following groups may be of interest to the pro- Language-impairment . Speech dif- ject. ficulties may also be caused by a reduced function of the speech center of the brain. Such an impairment is Phonation- impairment . Persons called aphasia and is often combined with phonation impairment have a re- with other impairments. The need duced function of the vocal cords and for aids has not been investigated. thus have difficulty in producing a vocal tone useful for speech. One group of these is the laryngectomized The Aid Situation who due to an operation in the larynx completely lack vocal cord function. As shown, the access to aids for This group comprises approximately speech-impaired persons is not satis- 500 persons in Sweden. (The popula- factory. It is known that individual tion of Sweden is approximately 8 m.il- aids have been developed in several lion. ) For them, there exist certain parts of the world. However, they tone-generating aids to be applied are not very well-known and there ex- outside of the throat, or in the oral ists no collected information on work cavity. These aids should be devel- in the field. Considering this fact oped and complemented with others. and the fact that speech-impaired per- The number of other persons with vo- sons form a relatively small handi- cal cord problems, such as a weak capped group in each country, inves- voice, is very little known as well tigation work in the field should be as the need of aids for these per- carried out on an international ba- sons. sis.

Stutterers . Problems with stut- Project Organization tering are relatively widespread, but may differ in character. Certain The Swedish Institute for the aids for stutterers have been devel- Handicapped, therefore, suggests oped in various places in the world an international project to investi- but have not been put into production gate the aid situation for speech- and have thus not reached wide dis- impaired persons , organized on the tribution. lines of an international reference group with a Swedish secretary, employed by the Institute. The project Articulation-impairment . Articu- is calculated to run for three years, lation-impaired persons are those with of which the last year to a great ex- a reduced mobility and coordination tent will be devoted to the creation ability of the moving parts of the of a permanent organization, intended speech organ; the lips and tongue. to work siobsequently in close contact The injury may be congenital or ac- with the Institute and ICTA. In the quired, following a hemorrhage. One first period of the project a Swedish investigation has estimated the num- administration group for detail plan- ber of articulation-impaired persons ning with representatives from the in Sweden in possible need of an aid Institute, the Department of Speech at between 2000 and 3000 persons. For Communication, Royal Institute of these there are no known aids beside Technology, Stockholm, and the Swedish conventional typewriters. Board for Technical Development will

218 . , , : . . . .

be established. Furthermore, the to a word memory, cind need for dif- Institute for the Handicapped will ferent kinds of presentation (writ- organize a Swedish reference group ing, synthetic speech, etc.) may be with representatives from different mentioned. specialties in the field, such as speech transmission, phoniatrics linguistics, and neurological rehabilitation.

The project slated to begin in the Fall of 19 74 will be constructed accordingly:

Institute for .Reference the Handicapped Group

Administration Group "" ~

I Institute for Dept. of Speech Dept. of Med. Rehab., the Handicapped Comm. , Royal Inst, of Sahlgren Hospital Technology

Project Work The project work is estimated to run according to the following The work will include the fol- schedule lowing: 1. Establishing project organiza- 1. Survey of the speech-impairment. tion and Swedish reference group (year 1) a. Type of impairment, diagnosis, symptoms 2. Existing documentation (year 1)

b. Number of persons and distribu- Collection and studies tion. Creating a documentary system. c. Need for aids, physical and mental prerequisites. 3. Present activities (year 1-2).

2. Survey of the situation for tech- Inventory by inquiry. nical aids (including training

aids) . Studies by visits to foreign institutions concerned. a. Existing aids; testing, judging. 4. Establishing an international b. Present development work. reference group (year 2).

c. Desired development work. 5. Symposium and work plan for the future (year 2) d. Prerequisites for initiation of development work. Outline of a work plan for the future. Apart from the work of the secretary and the reference group, cer- Outline of an international sym- tain special investigations will be posium and exhibition.

ordered from consultants , such as persons with phoniatric education. Symposium in Sweden. As an example of such investigations an analysis of the ability of certain Work plan for the future. speech-impaired persons to manage writing equipment, need for access

219 :

6. Permanent organization (year 3) synthesis equipment and writing equipment with word memory. Establishing period.

Steady state. The following persons are working within the project:

Costs K. Kalyas , project leader The costs for the project are calculated for year 1 to a total of P. Branderud, technical 82,00 Swedish crowns according to development the following calculation: L. Bondelind, technical Secretary development Salary, 10 months 40,000 Social expenses 12,000 The bulk of the work has been 15,000 Travel expenses concentrated so far on part 1. Documentary system 15,000

82,000 The Department of Medical Rehabilitation In addition to this there will , Sahlgren Hospital , be external investigation work as Gothenburg mentioned above at a total cost of 50,000 crowns maximum. For years 2 and 3 the costs are calculated at At the Department of Medical 111,000 crowns per year due to in- Rehabilitation in Gothenburg for creasing expenses for salary and the past year work has been in travel. progress on writing equipment for persons with anarthria; that is no speech ability due to injury of the cranial nerves and/or the ASSOCIATED PROJECTS nerves controlling the mobility of the tongue. For these persons The following projects will be special writing equipment consist- associated with the activities of ing of a keyboard and a display the Swedish Institute for the Handi- unit has been constructed. The very capped. , equipment has been received .1 positively by the handicapped, hospital staff, and others. It The Department of Speech has been determined that a valuable Communication, Royal Institute feature would be for the equipment of Technology, Stockholm to present whole words by pressing single keys. The speed of At the Department of Speech Com- the system would be thereby in- munication a project in this field creased substantially. Future has been running for four years. It work on the equipment will thus is divided into two parts aim at providing the system with a memory. The work is led by 1. Aids for phonation- impairment. Dr. Bengt Lindberg. Development of amplifier and vibrator with different types of frequency and amplitude control.

2. Aids for articulation-impairment. Technical investigations concerning hand- maneuvered speech

220 . ,

ASPECTS OF SOME PROBLEMS RELATmG TO THE TRANSLITERATION OF THE HAUSA LANGUAGE INTO BRAILLE

James I. Murnane*

An opportunity would, therefore, The subject was first suggested appear to exist for developing a pro- 1973 by Dr. D. J. M. Muffett, early in gram based on the research reported Assistant Professor at the Institute here, optimizing its potential for African Affairs at Duquesne Uni- of assisting in the rehabilitation of versity. He recalled a former blind blind in considerable numbers. American student who was studying the Beyond the initial inquiry a great Hausa and for whom a pilot braille deal of research still remains to be system had to be developed by the done and it is clearly not merely Institute in cooperation with volun- — desirable, but in fact quite teer workers of the Rodof Sholom Con- highly that it should be done gregation in Pittsburgh. A more de- imperative, Nigeria. tailed examination of some of the in problems which arose then seemed the summer of 1973 some clearly indicated. During experiments were conducted on certain of the problems expected to arise NEED ESTABLISHED during practical attempts at transliteration. We were aided by a blind who was fluent A further impetus arose when Al- American teenage girl braille, but who had no previous haji Tijani Malamfashi of the Mili- in whatsoever in Hausa or Governor's Office in Kaduna experience tary linguistics. With the aid of Nigeria, undertook to interest the in a pilot series of designators, however, Federal Government of Nigeria in the she was able to reproduce the appro- prospects of a transliteration of the with a more than rea- language into braille, especial- priate phonemes Hausa adequacy, except for Nigerian ex-servicemen blinded sonable degree of ly for variation in ton- War (1966-70). His ob- an occasional minor in the Civil became clear that servations confirmed that although ality. It thus further work with a Hausa-speaking there are presently well-developed and a language laboratory institutes for the blind in Northern instructor correct this deficiency. Nigeria, rehabilitation does not as could readily same problem would not confront yet include substantial programs for This Hausa speaker, of course, any the Hausa-speaking blind. None of a blind more than it would confront a sighted the necessary groundwork for a braille learning to read in Ro- system appears yet to have been under- Hausa speaker taken. Further confirmation of this man characters. view has come from the Principal of into the problems Army Blind Centre in The research the Nigerian transcription into Kaduna, Captain Waziri Mahdi arising from Hausa braille was continued. The thesis presents some of the major aspects of these problems, together with proposed solutions for them. The study appears to be unique, since no African have been *Duquesne University, Pittsburgh, languages as yet appear to analysis and Pennsylvania. subjected to such an examination.

221 , — ,

since the advent of colonialism. 200,000 blind Hausa speakers in North- Islamic institutions which formerly em Nigeria alone. had provided for the maintenance, assistance, and rehabilitation of the In the Hausa language area, the blind seem to have dwindled away. most significant cause of blindness Sporadic attempts have been made to are, in their order of prevalence: rehabilitate selected persons, some by national movements (such as the 1. Onchocerciasis,

Nigerian Federal Government) , some by Christian missionary groups or by 2. Trachoma and trichiasis, the Ahmidiyya of Islam; others along multidenominational lines usually 3. The prurulent keratides (follow- under the auspices of philanthropic ing smallpox, measles, and even societies, such as the Royal Insti- chickenpox) tute for the Blind in Britain (which has been active in Hausaland, and 4. Cataract, established a Northern Nigerian So- ciety as early as 19 52). None of 5. Eye injuries. these attempts, however, have so far considered the problems created by Greatly increased efforts have attempts at vernacular translitera- been undertaken recently by the Ni- tion. gerian Federal Government to rehabilitate the blind, although as yet there has been no attempt at the structured THE HAUSA LANGUAGE establishment of a system, of braille for the Hausa-speaking blind. Hausa is a major language not only in Nigeria, but also throughout However, the situation is much West Africa. It is spoken as a pri- more desperate for the blind Hausa mary and secondary language, or as a speakers living in Francophonic West commercial lingua franaa, by many Africa. In this region (Upper Volta, millions of persons in Northern Ni- Niger, Chad, Dahomey, Togo, Cameroons) geria, Niger, Northern Dahomey, the worst endem.ic area in the world Northern Ghana, Northern Togo, Cam- for onchocerciasis, there appear to eroons. Upper Volta, Chad, and the be no rehabilitation programs for Sudan. According to BBC statistics, the blind at all. it may be shared by as many as 100,000,000 people. It is a well- established written language. It A BRAILLE SYSTEM was transcribed in Aj ami (Arabic script) centuries ago, although its Certainly, it would be naive to transcription into the Roman alpha- expect that the majority of these per- bet dates only from about 1910. sons can be made readers of braille. Clearly, Hausa will not lose its Possibly, only between 15 and 30 per- status as a major language of West cent of 200,000 blind Hausa-speaking Africa within the foreseeable fu- Nigerians (75,000) could become pro- ture. ficient. However, even if only that number of persons can be helped, then the contributions which they would BLINDNESS IN THE make to the nation would be enormous. LANGUAGE AREA And although all the results of the system may not be put to utilitarian Meaningful statistics on blind- purposes, its adoption would at the ness in Hausaland are not easily ob- very least assist in putting an end tainable. Those most readily avail- to the terrible physical and social able, contained in Annales d' Oaulis- isolation which blind people suffer

tique , state that 0.8 to 1.0 percent which constitutes the major tragedy of the population in Northern Nigeria of the affliction. is blind. Of the 25,000,000 or so people in this region, at least The proposed Hausa braille sys- 2 2,000,000 may be regarded as compe- tem will conform as much as possible tent in Hausa for the purposes of with the English model, should a this project, and it can thus be es- blind Hausa speaker desire to study timated that there may be at least English after having mastered the Hausa system.

222 . ,

It is recognized that there is a 3. The essential Hausa phonemes B, French phonetic system for Haoussa D, and K must be represented, which is used in Francophonic West and the nonessential, but sig- Africa. In Niger alone, there is a nificant, phonemes TS and SH population of 1,800,000 Hausa whose will be included. phonetic experience in the language has been essentially Francophonic; 4. A number of syntactical peculi- but Haoussa in such localities has arities should be represented in survived for all practical purposes the adaptation of braille to the Among these are as an oral language , and French is Hausa language. always the official tongue. In any the conjunction (DA), identifi- case, the English phonetic system of ers (NE , CE , and KE) , locators Hausa would obviously be entirely in- (GA and MAI) , negative {BA) , the telligible to the Francophonic blind preposition and conjunction Hausa since the probability of his (SAI) , and the future particle being illiterate is overwhelming and (Z/1) . his instruction would thus take place ab initio. However, the divergence 5. All of the English "wholewords" between the French braille system and have been deleted, except for the the proposed Hausa system would be Hausa correspondent of "and" (the considerable, as a number of unique conjunction DA). Similarly, all French phonemes are to be deleted in the English "partwords" except the latter system. the phoneme (SH) have been dropped. In the proposed Hausa braille system, The Hausa braille system may be developed even further with the assig- 1. The entire Roman alphabet and nation of "wholewords" to the letters, the number system, qualified by as in the English system. However, the numeral sign, must be main- such future changes must await contained. Although the letters siderable research into frequency of Q, V, and X do not occur in usage and testing in the field. Hausa, and P is vestigial, it is necessary to retain these letters It is believed that this Grade in case the blind student should 1-1/2 braille scheme can minimize the desire to learn English braille space needed without obscuring the at a later date readability of the text.

2. The punctuation is to be retained Hopefully, all of the problems for the most part. Exceptions in this first transliteration attempt are in the separation of the have been solved, although of course English braille symbol for the the final determinant of the validity open quotation mark from the of each braille symbol will be ascer- quotation mark, which may have tained in the field study in Nigeria. led to some confusion; and the association of the Hausa particle KOI and TO! (with the question BIBLE TRANSLITERATIONS mark and the exclamation mark

respectively) . In the Hausa sys- It has been reported by Captain tem, one braille symbol will rep- R. White, Deputy Director of the Royal resent the quotation marks. Commonwealth Society for the Blind, that some braille transliterations of Other changes include the intro- the Hausa language already do exist. duction of addition, multiplication, However, further inquiry showed that and subtraction signs; a Hausa equiv- the documents in question are trans- alent of "italics" is substituted; and criptions of the Christian Bible, and the glottal stop is associated with are used in "Vernacular Bible Schools." the English symbol for the apostrophe. These materials would be of minimal interest to the Hausa-speaking blind, It must be mentioned that Hausa the vast majority of whom are Muslim. punctuation is not presently as so- Furthermore, the dissemination and phisticated as English punctuation, implementation of these braille tracts although in the future it is believed have been forestalled by contradic- that it will become more complex. tions among the three different braille

223 — systems used by proselytizing groups. reliance on electronic aids, includ-

For example, the transliteration of ing radios, phonographs, tape , record- the Christian Braille Ambassadors is ers , the Optacon, and talking books. not compatible with that of the Luth- Such equipment, however, would be eran Braille Workers, nor that of the prohibitively expensive for the vast Scripture Gift Mission. Regrettably, majority of Hausa speakers , whereas an official at the Sudan Interior Mis- braille documents would not, espe- sion School for the Blind was unable cially if their circulation were to to determine who compiled the respec- be s\±)sidized by the Federal Govern- tive systems, or account for the di- ment. Conceivably, therefore, the vergence of braille symbols. proportion of blind Hausa-speaking persons who would become literate in braille might be expected to be very CURRENT USE much higher than that reported for the United States. Captain White's colleague in Ni- geria, Austen N. Chibututu, Admini- One blind rehabilitation worker strative Secretary of the Social De- placed considerable stress on the velopment Division of the Federal fact that the social environment of Ministry of Labour, has asserted that the blind person would counteract the " Hausa braille is in use in . . . Resi- attraction of literacy to them,. Her dential Schools for the Blind as well concern was based on experience in a as in the Open Education Scheme for program for the introduction of a the Blind in the Hausa-speaking areas Spanish braille system in rural Mexi- of Nigeria." Unfortunately, no de- co. In that setting, there was con- tails concerning these institutions siderable resistance to the introduc- have been found to conform or amplify tion of braille on the part of the this statement. There is thus some community, which was principally il- doubt as to whether or not a secular- literate, as well as by blind sub- ized use of Hausa braille, at either jects, who likewise had little desire Grade 1 or Grade 1-1/2 levels, already to become literate, and who refused is current in Nigeria, or the extent to act independently of the local of its use, if any. mores. Rather, the blind preferred that others help them and approve the Because of such contradictory relatively limited scope of their information on this point, therefore, daily activities. The community did the exact status of Hausa braille can- see, in fact, that blind persons were not readily be determined. At best, not completely neglected, and assured its implementation would appear to be that at least minimal needs were taken extremely limited, since opinion on care of. it leads to some contradiction between officials in the Military Governor's There might be some resistance Office (whose information on the situa- to changes in "traditional" Hausa tion at the Nigerian Army Blind Centre attitudes toward the blind at the must be considered to be substantially present moment, but the modernizing reliable) and that of a Federal offi- elements in the culture ought to dem- cial in the more remote Social Devel- onstrate a compensatory receptivity. opment Division in Lagos (although This should be so especially in view his information must of course be of expressions of concern over blind taken seriously) . To resolve this beggars which some government offi- matter, further on-the-spot inquiry cials have voiced already. will be necessary. Without doubt, the blind could The response of some of the blind only benefit from literacy, even if rehabilitation workers at the Insti- they were only to attain a minimal tute for the Blind in Albany, New understanding of basic mathematics York, to the prospects for such a for simple personal transactions, or program has been mixed. We were re- in using braille labels. Its intro- minded that presently only about duction is essential for the primary 5 percent of the blind in the United education of young blind persons (or States today are fluent in braille for its continuance) and, in the case a figure which is startling in its of the Hausa of Northern Nigeria, it limitation — and that this low per- is especially important as a tool in centage was due in part to extensive the meaningful rehabilitation of blind war veterans.

224 To some extent, a knowledge of population on the present basis will braille would free the blind from undoubtedly become increasingly more their absolute dependence on members burdensome for the nuclear or the ex- of the local community. The family tended family, as has been the case and the local community may continue everywhere. For these reasons, blind to maintain and assist the blind, but persons may have eventually to assume this will be only at the present level greater responsibility for their own of deprivation. Any increase in pro- independence. It is to their long- ductivity or standard of living of term advantage, therefore, to take the blind will entail wider involve- advantage of and to adapt to the best ment than local initiative can provide. facilities for rehabilitation that Furthermore, as Nigeria becomes in- can be made available. creasingly modernized and urbanized, the task of coping with the blind

225

. " ,

BLIND PROGRAMMERS AT ROLLS ROYCE — NEWSLETTER 1, JULY 1974

Terry Tate*

Editor's Note. The increasing Centre is November 1973 and follow- number of enquiries about the Rolls- ing training, is now a member of our Royce program to train and utilize Numerical Analysis Group. visually impaired personnel in its data processing activities prompted We took delivery of an M.I.T. Terry Tate to compile the first issue BRAILLEMBOSS in 1971. Connected on- of a Newsletter on their work. He line to our computer complex (AMOS) has very kindly given us permission this device parallels the operation to reproduce it for readers of the of a normal teletype but produces

Bulletin , allowing us to bring it to paged BRAILLE instead of normal the attention of a yet wider audience print. Thus our blind programmers of interested persons . Further is- are provided with the same on-line sues of the Newsletter will be re- computing facilities as their sighted published here as they appear. colleagues

Languages used in the centre in-

clude FORTRAN, KDF9 USERCODE , PDPIO MACRO and a of special This newsletter is produced in number purpose response to the increasing interest in-house autocodes. The system offers interactive and batch operating fea- being shown in o\ir activities in the field of the visually handicapped. tures. Our two KDF9's are to be replaced one 74 19 75. As this is our first issue, we have by CYBER in included some items of historical in- We have produced a grade II terest for the benefit of those who BRAILLE translator (written in MACROIO). may not be aware of our experience in employing blind programmers. A special purpose line formatting programme (DOT) overcome certain limitations of BRAILLEMBOSS. Terry Hicks began to lose his the sight while serving a company ap- Other enhancements include a mod- prenticeship as a production engineer. ification to substitute Standard Eng- He joined the Engineering Computer lish Computer BRAILLE for the American Centre as a trainee programmer in equivalent, which the BRAILLEMBOSS 19 70. He is now competent in a num- originally output, and the provision ber of languages, low and high level, of a switch to suppress the output and has attained the rank of Senior from the teletype, if desired. Systems Programmer. After two years of virtually trouble free operation, we have had Ian Dawson was educated at Wor- some difficulty with the BRAILLEMBOSS cester College for the Blind and sub- recently, due mainly to worn parts, sequently gained an honours degree but thanks to assistance from in Mathematics at Lanchester Poly- prompt MIT and the our engineers, technic. He joined the Computer efforts of our problems have largely been overcome. We are giving a presentation to the British Computer Society on Octo- *Rolls-Royce (1971) Ltd., Bristol Engine Division, Engineering Computer ber 16th 1974 entitled: "Blind Pro- Centre. grammers in Computing.

227

MISCELLANEOUS

ABSTRACTS OF LITERATURE ON CURRENT RESEARCH ON EDUCATION FOR THE VISUALLY HANDICAPPED

Michael J. Tobin

Editor's Note. Dr. Michael J. handicapped children fail to identify Tobin, Director of the Research Cen- causal factors or the nature of prob- tre for the Education of the Visu- lems. In previous studies of EEG ab- ally Handicapped, University of normality, no positive relationship Birmingham, Birmingham, England, between neurologic findings and be- initiated a service for British edu- havioral or intellectual development cators of the visually impaired some has emerged. two years ago which extracted the substance of current research of The developmental charts and interest to them. We thought the EEC's of 38 children with a variety output of this unique service to be of visual handicaps were studied. valuable to a much wider audience The greater the loss of vision, the of users. We solicited, and received, more abnormal the EEC's, but no re- permission from Dr. Tobin to repro- lationship was found between abnormal duce these abstracts in the Research tracings and the occurrence of seiz-

Bulletin . Although documentalists ures or of retarded intellectual de- will be unhappy that the entries do velopment. Abnormal EEC's are not not have descriptors nor other iden- indicators of a poor prognosis for tifying terms, they will be cheered behavioral development. by the news that this data will be added when the material is entered into the forthcoming machine-readable Research Index that IRIS is in the Anderson, C. B., & Rogers, D. W. process of generating. We plan to keep these entries up-to-date by An inexpensive braille terminal publishing them as they are received device. from Birmingham. We welcome your comments on their utility, and, of American Foundation for the Blind course, welcome your additions or Research Bulletin No. 22, 1970, corrections, to the material pub- 111-117. lished. The active use of tim.e-shared facilities for blind programmers requires a braille terminal system. Details are given for the construction of a brailler from a Model 33

Akiyama, Y. , Parmalee, A. H., Jr., & teletype by modifying the print head Flescher, J. and increasing the resiliency of the platen. A description of the pro- The electroencephalogram in visually gramming needed to drive the brailler handicapped children. is presented. Possible improvements are discussed; it is concluded that

The Journal of Pediatrics , 1964, 65, the brailler has proved very effec- 233-242. tive in use, the braille being easy and convenient to read and the relia- The abnormal or unusual develop- bility of the overall system approach- mental behavioral patterns in visually ing that of a non-braille system.

229 , .

Apple, M. M. form of the HIS can be used with confidence both in the research studies Kinesic training for the blind: A of groups and in clinical evaluations program. of individual intelligence; the authors feel that the obvious advantage Education of the Visually Handi- of the abbreviated form is its time- capped, May 1972, 2, 55-60. saving aspect.

Seven congenitally blind adolescents from a residential school for the blind participated in a program Axelrod, S. of kinesics training (interpersonal communication expressed through Severe visual handicap and kinesthetic movements of parts of the body) of figural aftereffects. 6 weeks duration, the purpose of which was to acquaint the subjects American Foundation for the Blind with common facial expressions and Research Bulletin No. 17, July 1968, body gestures. Instruction was con- 1-4. ducted both individually and in a group setting. Observations made of Thirty-three severely visually the experimental group and the con- handicapped (legally blind) high trol groups at the conclusion of the school students were compared with program revealed that the experimental 33 normally-sighted students in amount group did not show a greater frequency of kinesthetic figural aftereffect of "test" gestures (i.e., new ges- (KFAE) at 0, 1, 2, 4, and 6 minutes tures learned during the program) after a 60-second inspection period. than either the sighted or blind con- KFAE for both groups were negatively- trol groups. The author feels that accelerated decreasing functions of the negative test results do not nec- time since inspection. KFAE for the essarily invalidate the program, how- visually handicapped group were sig- ever, difficulties in methodology and nificantly smaller than for the con- procedure could well be masking what trols. Differences due to degree of may be an effective program for the residual vision, age at onset of vi- congenitally blind. sual loss, and sex, were not significant.

Avery, C. D. , & Streitfeld, J. W. Barraga, Natalie. An abbreviation of the haptic intelligence scale for clinical use. Increased visual behavior in low vision children Education of the Visually Handicapped

May 1969 , 2, 37-40. New York: American Foundation for the Blind, 1964, 199 pp. The Haptic Intelligence Scale (HIS) , which had been administered to Through a controlled experimental 32 students at a residential school approach to specialized teaching, for the blind, was rescored for each braille reading children with low subject using an abbreviated form of degrees of visual acuity (6/200 or the test. Each of the 6 si±)tests of less) were visually stimulated for a the HIS was rescored, but on 5 of the short period in an attempt to increase 6 subtests only 46 percent of the their visual functioning despite • their items were utilized; the 6th utilized assumed lack of usable vision. Highly 100 percent of the original items. Re- significant gains in visual efficiency sults indicated an almost identical were demonstrated by all experimental distribution with that of the HIS, subjects. Findings verified the va- and a correlation of 0.99 between the lidity of stimulation programs to en- 2 forms; 9 4 percent (30 out of 32) of hance learning and suggested the need the scores varied no more than 4 IQ for continual appraisal of children points and the largest variation was in regard to appropriate reading media. 7 IQ points. The correlations between A Visual Discrimination Test for mea- the 2 forms of each subtest were also surement of visual abilities and de- found to be highly reliable. The re- tailed lesson plans for a stimulation sults suggest that the abbreviated program are included.

2 30 , .

U. Bateman , B Beermann,

Reading and psycholinguistia proces- Visual exercises and sight educa- ses of partially seeing children . tion .

Washington, D.C.: The Council for The Visually Handicapped , April 1972, Exceptional Children, NEA, 196 3. 1, 15-16. 46 pp. Since the turn of the century The purpose of the study was to teachers in the field of the educa- investigate the effect of visual de- tion of the visually handicapped have fect on the reading and psycholin- engaged themselves more and more in guistic processes of children with a solving the problem of how the vision variety of visual problems. Instru- of the handicapped pupil can be in- ments used were : Monroe Reading Ex- creased. After the theory of saving amination, Illinois Test of Psycho- sight had lost its dominant influence linguistic Abilities; and the Stan- on education, special visual exercises ford-Binet. with mainly medical character were introduced in many schools for the Little relation was found between partially seeing. Partly these exer- the degree of visual defect, reading cises corresponded with those of the ability, and the intelligence of the sight training schools of Bates and children. Apparently, central proc- called forth the criticism of oph- esses are not grossly impaired by thalmologists and educators. With limitations on sensory input; there- the introduction of pleoptic and fore, a justified conclusion is--help orthoptic schools for visual train- each child use centrally what he has ing the intersections between medical peripherally. and educational fields of tasks advanced so far, that a clear marking off seems urgently necessary. Al- though the measures even partly re- Bauman, M. K. semble each other in the two scientific areas they show their own char- Clinical interpretation of personality acteristic accents: visual exercises inventories in pleoptic and orthoptic pursue the aim of a physical improvement in vi- Education of the Visually Handicapped, sual power; the educational exercises (P. A. Vol. 49, No. 6, June 19 73, No. are primarily directed to improve the 10240, Vol. 3(3), October 1971, pp. psychological components of the vi- 82-87.) sual act. Special visual exercises therefore have a justified place in Discusses the interpretation of the education of the partially seeing. personality inventories by clinical As its main aim, however, visual edu- psychologists with special reference cation is to be regarded as a princi- to the A. E.F.I. (Adolescent Emotional ple of instruction, which means that

Factors Inventory) , a measure of ad- in each lesson the efficiency of vi- justment for visually handicapped ado- sual performance should be evaluated lescents. It is proposed that while and stimulated as much as possible. factorial studies have their uses, the clinician works with individuals, not groups; he is concerned with the individual context (i.e. the personal his- Benjamin, J. M. , Benham, T. A., tory, interview data and other test Bolgiano, D. R. , & Meeks , E. D. results) ; and he interprets subscale scores as a configuration. The A laser cane for the blind. A. E.F.I, subscales and scoring system are briefly described. Five cases American Foundation for the Blind, are described in which the interview Research Bulletin No. 22, 1970, data served as validating evidence 139-142. for the A. E.F.I, subscale scores.

A collapsible , lightweight (1-1/3 lbs.) cane, incorporating a laser beam generator and receiving

optics and photodiodes , is described.

231 . .

The device emits pulses of infrared Berla, E. P. light in three beams to look simultaneously down, straight ahead, and Effects of physical size and complex- upwards. The beams are reflected ity on tactual discrimination of from any object in front of them and blind children. detected by a photodiode placed behind a receiving lens. The downward Exceptional Children , 1972, 39, 120- beam warns of any drop-off larger 124. than nine inches at a distance of two paces in front of the traveler; a Studied the tactual discrimina- low-pitched tone notifies the user. tion performance of 36 blind 1st and The straight-ahead beam, about two 2nd graders. Metric figures from feet high, warns of objects at a three levels of complexity (3x3, maximum (adjustable) range of out 4 X 4, 5 X 5) were combined factori- to twelve feet in front of the cane; ally with three physical sizes (1, 2, warning is by means of a tactile stim- 4) in a same-different pair-comparison ulator on the index finger. The up- discrimination task. Increasing lev- ward-looking beam detects obstacles els of physical size and complexity at head height appearing directly had no effect on discrimination accu- above the cane tip, as earlier warn- racy, but increasing levels of either ing proved confusing; warning is physical size or complexity signifi- given by a high-pitched tone. j cantly increased task time. Overall accuracy of performance was low (62 percent correct discriminations) and it appeared that unsystematic and

' Bentzen , B . L _ . inappropriate hand and finger movements were responsible for the low Production and testing of an orienta- level of performance. tion and travel map for visually handicapped persons.

The New Outlook for the Blind, 1972, Berla, E. P. 66, 249-255. Behavioral strategies and problems To test the hypothesis that a in scanning and interpreting tactual tactual map can facilitate independ- displays. ent travel by visually handicapped people in an xmfamiliar area, six The New Outlook for the Blind, 1972, subjects, all proficient blind trav- 66, 277-286. elers, were provided with a tactual map of the canpus of the Perkins Interviews with 12 blind adults, School and asked to travel to each all skilled users of tactual maps, of three objectives. The map, de- provided information on a number of signed to utilize symbols and criteria behavioral techniques and problems found effective by various researchers related to tactual displays of in- (Angwin; Leonard; Nolan; Shiff; and formation, with the aim of ultimately Wiedel) was constructed by hand and increasing the efficiency of the meth- reproduced by thermoform, with the ods used to teach children to use braille labels located on a tactual tactual maps. Subjects were asked to overlay. After training in map use, describe their strategies in locating the subjects, all unfamiliar with symbols and following a track on a Perkins School, were given points for pseudo-map. The discussion includes reaching each of the three objectives description of various scanning tech- in the allotted time, with points de- niques (one-hand, two-hand; horizontal ducted for each "checkpoint" along and vertical; perimeter scan; bounded the route which was missed or each search; spoked wheel, etc.), as well wrong turning made. Of a total pos- as differences in hand and finger uti- sible score of 27, the range of scores lization, micro-motions, distances was 15-22 and the mean score was 19.1. between hands, etc. Types of symbols, All subjects succeeded in reaching orientation, keys, and map "clutter" all three objectives within the time are also discussed. No attempt is limit. Reactions to the map by the made to quantify the information re- subjects indicated that, for a variety ceived from the subjects in the in- of reasons, each found it preferable terviews; rather, the major commonali- to verbal directions ties and differences in the above-

232 ; mentioned areas are discussed. Basic Best, B. D. A. concepts (with accompanying activities) felt by subjects to be neces- School braille--Suggestions for a new sary in teaching children to utilize approach. tactual maps were: meaning of a sym- 113- bol ; lines , points , and areal symbols Teacher of the Blind, 1973, LXI, discriminating bounded space ; repre- 116. sentations of the environment; and spatial relations and memory. Recommendations for making braille easier. Best believes that the Standard English Braille Commit- tee has been too clever and made the Berla, Edward P. rules of braille too complicated-- "It is vital to bring braille into Strategies in scanning a tactual popular use since it is the only med- pseudomap. ium offering the blind a chance to both read and write." School braille Education of the Visually Handicapped, as Best visualizes it, consists of 1973, 5, 8-19. the alphabet and the five extra wordsigns "AND FOR OF THE WITH." The Purpose: (a) to determine whether simpler upper wordsigns represented a brief period of instruction on scan- by letters of the alphabet are re- ning tactual displays increases effi- tained with the exception of k , x, ciency; (b) to determine which search and z. Now k represents "keep" in- pattern is most effective. Method: stead of "knowledge," which is not 108 braillists from blind schools phonetic. x and z are dropped alto- (36 grade 4-6, 36 grade 7-9, 36 grade gether as they do not bear any re- 10-12) were presented with a tactual semblance to "it" and "as" respec- pseudomap with point and linear sym- tively. Initial and final contrac- bols. Three conditions of training tions are out— partly to help spell- (horizontal scan, vertical scan, no ing and partly to do away with argu- training) were combined factorially ments about precedence of contrac- with two conditions of scanning (one tions. A list of abbreviations to hand, two hands) — training being an be retained is included. inter-subject's variable and scanning a within-subject 's variable. Subjects with residual vision were blindfolded.

Measures obtained of (a) total number Bevans , J. E. of objects isolated correctly, (b) number of duplication errors, and (c) time Development of a recreational music taken. Results: Training blind stu- program at Perkins School for the dents to scan in vertical columns re- Blind. sulted in the most accurate and consistent performances compared to a International Journal for the Educa- group trained to scan in horizontal tion of the Blind, 1965, 14, 72-76. rows, or an untrained control group. A high level of performance is obtained A discussion of a music program with one hand and two handed vertical in which the emphasis is on playing scanning, with the latter being fast- and experiencing music, rather than er. Effects of training are moderate on learning about music in its tech- compared to the control group (10 per- nical sense. The program was devised cent increase in number of objects to give a creative outlet to those located, 200-300 percent decrease in students who haven't mastered the variability with respect to duplica- fundamentals of literary braille and tion error) . Control group signifi- therefore cannot receive braille mu- cantly, but moderately, faster. In sical instruction or participate in vertical scan, the use of a reference the school's regular musical activi- hand has no effect. In horizontal ties, but who nevertheless should, scan the reference hand leads to more it is felt, be offered the chance to symbols located, more errors of dupli- enjoy and be successful in performing cation made. music. An outline is given of the programs in both the lower and upper schools. The lower school program involves extensive rhythm band work

233 . ,

and musical games (for the younger Blasch, B., Welsh, R. L. , & David- children) ; the older children develop son, T. more sophisticated rhythm techniques and also work with flutophones. Op- Auditory maps: An orientation aid portunities for performance are numer- for visually handicapped persons. ous and much use is made of tape recorders as teaching aids. The upper The New Outlook for the Blind, school works exclusively with re- 1973, 67, 145-157. corders; all fingerings are taught verbally. Any learning of the tech- Auditory maps , recorded on cas- nical aspects of music is incidental sette tapes, can be used to provide and thought to be of secondary im- a visually handicapped person who is portance. The discussion includes fully trained in the use of long-cane several ideas for adopting the Perkins or dog-guide mobility skills with a program to the needs of a particular verbal description to orient him to school a specific travel area (district map) or step-by-step instructions to guide him to a particular objective (route

map) . It is suggested that the in- Blackhurst, A. E., Marks, C. H., & formation included on such maps can Tisdall, W. J. best be prepared by an orientation and mobility specialist and that the Relationship between mobility and di- content of the map can best be ex- vergent thinking in blind children. pressed using one or more of the following orientation reference systems: Education of the Visually Handicapped ego-centric, topo-centric , carto- 1969, 1, 33-36. graphic, and polar-centric. Sugges- tions are made for the use of spe- In order to test the hypothesis cific mobility techniques and re- that high mobility in blind children cordings of particular sound cues might have a positive relationship that might be encountered. Sample with divergent thinking, 76 blind scripts for auditory district and children from day school programs and route maps are included. The rela- 76 blind children from residential tive merits of auditory versus other schools were rated by their teachers types of maps are fully discussed, on mobility in the classroom, in the as are the wide variety of potential school, and on the school grounds, uses of such recorded travel aids. using a 7-point scale. Sixteen subjects from 6 tests of divergent thinking were then administered to the subjects. Results from the day school Bledsoe, C. W. group showed that 40 of 41 correlations were significant, but all were For parents looking ahead to future needs of their blind chil- low (between 0.23 and 0.33) ; and the mobility author suggests that although this dren. may be statistically significant, it may not be psychologically meaningful. International Journal for the Educa- With this in mind, and since the high- tion of the Blind, 1963, 13, 13-16. est significant correlation obtained was 0.33, it was concluded that any The author discusses the bene- relationship which does exist between fits of mobility training for the mobility and divergent thinking in blind and feels that parents should day school children is at best a be urged to put their blind children slight one. in the hands of competent instructors for such training. A brief history For the residential school group, of the development of mobility instruc- only 3 of 123 correlations were sig- tion is given, along with some of the nificantly different from zero; it arguments advanced by critics of such was thus concluded that for this group instruction. The author suggests no relationship between mobility and that one of the most important bene- divergent thinking existed. fits which independent mobility gives to the blind is privacy—being able to travel without the inevitable accom- paniment of another person— and it is

234 benefit, he feels, which outweighs surface and an illuminated area con- the arguments against the need for taining two pieces of glass--one plain mobility training. and one frosted. Designs to be copied were placed between the two pieces of glass and the child received a light reward as he performed tasks. Con- Bohman, R. V., Bryan, W. H. and clusion: (i) Visually handicapped Tapp, K. L. retarded children can and must learn to attend to objects and tasks before The auditory quiz board: An orienta- they can be expected to develop visuo- tion and mobility game for visually motor skills. (b) Restricted visual handicapped schoolchildren. environments which direct the child's attention to the task are effective The Neu Outlook for the Blind, 1972, in initiating purposeful performance. 66, 371-373. (c) Light is a powerful reinforcer for these children. (d) The effec- Describes a tool devised by and tiveness of a technique can be evalu- used at the Wisconsin School for the ated by monitoring the appearance and Visually Handicapped. It uses multi- spontaneous use of the skill by the ple choice questions, brailled or child. printed onto a wired form which is clipped to the surface of a battery- operated box designed to buzz when contact is made between a small probe Bottrill, J. H. and the correct answer to the question. The board has proved to be a Difference in curiosity levels of good instrument for teaching cardinal blind and sighted subjects. directions and other basic environmental and geographical concepts Perceptual and Motor Skills, 1968, preparatory to advanced orientation 26, 189-190. and mobility training. In order to determine whether blind people are less curious than Bongers & Doudlah, sighted and whether curiositv be- Techniques for initiating visuo-motor havior is influenced by the degree behavior in visually impaired retarded and severity of blindness, a group children. of 28 blind subjects (16 students and 12 older non-students) and a group of the Visually Handicapped, Education of 43 sighted subjects (33 students and IV(3) , 1972. 10 older non-students) were tested Initially, these authors felt for locomotor curiosity and perceptual that the major developmental problem curiosity, using the pyramid maze and of the visually handicapped, multiply a modification of the Stimulus- handicapped child was one of failing variation Seeking Scale, respectively. to extract and integrate sensory in- Results of the perceptual curiosity formation from transactions with the test showed the sighted subjects to environment--but such transactions be significantly more curious than in an institutional ward are routine the blind subjects, but the locomotor and task specific and there are few test showed no such significant dif- opportunities for purposeful, multi- ference. No significant variation sensory learning experiences. Harnon was found among blind subjects as to had found that retarded visually han- source or degree of blindness . It is dicapped children had a strong desire suggested that these results indicate to look into a light source. A high that while stimulation from the ex- intensity lamp was used to direct the ternal environment may be the primary child's attention in a dark room. motivating factor in perceptual curi- Tasks learnt could be generalized osity, the motivation toward locomotor by opening the door gradually till stimulus-seeking behavior may be in-

they were performing in a uniformly ternal . lit area. The lamp was also used as a manipulative object — the child was taught to switch it on and off. This was generalized to pushing the button on a water tap. Next he used a portable light table with a slanted work

235 .

Bottrill, J. H. refused to answer questions relating to sex on the Worchel Scale, and Locomotor learning by the blind and these items were omitted in all sighted. cases; the author suggests that as the self-concept appraisal was thus Perceptual and Motor Skills, 1968, incomplete, this may have accounted 26, 282. for lack of significant differences. He suggests further that although Thirty-three blind and 33 sighted the significant t ratio for the ad- subjects were tested on their ability justment class may indicate that the to learn a raised, semi-linear U- test is an inappropriate measure of finger maze, in order to determine improvement, the smallness of the whether blind people's continual need sam.ple argues against too much reli- to locate themselves in their envir- ance being placed on the results. onment leads to the development of a "locating skill" which sighted people do not possess. Each subject's score was the total number of mistakes be- Bourgeault, S. E. fore he learned the maze, or before the end of the 20-minute test period, Self-imposed limitations in creative whichever was shorter. A t-test on teaching. the learning scores showed a significant difference between the blind and International Journal for the Educa- sighted. Only 34.4 percent of the tion of the Blind, 1963, 12, 115- blind and 39.2 percent of the sighted 116. subjects managed to learn the maze within the allotted time. The author The author describes five symp- suggests that this, and the lack of toms of "professional rigidity" which group differences, might lead one to he suggests limit the performance of assiome that blind people do not have potentially creative teachers of the a better cognitive idea of their sur- visually handicapped: 1) strict ad- roundings than sighted people; how- herence to certain presentation tech- ever, the maze used in the experiment niques; 2) prolonged use of "success- may provide insufficient cues, as it ful" techniques; 3) tendency to aban- requires only simple arm and finger don too quickly techniques which have movements and thus limits the amount failed; 4) idealization of objectives, of kinesthetic sensations available, both for children and for the teachers and these sensations may be of great themselves; 5) the stereotyping by importance to the blind in spatial teachers of the visually handicapped location. children they teach. It is suggested that a constant refusal to identify emotionally with specific techniques and procedures can lead to a more Bottrill, J. H. flexible approach to teaching and ultimately to a more creative solu- Effectiveness of an adjustment course tion to the problems of teaching for those recently rendered blind. methods

Perceptual and Motor Skills, 1968, 26, 366. Brambring, M. In an attempt to evaluate an adjustment course for the newly-blind, Technical and practical utilization 7 subjects and 2 control groups, one of electronic mobility aids for the of 5 students blind since birth, and blind. the other of 26 sighted students, were given orally the Neuroticism American Foundation for the Blind, Scale Questionnaire (NSQ) and the Research Bulletin No. 26, 1973, 257. Worchel Self-activity Inventory, at the beginning and again at the end of Three matched groups of 8 sub- the 1-month course. Computation of t jects each (blind, between 14 and ratios revealed that the only signifi- 22 years old) were assigned to the cant ratio was on the NSQ for the ad- Kay Sonic Aid training scheme, the justment class (P < 0.01) 5 of the 7 Laser Cane training scheme and a subjects of the experimental group control group with no training.

2 36 . ; , — ,

Training was 20 hours individualized Brothers, R. J. instruction, locomotor proficiency being measured at three stages of Arithmetic computation by the blind learning. Conclusions: (a) locomotor A look at current achievement. proficiency is substantially influenced by training; (b) extent of Education of the Visually Handicapped training determined level of success 1972, 4, 1-7. attained in operating the aids (c) no positive effects were demon- In order to determine: (1) the strated for the Kay Sonic Aid in current achievement level of braille these circumstances; (d) given com- students (in U.S.A.) in arithmetic parable training, the Laser Cane led computation; (2) the difference, if to objective and subjective improve- any, between achievement scores of ment of performance; (e) however, braille students on the SAT test of these improvements appeared to depend arithmetic computation for 1959 and on training effects rather than on 1970; and (3) possible relationships the electronic system; (f) thus the between SAT scores and type of compu- correct procedure seems to lie in ex- tational device or strategy used by tensive mobility training to optimally braille students, 269 blind students exploit the natural capabilities of aged 6 to 20 were tested using the blind people; (g) further attempts to SAT (Scholastic Aptitude Test) . Re- design technical aids should incor- sults indicated that the achievement porate human factor principles , and levels of the 1970 group were con- take into account man's psychological sistently below those of 1959, al- characteristics though at the higher grade levels students attempted more responses in 1970 than in 1959. A relationship was found between achievement and Brothers, R. J. type of device or strategy used in computation for the 19 70 8th-grade Aural study systems for the visually group only, where 43 percent of the handicapped: Effects of message length. students using an abacus achieved at or above the expected grade level of Education of the Visually Handiaappe d 8.2. Educational implications are

3(3) , 65-70. discussed.

Assigned 6 braille readers and 4 large-type readers--blind high school students — to each of four Brown, G. D. , & Jessen, W. E. groups. Subjects heard a 24-minute tape in 1, 2, 3, or 4 sections and Preliminary performance test battery were immediately tested after each of orientation, mobility, and living section for information recall. skills . Seventy- four hours later, subjects were again tested for recall. Re- American Foundation for the Blind, call was not significantly affected Research Bulletin No. 24, 1972, 1-20. by message length but immediate recall was significantly more accurate Sixteen blind children, median than delayed recall (p < 0.01). age 13, attended a three-week work- Braille readers had significantly shop of daily classes in orientation, (p < 0.05) higher scores than the mobility, and living skills. Twenty- large-type readers on delayed recall, one preliminary performance tests Limitations of the study are noted. were developed to evaluate the workshop. Students achieved significant increments in performance on four of the living skills tests and on one of the orientation mobility tests. Staff members were in close agreement when ranking students for proficiency and improvement in orientation and living skills. Comparison with performance measures indicated that staff gave almost equal weight to orientation and mobility and to living skills when

237 i . , ranking students for proficiency. light reflected from the screen is There was little relationship between not sufficient, they are provided subjective ranks for improvement and with desk lights. The lens is fo- distributions of differences obtained cused for the majority of students, from pre- and post-test scores. The then re focused for those who require feasibility of developing a perform- it. Those sitting near the screen ance battery of orientation, mobility can only see the bottom half clearly and living skills was demonstrated. so the lens housing is tilted to fo- A preliminary manual for the OMTB cus the whole image on the bottom is presented in an appendix. half of the screen. Students with very low vision stand next to the projector and look directly at the film (or at the wire, geometrical

B rown , J . D . objects which can be projected onto

the screen) . Blunt felt tips are Storytelling and the blind child. used to make the lines thicker and easier to see. The Overhead Projector The New Outlook for the Blind, 1972, has been found to focus the attention 66, 356-360. of the students and is flexible and adaptable to the needs of the teacher Obtained 72 questionnaire re- and pupil. There is a range of in- sponses from 160 mailed to librarians, struments which can be used and the teachers and specialists. Results simultaneous use of auditory and unequivocally indicate that story- visual modes improves retention of telling is an effective tool when material taught. properly utilized for aiding children's mental development. It is noted that story tellers must realize the importance of their voices and Calek , . differences in reactions of blind children as against sighted ones. It Distressful life situations and the is suggested that story tellers emotional strain felt by subjects (a) know their audiences beforehand with severe visual defects. in regard to average age and attention span of the group, and home and American Foundation for the Blind, school interests; (b) develop themes; Research Bulletin No. 25, 1973, (c) select suitable books; (d) read 59-69. material several times aloud and commit to memory; (e) make use of in- Eleven subjects between 18 and flectionary and modulatory voice and 20 years old, all totally blind, psy- tones; and (f) encourage audience chologically naive, prospective music participation. A list of 37 stories teachers. Several problems relating enjoyed by blind children is in- to the inner life of the visually im- cluded. paired were studied. An attempt was made to describe the emotional strain in the life of the severely visually handicapped and to discover whether Bruce, R. it was as stressful as the objective visual deficiency. The causes of Using the overhead projector with emotional strain were sought among visually impaired students. the stresses, frustrations and conflicts of their life situations. Education of the Visually Handicapped Their sensitivity to the so-called 1973, 5, 43-45. higher or social needs was analyzed and compared with frustration levels For six years the ove rhead pro- at the level of lower, biological jector has been used for h igh school needs. Investigation was made into maths at the Virginia Deaf-Blind the character of modes of coping with School. The classroom is darkened to stressful life situations. Special cut out the glare which is troi±>le- attention was paid to defense reac- some to partially sighted children tions showing clearly indications of Students move their desks to places maladaptation. The causes of physi- suitable to their needs — n some cases cal appearance and the ways they were between the projector and screen. If acquired were also investigated.

238 . . .

Lastly, the role of the life history Chase, J. B. , & Rapaport , I. N. of the visually impaired was considered, including the influence of per- A verbal adaptation of the draw-a- sonality traits on modes of coping person techniques for use with blind with stress situations and on the subjects: A preliminary report. specificity of behavior. International Journal for the Educa- tion of the Blind, 1968, 18(4).

Carolan, R. A verbal adaptation of the Draw- A-Person Test (VDAP) was administered

Sensory stimulation in the nursing to 75 subject s . An equal number of home . sighted subje cts took the DAP and VDAP and thre e judges later blind- The New Outlook for the Blind, 1973, matched a ran dom sample of ten of 67, 126-130. these drawing s with their verbal counterparts Correct matchings were To continue functioning adap- found to be s ignificant beyond the tively, the individual needs constant- 0.001 level, suggesting a congruence ly varying forms of sensory stimula- of clinical i nterpretation of both tion. Since such stimulation is often instruments Further validation re- neglected in the nursing home envir- search is ant icipated. onment, producing particularly devas- tating psychological effects in the sensorily impaired person, specific suggestions are provided for improv- Cicenia, E. F. , Belton, J. A., ing this situation. Orientation and Myers, J. J., & Mundy, G. mobility specialists can be invaluable catalysts in providing the skills The blind child with multiple handi- and motivation necessary for enrich- caps: A challenge. Part II. ing such an environment. Specialists and regular staff can, by analyzing International Journal for the Educa- the person's entire daily routine, tion of the Blind, 1965, 14, 105-112. provide many opportunities for additional sensory stimulation. Johnstone Training and Research Center has developed a continuum of therapeutic educational experiences for multihandicapped children. Carolan, R. Failure of the educational pro- Sensory stimulation and the blind in- gram was recognized in the beginning fant. due to the emphasis on the "three R's." Likewise, an emphasis on self- The New Outlook for the Blind, 1973, care and activities of daily living 67, 119-126. was unsuccessful because the emotional needs such as security promotion, To develop properly, it is neces- acceptance promotion, fear allevia- sary for the infant, and particularly tion, and human interactions were the infant who is blind, to be pro- necessary before learning as such vided with a rich environment of sen- could begin . An educational se- sory stimulation. When such stimula- quence of management skills, motor tion is absent, as evidenced by the experiences, selected academics and findings of researchers in sensory group integration has been devel-

deprivation, a variety of undesirable oped . behaviors can result. Parents can be helped in a variety of ways to provide their child with adequate stimulation. The concept of an infant curriculum, as developed by Barsch, is suggested as a particularly useful means of insuring that appropriate kinds and amounts of sensory stimulation are provided at each developmental stage

239 Clegg, G. Condricek, R. , Meehan, F. , & Love, J.

A study of the tactual discrimination A new braille medium. test for measuring tactual ability of the visually handicapped. American Foundation for the Blind, Research Bulletin No. 25, 1973, American Foundation for the Blind, 69-94. Research Bulletin No. 25, 1973, 259. A report on the development of Hypothesized that level of per- a new system converting coded ink- formance on TDT would be positively print into braille. There are three related to ability to discriminate components: (a) The ink-print medium braille characters. Tested 52 legally --composed by computer, typeset by blind participants who had completed photocompositer , printed by photo a course in braille reading. They offset. A standard 3x2 braille had to sort 56 pieces of sand paper cell is coded into a 1 x 6 array. of various sized grits (14 each of Each dot is contained in a square of extra fine, fine, coarse, extra side 1/100". To allow separation of coarse) into a box. This box was lines, the top half of the square is sectioned into 4 small bins , one large white. Presence of a dot is repre-

(closest to the subject) , the box be- sented by black bottom left hand ing waist high at an angle of 30 de- quarter, white bottom right hand grees. A sight occluder was used. quarter. Conversely for absence of T%r7o measures--time of sorting and dot (i.e. 25 percent of the square number of correct responses were is always black — this facilitates found to be significantly correlated tracking) . A standard sheet of coded with braille rankings (p < 0.01). ink-print can contain more information than a standard volume of braille --an enormous reduction in bulk and inconvenience. (b) The optical read- Coffey, J. L. ing device. This reads one line at a time, collects groups of 6 dots, Programmed instruction for the blind. holds a number of them and feeds them into a transducer. This device will International Journal for the Educa- change line automatically and read tion of the Blind, 1963, 13, 38-44. in either direction. It also provides the number of the media line being A series of objectives to explore read to the transducer. (c) The the possibilities of programmed in- transducer. This converts the elec- struction for use of blind students trical impulses generated by the op- is outlined by Battelle Memorial In- tical reader into a form acceptable stitute. Evaluation of an experi- to the blind reader— so far attention mental program led to these conclu- has been concentrated on braille. sions : (a) programed instruction is Bumps are activated to simulate em- feasible and as desirable as for bossed braille and pass beneath the sighted students; (b) programs written reader's fingers. A string of char- specifically for the blind should be acters is continuously available to tested; (c) verbal responses were bet- the reader who can retrace misunder- ter for junior high students, but stood words by backing up. braille responses for senior high students are effective; and (d) mode of presentation (aural or braille) is not significant. Conner, G. B.

Blindness 1964.

Washington, D.C.: American Associa- tion of Workers for the Blind, Inc., 1964, 175 pp.

The title of every government sponsored research project relating to blindness is listed along with a short resume of the study and the

240 . , . , principal researcher as well as pub- predictor of academic achievement. lication sources. Until such data is available, the Perkins-Binet should be a very use- A variety of definitions from ful resource for the clinician, pro- numerous 'countries provides strong viding pertinent educational data support for an examination of de- not accessible through the use of fining procedures and highlights the verbal instruments such as W.I.S.C. pressing need for a more communica- and Haynes-Binet. tive vocabulary in regard to visual handicaps. A realistic discussion of the newer concepts in mobility precipitated by technological advances Craig, E. M. places a new perspective on travel techniques and devices. Variables Role of mental imagery in free re- other than the objective analysis of call of deaf, blind, and normal the amount of residual vision must subjects be considered in mobility training of persons with partial vision. Journal of Experimental Psychology 1973, 97, 249-253.

Tested A. Pavio's dual coding (see Psychological Ab- Coveny , T. hypothesis stracts, Vol. 43: 10753) in an ex- A new test for the visually handi- periment with 40 undergraduates, capped: Preliminary analysis of the 40 deaf adolescents and 40 blind reliability and validity of the adolescents. It was hypothesized Perkins-Binet that the deaf store information almost exclusively in a non-verbal Education of the Visually H andi capped code, while the blind primarily use 1972, 4, 97-101. an auditory-motor code (verbal). Differential retention effects were Subjects: 55 students (grades therefore expected for the two groups differ in the 3-6) : 30 braille readers 8-9 to 14-9 for words which seem to age range) and 25 print readers (age ease with which they elicit non-verbal range 8-0 to 16-5) , all at blind imagery. All the groups recalled school. The Perkins-Binet has two high-imagery words better than low- forms — form N (no useable vision) and imagery words; this was predicted not for form U (useable vision) . These were from the normal and deaf but administered to the appropriate the blind. A significant Group X groups together with the W.I.S.C. imagery level interaction was found (half had the W.I.S.C. first and for the normal-blind comparison, but half the Perkins-Binet first in each not for the normal-deaf comparison. the group) . The reliability of the test Serial position effects support was tested with a split half relia- dual coding hypothesis. bility technique. Results: Both forms of the Perkins-Binet showed a high degree of internal consistency and the highly significant correla- Cull, J. C, & Hardy, R. E. tions between the Perkins-Binet and the W.I.S.C. show that both scales Language meaning (gender shaping) are measuring, to some extent, the among blind and sighted students. same abilities. But the Perkins- Binet and W.I.S.C. scores are not Journal of Psychology , 1973, 83, interchangeable, standard deviations 333-334. on the Perkins-Binet being larger than with the W.I.S.C. Coveny recom- Examined whether substantially mends reporting Perkins-Binet scores different word meanings exist within as mental age rather than I.Q. The the language systems of blind and Perkins-Binet appears to be a valu- sighted students. Twenty-two totally able clinical instrument, giving in- blind and 64 sighted high school stu- formation about degree and quality dents were randomly selected and ad- of residual vision as well as tactual ministered the Gender Association perception. More research is needed, Survey. Blind subjects showed sig- though, to determine its value as a nificant differences (p < 0.05) from

241 . their sighted counterparts in assign- Davidson, P. W. ing gender to 17 out of 50 commonly used words . Language seems to help Haptic judgments of curvature by blind condition perception and people who and sighted humans are deprived of vision seem to have different interpretations concerning Journal of Experimental Psychology , the meaning of words in the language. 1972, 93, 43.

Three experiments investigated the relationship between active hand- Curtis, J. F., & Winer, D. M. ling and veridical haptic curvature perception. Experiment 1 showed that A comparison of the efficacy of two blind subjects made more objective methods of mobility training for the judgments of curves than did sighted blind, using blindfolded sighted sub- subjects. Videotape recordings of jects . exploratory scanning during the judgments revealed that the blind subjects American Foundation for the Blind, used a scanning technique allowing Research Bulletin No. 22, 1970, more global apprehension of the stimu- 119-129. lus parts. In experiment 2 sighted subjects' judgments became more ob- Two methods of mobility training jective when they were restricted to currently being used with the blind, using the scanning technique charac- tactile maps and binaural recordings, teristic of the blind. Experiment 3 were evaluated by measuring the travel suggested that sighted subjects' er- performance of 7 groups of blindfolded rors in experiment 1 probably stemmed sighted Ss following training with from using movements that obscure various combinations of training stimulus curvature. It was concluded methods . The Ss ' travel performance that the blind's scanning technique was measured on a one-block course in (a) focused attention to an informa- a small town. Each S made 2 round tive stimulus feature, and (b) pro- trips on this course. A balanced de- vided a way to pattern the ends-to- sign was used, with some groups re- middle relationship that specifies ceiving no training, some receiving curvature. Implications for a theory only one type, and some receiving of the role of exploratory activity both types , either between the 2 in perception were discussed. trips , or prior to them. The results indicate that there is no statistically significant difference in travel time or path on the sidewalk among Davidson, P. W. any of the groups. The role of exploratory behavior in haptic perception: Some issues, data, and hypotheses.

Dauterman , W. American Foundation for the Blind, Manual for the Stanford multi-modality Research Bulletin No. 24, 1972, test. 21-27.

New York : American Foundation for the Reviews recent studies of tactile- Blind, 1972. kinesthetic perception. The accuracy of such perception is discussed in Details the s tandardization terms of focus-of- attention, scanning studies for the Stanford multi-modality materials, autogenetic development, test which measure s the blind person's and experience factors. It is felt ability to use ima gery in solving mo- that understanding of these inter- bility and other d aily living problems, relationships may result in improved Methods for scorin g and interpreting information gathering and coding ac- are presented. In dexes on testing the tivities of the blind. deaf-blind and on the blind examiner are included.

242 . . . . . , —

Davis, P., Asarkof, J., & Tall- Dolan, W. man , C . B The first ten months of the rubella A closed-circuit television system as living unit. a reading aid for visually handicapped persons The New Outlook for the Blind, 1972, 66, 9-14. The New Outlook for the Blind, 1973, 67, 97-102. Describes operation of a Rubella Living Unit established to halt re- Although most visually handi- gression and teach basic self help capped patients can be adequately skills to a group of pre-school deaf, rehabilitated visually through the blind, mentally retarded children, prescription of low vision optical mostly confined to beds in a residen- aids, some who have very poor vision tial school. Originally 6, then 10, and who need to do a great deal of children with 9 staff members working reading may be helped through the 12 hours, 7 days a week formed this prescription of a closed-circuit tele- unit. Improvement of the child's life vision reading system. It was found style, helping him reach his maximum that 6 of 17 patients tested with such potential and determining the most a system could benefit from it. The effective teaching techniaues with system used includes a push-button this type of child were the objec- operated scanning device for shifting tives. Means for achieving them were the image on the page instructing the children in daily living skills, enlarging experiences by field trip and improving interactions with peers and staff and evaluating Dickey, T. W., & Vieceli, L. the educational potential. Six case histories are presented, emphasizing A survey of the vocational placement that such children are able to learn of visually handicapped persons and if someone believes in the child's their degree of vision. potential and gives enough of himself to the learning situation. The New Outlook for the Blind, 1972, 66, 38-42.

In order to determine the occu- Drever, J. pational areas in which visually handicapped clients are being placed, Early learning and the perception of the range of visual acuity among those space being placed, and the kinds of questions employers were asking during job American Journal of Psychology development interviews, 225 question- 1955, 68, 605-614. naires were mailed to graduates of the Placement Counselor Training Program Subjects: 2 groups of 37, matched of Southern Illinois University. The for age, sex and intelligence— one 77 usable returned questionnaires blind, one sighted. Average age placed a total of 1733 clients in 18 14-11, 15-3, respectively. Average specified occupational areas and in intelligence 111.7, 116.1 respectively. the category "other" for the period The blind group was subdivided into 1 July 68 - 15 April 70. The occupa- 2 matched groups — early blind (before 4 tional area, "industrial," showed the age 4) , and late blind (after age the ex- largest numbers of placements (600) , but at least 2 years before

. were with self-employment second (191) , and periment) Sighted subjects A clerical third (160) . With respect to blindfolded for the 4 tests. (1) vision, 43.4 percent of those placed Figure Recognition Test. Subject had visual acuity of 20/200 or better, given 2 parts of a figure (circle, 42.9 percent had 20/200 or less but semi-circle, etc.) to handle separ- were not totally blind, and 13.7 per- ately but not to put together. Sub- cent were totally blind. Among the sequently he had to decide which of legally blind persons, 24.3 percent 4 figures would have been made from were totally blind, a figure which the parts he had. In this test, the closely approximates the 25.5 percent sighted were slightly superior to the of totally blind persons in the legally late blind, who were far superior to blind population who are 20 to 60 years the early blind. (2) An Orientation of age 243 . . i

Test. This involved a 10 x 10 peg- other than visual function appeared board with 1" between holes, pegs in 16 percent of the si±ijects. sticking out 1/4", Subjects had to feel a configuration of pegs, hold on to the last one while the board was rotated through 180° and then, Dumas, G. after the pegs had all been removed, replace them in the holes they occu- Mimicry of the blind. pied before rotation. In this test the late blind were considerably su- . . . And there was light, 1932, 2, perior—with the early blind and 30-33 (in French) sighted very close to one another. Psyahologiaal Abstracts , 1933, 7, Test. (3) A Short Classification Item 2205. Subjects had to choose the odd one in the out of three configurations "To find out whether in mimicry were similar peg-board— two of which we are imitating ourselves, by means numbers in shape but had different of consciousness of our spontaneous of pegs, two of which had equal num- expression through our muscular and bers of pegs, i.e., .*./.*./. '. cutaneous sensibility, or are imi- shape, Here, the blind classified by tating someone known to us visually, by number. The sighted the sighted some 30 observations have been made classified by when vmblind folded, on people born totally blind. The shape like the blind. (4) A Test results indicate that mimicry is of the Tactile-Kinesthetic Percep- based on visual observation although tion of Straightness . This showed possibly the blind might be taught a that the blind are superior to the certain amount of mimicry through con- sighted in both objectivity and tactual observations." sistency. In none of the tests was age of onset after 4 significant, nor age, nor intelligence. The tests seem to indicate that some skills are Duran, P., & Tufenkjian, S. built up early, and later learning has little effect. The measurement of length by congenitally blind children and a quasiformal approach for spatial concepts Duche , D. J., Rausch de Traunen- berg, N. , & Bouras , A. American Foundation for the Blind, Research Bulletin No. 22, 1970, Value of psychological techniques in 47-70. the examination for detection of vi- 3-6 old sual disturbances in the year The major task of this study was child. an analysis and description of the use of the haptic sense as it bears Psychological Abstracts, 1973, 49(4), on the blind child's perception of No. 7129. length. Congenitally blind children were chosen as Ss in order to isolate Studied 84 3-6 year olds to de- those perceptions derived through tac- help in termine factors which might tile learning. This paper presents a validating ophthalmological screening quasiformal approach and point of Stan- examinations. Results of the view for investigating spatial con- Form L, ford-Binet Intelligence Scale cepts and describes an experiment and of the Columbia Mental Maturity inculcating this point of view. In Scale show that two types of children the quasiformal approach the behavior in- present difficulties which might of an organism is differentiated into validate ophthalmological examina- four parts: input, memory, logic, and tions — those with slightly deficient display. It was found that the major intelli- I.Q.s, or those with normal classes of methods for measurement gence but severely inhibited person- length by congenitally blind chil- general of alities. It is concluded that dren are: juxtaposition and noncoin- psychological maturity and level of cidence of endpoints , body part as perceptual -motor integration determine measuring instrument, kinesthesis, the capacity to adequately respond to time duration, and physical princi- of the test-types. Er- presentation ples. The principal conclusions were rors of estimation due to factors

244 . , , ,

(1) The physical size of the objects Deficient color vision was found in will determine the technique used for 13.25 percent of educationally handi- the measurement; and (2) as the physi- capped and 5.05 percent of regular cal length of the object increases, class (p < 0.05). Hypothesis con- the corresponding difference thres- firmed. hold increases independently of the technique used.

Evans, R. , & Simpkins, K.

Elonen, A., & Cain, A. C. Computer assisted instruction for the blind. Diagnostic evaluation and treatment of deviant blind children. Education of the Visually Handicapped 1972, 4, 83-85. Ameriaan Journal of Orthopsychiatry 1964, 34, 625-633. Fourth, 5th, and 6th graders use computers in math lessons at a school Visually handicapped children in Philadelphia. The program has 6 with varying types and degrees of grades and the child goes through at developmental problems are often mis- his own rate, reviewing when neces- diagnoses. Even if diagnosed correct- sary. Each child is first taught to ly, they continue to bewilder parents, use the keyboard on a thermoform psychologists, medical experts, and reproduction. When a child answers educators because of the lack of incorrectly, the computer types treatment plans , programs , or cen- "wrong--try again." After 3 tries ters . the answer is typed in and the student types it again for reinforcement. A therapeutic approach in a psy- He is told at the end how many correct chiatric clinic revealed notable responses he's made. If he has only progress in behavioral development made 2 errors, he goes on, otherwise with eventual entrance into local and he repeats the lesson. The braille residential educational programs for characters are printed out on thin many children. The study verified paper by means of a braille adaptor. that pessimistic diagnoses and prog- It is in grade 1 braille which is noses are not acceptable without difficult for students accustomed to (a) prolonged study and therapeutic grade 2. Also no number signs pre- care; (b) cooperative communication cede numbers which is confusing. and planning of many disciplines is Thirdly the multiplication and di- indicated; and (c) older concepts revision on the abacus is done from garding the influence of visual handi- left to right but the computer must cap on personality development deserve have input and output right to left. re-examination But all these difficulties have been overcome with time and patience and the children are all very enthusiastic. Espinda, S. D.

Color vision def iciency--A learning disability? Flanigan, P.

Journal of Learning Disabilities Automated training and braille read- 1973, 6, 163-166. ing.

Hypothesized that color vision The New Outlook for the Blind^ 1966, deficiency would be associated with 60, 141-146. impaired learning, inappropriate classroom behavior and a higher fre- Braille reading of 15 experi- quency of referrals to programs for mental and 15 control subjects was the educationally handicapped. Diag- compared as a function of training, nostic color vision screening was con- the experimental subjects using an ducted for 83 males in 11 classrooms automated se lf-learning device,* for the elementary level educationally *This introduces braille on a tape handicapped and for 139 males in 11 moving from right to left under the regular 3rd and 4th grade classrooms. subject's fingers.

245 ,

and control subjects reading tradi- blind as do the motor patterns for tional braille. Analyses of the data fighting and the ability to discharge indicated (1) performance of experi- anger. A 14 month old not yet creep- mental subjects was significantly- ing, after separation from his mother superior to that of control subjects for 3 days developed pathological on certain variables (retracings, ver- screaming and regressed. His mother, tical movements) involved in the tra- herself somewhat unstable, was help- ditional braille reading process; less. When rage was recognized as (2) following a non-instructional one component of the screams, and the period of 3 months , the experimental child was encouraged to use banging gains remained constant; (3) non- and pounding toys, screaming stopped significant treatment effects were and normal development resumed. For noted on cortprehension and reading a 19 month old, all factors were more achievement level as a function of favorable. He was precocious, espe- the method of reading; and (4) the cially in language development, mo- appropriateness of sequenced grade- bile, and could follow his mother level reading materials for reading about. Substitute care during the instruction was pointed out. Sugges- period of separation was superior to tions for future research were given. the previous case. His mother was more skillful in translating his messages and in making the appropriate response. No regression or unfavor- Fonda, G. able symptoms developed.

Management of the patient with sub- normal vision.

Franks, F. L. , & Baird, R. M. St. Louis C. V. Mosby Company, 1965, 160 pp. Geographical concepts and the visually handicapped. Not only does this book provide

comprehensive coverage of clinical Exceptional Children , 1971, 38, 321- procedures for examination of low 324, vision patients, it also includes a

readable explanation of magnifiers, ' Reports the third in a series of telescopic systems, and visual aids studies exploring geographical concept in general. Four chapters are de- attainment in visually handicapped voted to information which is invalu- students as a basis for improving able to teachers of visually handi- ability to interpret embossed maps. capped children. Covered are such Studies 1 and 2 are briefly summar- topics as (a) classifications of de- ized as background. In study 3, 24 grees of vision, type size, and re- braille and 24 large-print readers lated magnification, (b) training and were tested on three-dimensional instruction in the use of optical aids, raised surface models (land forms) and (c) a valuable question and answer which were tactually and chromatically section which should clarify many mis- coded. Results indicate that the conceptions . landforms provided highly discriminable illustrations of 40 basic geographical terms and the codes furnished effective cues. It is con- Fraiberg, S. cluded that the materials should greatly facilitate both instruction Separation crisis in 2 blind children, and evaluation of geographical concept attainment. Psyahoanalytia Study of the Child, 1971, 26, 355-371.

Describes the factors which ag- gravate the separation crisis for the blind child and his mother. Lack of vision limits the perceptual field within which the child is aware of his mother's presence. Evocative memory develops more slowly in the

246 . —; . ,

which also Friedman, J., & Pasnak, R. field, we need a system accounts for the functional capability Attainment of classification and of an individual. Genensky proposes seriation concepts by blind and four categories classifying on two sighted children. criteria: (a) whether or not a person can maneuver safely in an unfamiliar guide dog, sighted Education of the Visually Handi- environment without whether or not they capped, 1973, 5, 55-62. guide, etc. (b) can read ink print, and write like a feels it un- Subjects: sighted - 10 first sighted person. Genensky fortunate that the legally blind are graders (age 6) , 10 third graders treated as a visually homogeneous (age 8) , 10 sixth graders (age 11) population this creates functionally blind - 21 grades 1-6 (ages 6-14) . — Sighted subjects were blindfolded blind people who could have been en- sighted. in non-verbal tests. Classifica- couraged to be functionally discovery of class, classifica- It also tends to involve stereotyping tion: thought tion of form, size, texture, orienta- of jobs, hobbies, lessons is no tion. Seriation: size, length, weight, suitable for the blind. There functionally seriation of recognizable forms. necessity for any but the Verbal: analogues of discovery of blind to be taught to read braille. new class and seriation of recognizable This group is becoming smaller as "Light forms, transitivity. Three sessions, aids are discovered (e.g., one read ink- in this order, of testing were car- Perception" person can now Those ried out with each child, individual print-x-4 on the C.C.T.V. aid). testing. Results: in general, sight- who can be taught to read and write ed perform more adequately than blind ink-print with whatever aid is re- it is vi- children. (1) Conceptual abilities quired should be taught to— of classification and seriation con- tal to the emotional and mental health tinue to develop in all children at of the visually impaired that they be- possible. least to age 11. (2) Perceptual come as independent as factors intrude so much that they may not be good indices of conceptual ability. Changing perceptual quali- H., & ties (objects used) a little, often Gescheider, G. A., Wright, J. changed error scores considerably Evans , M. B apparently because the children's of mastery of the concepts was incom- Reaction time in the detection signals. plete. (3) Blind and sighted chil- vibrotactile dren are approximately equal in Psychology classification and seriation at 8 Journal of Experimental years of age, but afterwards the 1968, 77, 501-504. blind tend to fall behind, especially made judgments of on verbal tasks. (4) Lack of vision Three subjects burst of impairs performance on visual tasks the presence or absence of a the index finger- as well as manipulative tasks when 200-cps vibration on re- these tasks involve conceptual abil- tip. The probability of subjects was Conclusion: if acceleration of porting the presence of a signal ity. proba- conceptual ability can be achieved by found to be influenced by the signal training children on Piagetian indices bility of signal occurrence and ROC curves of conceptual ability, the blind child intensity. A family of will profit most. describing the effects of signal probability on response probability for each signal intensity level was interpreted as support of the applicability of signal detection theory to Genensky , S the judgment of cutaneous stimuli. signal intensity and classification of the Manipulation of A functional probability also led to changes impaired to replace the signal visually subjects' reaction time for saying legal definition of blindness. in "yes" and for saying "no" when the signal was present and when it was Teacher the Blind, 1973, LXI , 82-91, of absent, supporting the conclusion that subjects' decision time was longer the Rather than a system which classi- closer on the sensation continuum a as blind on the basis of a fies a man sensory observation was to specified visual acuity or visual particular his criterion.

247 ,

Goldblatt, M., & Steisel, I. M. Graham, M. D.

Behaviour modification with multi- Wanted: A readiness test for mobility handicapped blind children. training.

Proceedings of the 81st Annual Con- The New Outlook for the Blind, 1965, vention of the American Psychological 59, 157-161. Association, Montreal, Canada, Vol. 8, pp. 809-810, 1973. The lack of an entrance exam prior to admittance for mobility Notes that although a large and training is a critical factor; with- growing literature exists on the use out such a measure one has no indica- of behaviour modification techniques tion of the degree of predicted suc- with children who manifest various cess. Existing data on 100 blind aberrant behaviours, reports of work adults not associated with agencies with blind and multi-handicapped chil- for the blind were explored. Those dren are rare. The behavioural analy- in the high activity group gave evi- sis treatment plan, and results of dence of certain specific character-

' ' programs designed to modify the be- istics . haviour of 3 children who attend a day school for the blind are present- Regardless of the age of the ed. The pinpointed behaviours were visually handicapped person, multi- assaultive acting out, rumination ple factors appear to comprise the during and following meals, and sleep- mobility patterns; therefore, success- ing during school time. The targeted, ful attack on the problems of mobility objectionable behaviours diminished, for children and adults is dependent pro-social behaviour was acquired and on the development of, experimentation the teachers were able to make more with, and standardization of rational positive, educational approaches to procedures for selection of trainees the children. and evaluation of performance.

Goldish, L. H. f^uldager , L.

The severely visually impaired popu- Using video tape in the education of lation as a market for sensory aids deaf -blind children. and services: Part one. The New Outlook for the Blind, 1972, The New Outlook for the Blind, 1972, 66, 178-182. 66, 183-190. The use of video-tape at the In this , the first part of a re- Perkins School in the education of the view of the question of sensory aids deaf -blind is discussed. The author the author presents the findings of a sees the main advantages of this meth- survey of the potential market for od to lie in the areas of teacher sensory devices among the visually self-evaluation, teacher training handicapped in the U.S. The article (including the development of a tape includes discussions on the defini- library of the various behaviors and tions of blindness , and the factors syndromes of deaf-blind children and to be considered when deciding upon a the techniques used in their educa- particular aid or device, as well as tion) , and the education of the deaf- a large number of statistics describ- blind children themselves, assuming ing the number, activities and char- that the children have sufficient acteristics of the visually handi- vision to see the monitor. In this capped in the U.S. latter area the author mentions video- taping outings to facilitate later classroom discussion, using videotape to present new concepts and give children the chance to experience things normally unavilable in the classroom, giving practice in speech- reading by taping 5-minute talks, and developing video-taped programmed

248 ,

instruction techniques . It is anti- Hapeman, L. H. cipated that in the future the video tape will also be utilized in the Developmental concepts of blind chil- auditory training of deaf-blind children between the ages of 3 and 6 as dren. they relate to orientation cind mobility.

International Journal for the Educa- tion of the Blind, 1967, 17, 41-48. Halpin, G. , Halpin G. , & Tillman, M. The author describes the various Relationships between creative think- concepts necessary to the development ing, intelligence and teacher- rated of mobility and orientation in the characteristics of blind children. blind child, with the aim of giving parents an awareness of these con- Education of the Visually Handicapped^ cepts in order that they might provide 1973, 5, 33-38. opportunities for the blind child to experience them. The concepts are di- Subjects--63 blind children from vided into 3 groups: (1) concepts residential blind schools , 20 from needed for understanding the true na- day blind schools . Purpose — to test ture of the environment (including the relationship between (a) creative body image, nature of objects, nature thinking and teacher ratings on mo- of terrain, etc.); (2) concepts need- bility, adjustment to blindness, so- ed for achievina and maintaining ori- cial acceptance, dependence, conform- entation (including path of moving ity, rigidity, curiosity, academic objects, directions, sound localiza- achievement, (b) I.Q. (WISC; Hayes- tion, etc.) ; (3) concepts needed for Binet) and teacher ratings, (c) I.Q. efficient mobility (including distance and creative thinking. (Creative and time, turning, detouring, etc.). thinking being made up of fluency, The author stresses that it is only flexibility and originality and mea- through concrete learning experiences sured by the Torrance Tests.) Re- that a blind child will develop these sults-- (a) Only curiosity was related important concepts. to creative thinking on all three dimensions. Rigidity and academic achievement correlated with flexibility, (b) Only curiosity is cor- Hare, B. A., Hammill, D. D. , & related significantly with I.Q. Crandell, J. M. (c) I.Q. is only correlated with flexibility. Conclusion--the blind Auditory discrimination ability of child scoring high on verbal flexi- visually limited children. bility is more likely to be rated by teachers as non-rigid, adaptable, The New Outlook for the Blind, 1970, acceptable to change, likely to try 64, 287-292. new ways of doing things and also to do good school work. A child scoring The purpose of this study was to high on all three variables in crea- investigate the relationship between tive thinking and with a high I.Q. is sound discrimination ability and likely to be rated high on curiosity, blindness, degree of visual acuity, desire to know more about self and tactile-kinesthetic ability, chrono- environment, and to seek new experi- logical age, and mental age. Eighty- ences . five visually limited and 77 sighted children were given Form A of the Irwin Sound Discrimination Test, and a tactile-kinesthetic discrimination test (Crandell, Haminill, Witowski and Barkovitch, 1968). Mental age was estimated by using either the Hayes-Binet Intelligence Scale, the Verbal Scale of the WISC, or the Slossen Intelligence Test. Analysis of the data yielded the following results: (1) There was no significant difference in sound discrimination

249 . ability between visually handicapped Harley, R. S., Pollen, J., & Long, S. and sighted Ss with similar mental and chronological ages. (2) No sig- A study of the reliability and valid- nificant relationship was found be- ity of the visual efficiency scale tween sound discrimination ability with pre-school children. and either degree of sight or tactile- kinesthetic skills. (3) A signifi- Education of the Visually Handicapped, cant relationship for both groups was 1973, 5, 38-42. found between sound discrimination ability and mental age (p less than Subjects: 104 nursery and kinder- 0.01), and for the visually handi- garten children, the former being in capped Ss between sound discrimina- the 4 year age group, the latter 5- tion ability and chronological age 6-1/2 years of age. Purpose: to test (p less than 0.01). the Barraga (19 70) Visual Efficiency Scale. Method: the Barraga Scale was administered individually to all the children, some of the younger ones Harley, R. K., & Rawls, R. F. having two short sittings. Results: the scale was found to have both con- Comparison of several approaches for tent validity and acceptable internal teaching braille reading to blind consistency. Item analysis indicated children that many of the items did not discriminate sufficiently between high American Foundation for the Blind, and low scorers (the top 27 percent

Kesearah Bulletin No. 23, 1971, 63-85. and the bottom 27 percent) . Results also indicated that some items are In order to develop and field- inappropriately placed according to test materials to be used in a 2-year efficiency rating and some areas of study to compare 6 approaches in teach- the profile are not represented by ing braille reading to blind children, discriminating items, e.g. object materials in 2 braille media--grade 1 contour, light/dark intensity, image and phonemic—were devised and tested, constancy of outlines, pattern de- along with the traditional grade 2 tails and objects. system, in 2 basal readers, using contrasting approaches , analytic and synthetic. The 39 Ss , aged 5-10 years, were located in 6 classes in Havill, S. J. 6 residential schools for the blind. Teachers received special preparation The sociometric status of visually prior to the start of the school pro- handicapped students in public school gram. Ss were introduced to the ma- classes. terials following a readiness program, and a daily progress report was kept American Foundation for the Blind, by each teacher; teacher reactions Research Bulletin No. 20, 1970, 57-91. were later used in the evaluation of each approach. Results of the Slosson The purpose of this study was to Oral Reading Test and the Gilmore Oral examine the peer-acceptance level of Reading Test administered at the end visually handicapped children inte- of the academic year indicated that grated in regular classes. Sixty- phonemic braille could be used suc- three visually handicapped children cessfully with these beginning braille from grades 4 through 12 with no readers. The analytic approach ap- other handicap formed the experi- peared to function more effectively mental group. Each visually handi- for the phonemic materials than the capped subject was matched with a synthetic approach. The effective- sighted classmate on the basis of ness of the grade 1 approaches was sex, race, age, socioeconomic level, not adequately measured. A study of and achievement level to form a con- longer duration with more Ss and trast group. All children in a class more adequate materials is necessary where there was a visually handicapped to make generalizations concerning child were asked to choose 5 desired the efficacy of approaches in braille companions from their classmates for reading. each of 3 situations: working, leisure, and seating. A chi-square statistical analysis of the data revealed a significant difference in sociometric

250 .

status beyond the 0.001 level of con- permit the adaptation of visual effi- fidence between the visually handi- ciency patterns for specific visual capped experimental group and the activities sighted control group. There was also a significant difference (at least at the 0.01 level) between each of 3 visual impairment categories and Hopkins, K. D., & McGuire, L. the control group, and a significant difference in sociometric status be- Mental measurement of the blind: The tween visually handicapped children validity of the Wechsler intelligence of different levels of achievement. scale for children. Significant differences in sociometric status were also found between International Journal of Education visually handicapped children being for the Blind, 1966, 15(3). served by itinerant and resource room programs. None of the other variables The purpose of the present study examined (degree of impairment, grade was to obtain evidence on the valid- level, sex, socioeconomic level, time ity of comparability of the WISC as spent in regular classrooms , and a measure of intelligence for blind choice situations) revealed signifi- children. The study revealed that cant results. The author concludes despite the fact that the Hayes-Binet that the condition of being visually and WISC measure essentially the same handicapped, and thus different, re- abilities, and both yield normal dis- sults in low status and difficulties tributions of IQ scores, the corres- in social acceptance. ponding IQ scores are not comparable, i.e. there is a highly significant trend for WISC scores to be lower than H-B scores, the mean difference being Hoover, R. E. 8.5 points. (The differences expected at the extremes would appear to be even Visual efficiency as a criterion of larger since the standard deviations service needs. for the 2 scales are substantially different.) This study re-emphasizes American Foundation for the Blind, that a high correlation and strong re- Research Bulletin No. 3, 1963, 116- lationships between tests in no way 119. insure comparable normative data. An analysis of variance of the subtests An ophthalmologist gives his scale scores revealed a significant definition of visual efficiency as a tendency for relatively inferior per- complex of measurable visual charac- formance by the blind subjects on the teristics which, when combined v;ith comprehension subtests. This factor other sensory and physical character- attenuates the implicit validity as- istics, provides an opportunity to sumption of equal opportunity for utilize sight (opportunity implies learning and performing and suggests motivation or incentive) . Visual acu- some general visual bias on this type ity plays only a part in the deter- of scale. Results also show a tend- mination of visual efficiency. ency for lower intercorrelations between subtests for the experimental Visual acuity may and should be sample, implying greater intra- measured at several distances of both individual differences in intelli- far and near points. Visual versa- gences among the blind. No differ- tility measurements should include ences were found in subtest relia- optical power, accommodation, and con- bility between present sample and vergence and divergence ability — even standardization data. It is recom- in pathological conditions . Visual mended that users of intelligence capacity includes the measurement of test data be aware of the marked the length of time, the amplitude and lack of interchangeability between conditions for maintenance of visual WISC and Hayes-Binet scores and gear functioning. corresponding interpretations accordingly. The measurement of other physical and sensory characteristics and the establishment of patterns of visual characteristics required for the performance of various tasks would

251 . ,

Hopkins, K. D. , & McGuire, L. interpreted as genetically qualified but attributed to the effect of vi- IQ constancy and the blind child. sualization.

International Journal of Education for the Blind, 1967, 16(4).

Huff, R. , & Franks, F. The four-year constancy of IQ was investigated for a group of 30 Educational materials development in blind children. The initial Hayes- primary mathematics : Fractional parts Binet (H-B) IQ scores were much more of wholes comparable to subsequent WISC IQ scores than to the retest scores on Education of the Visually Handicapped the H-B, especially when actual IQ 1973, 5, 46-54. scores were considered. It is recommended that H-B scores be interpreted Subjects: 9 beginners, 10 1st- very cautiously since marked varia- graders , 10 2nd-graders . Each sub- tion was noted between test and reject was tested individually in a test performance. For a high degree series of manipulation (correctly of IQ constancy on a test, 2 condi- placing rubber circles and fractional tions must be evidenced: first, a parts of circles in circular nests in large amount of lasting variance, plastic form-boards) and discrimina- which indicates stable individual tion (of the "same or different," differences are being assessed, must "which is the largest" type) tasks. be present; this condition was met by Measures were taken of (a) the mean the tests used in this study as dem- number of manipulation trials re- onstrated by the high stability co- quired to correctly complete the efficients; second, adequate norma- tasks (for each grade level) ; (b) per- tive data, minimizing fluctuations cent of correct responses (mean for in the mean and standard deviation each grade) for each type of discrimi- of the IQs from year to year, must nation task. Results (a) decreased be present. This latter would appear over grade levels, all differences to have been the major factor which being significant; (b) for same/ accounted for the wide IQ discrepan- different tasks percent increased cies on the H-B. The study has im- with grade but only the difference plications for the renorming of the between beginners and 2nd-grade was H-B. significant. For "which is largest" tasks the percent of correct responses only increased between beginners and Ist-grade and this dif- Hudelmayer, D. ference was not significant. There were also several inconsistencies Concept-attainment with blind chil- between grade levels. All subjects dren. as-a-group failed to distinguish between 1/3 and 1/4 when placed in a

The Visually Handiaapped , 1972, nest in a form board. Perhaps super- 1, 49. imposition would be a better method. Further research is required to see The research should test in the if preliminary training would help area of nonverbal concept attainment, discrimination between 1/3 and 1/4. whether early visual deprivation has a general effect on the capacity of 9-15 year old blind children for information processing capacity. Forty Israel, L. congenitally blind or early blinded scholars were individually matched C.C.T.V. reading machines for visually with two fully sighted according to handicapped persons: A guide to selec- sex, age, and verbal IQ. With audition. tory tasks no over-fortuitous difference between blind and sighted chil- The New Outlook for the Blind, 1973, dren was encountered. On the other 67, 102-110. hand the sighted mastered the tactile tasks very significantly better than Closed-circuit television read- the blind. This difference is not ing machines are now available from

252 . .

several sources, with different mod- basic pre-Ccine skills and the addi- els offering a variety of features. tion of more intermediate steps in Among those criteria to be considered the standard teaching techniques. in selecting a device for purchase, the following are suggested as being vitally important: magnification, fo-

cus reversed image, sharpness (bright- Jones , J w. & Collins, A. P, ness) , clarity (contrast) , movable viewing table, adjustable monitor Trends in program and pupil placement (distance, weight, tilt) portability, practices in the special education of accessories, monitor selections, aes- visually handicapped children. thetics, warranty and service, long- term support, and user's manual and International Journal for the Educa- instructions tion of the Blind, 1965, 14, 97-100.

Organizational patterns in school programs for the visually handicapped Johnson, S. show an interesting change from a high percentage of special classes in 1946 Sources of agency referrals. to a very small percentage of children segregated in special classes in 1963. The New Outlook for the Blind, 1971, Resource rooms and itinerant teaching 65, 49-50. plans are used in a majority of pro-

grams . In this article, the author discusses how to encourage more people Teachers in both residential and to seek assistance, other than finan- local school programs work with both cial, from agencies serving the blind. blind and partially seeing children It is felt that more public education in combination units, and increasing is necessary, especially within the numbers of programs include visually medical field; but the author points handicapped children with additional out that public education will never problems . The ability to use print be the complete answer. Rather, she materials rather than the degree of states that people remember what they visual acuity is the basis for classi- see, and thus the utilization of com- fication in most schools. munity resources for clients, plus the integration of clients into community activities, would be more successful in making others aware of the Jurmaa, J. services available through agencies for the blind. The benefits of a The spatial sense of the blind. community-oriented program such as this would also be felt by the cli- American Foundation for the Blind,

ents themselves , in giving a sense of Researah Bulletin No. 24, 1972, participation in and belonging to a 57-64. community group. Presents a research protocol with 3 major problem areas proposed for study: (a) intercorrelations among the Johnston, B., & Corbett, M. various sense modalities of abilities for mental manipulation in space; Orientation and mobility instruction (b) the transposition of sense images for blind individuals functioning on from one modality to another from the a retarded level. viewpoints of--extent of attempts, likelihood of success, and the de- The New Outlook for the Blind, 1973, gree of help such a transposition is 67, 27-31. in performing a task; (c) intercorrelations among spatial ability, trans- Experience at the Orange Grove position and some control variables. Center for the Retarded, Chattanooga, The experimental approach is briefly indicates that blind youths who are outlined. functioning on a retarded level can be taught orientation and mobility skills through an emphasis on very

253 . . , .

Karnes, M. B., & Wollersheim, J. P. device (a cone of about 60 degrees) The pitch of the sound heard by the An intensive differential diagnosis user is proportional to the distance of partially seeing children to de- of the object and the quality of the termine the implications of educa- sound gives some indication of the tion. nature of the object. The report is based on data from 31 instructors Exceptional Children , 1963, 30, 17-25, 104 totally blind subjects (age range 11-55, average or above intelligence). By use of a variety of psycho- Ten subjects were novice travelers, logical and educational instruments, 74 trained in long cane, 20 trained highly relevant clinical information with guide dogs. Training in the use was obtained relating to the vari- of the glasses seemed to be a major able factors which influence success- variable —where training was good, so ful learning of the children with were performance and motivation. The partial vision. Implications of the sonic glasses greatly improve aware- study suggest the feasibility of a ness of the environment and reduce clinical type of teaching utilizing the stress of travel, since with special methods and techniques to training the user learns to distin- facilitate learning and achievement guish between the objects in his path. more in keeping with the potential This includes people and thus the abilities of the children acquired blind person can avoid embarrassing by a thorough interpretation of the collisions, can judge the space be- individual strengths and weaknesses tween himself and fellow travelers of the children. and can sense spaces to move into. ' Knowing that he has encountered a human object he can ask the way if need be. The major problem in the

- Kay , L : case of the congenitally blind is one of perception--he cannot mentally The sonic glasses evaluated. visualize the world in 3D, or readily relate to the idea of space. Even The New Outlook for the Blind, 1973, the simple concept of direction may 67, 7-11. be confusing. However, the evaluation brings out 2 important facts. A precis of the findings of two That almost all the teachers trained questionnaires (169 items for the 9 4 in teaching the use of sonic glasses users of the aid, 8 8 for the 21 train- wish to continue to do so. Success ers) to evaluate the ultrasonic bin- rate among subjects was over 90 per- aural sensory aid for the blind, a cent—only 8 subjects dropped out mobility and orientation device in- (maybe because of inadequate teaching vented by the author. The results, in this initial stage of introduction summarized in three lengthy tables, of the device) indicate that 8 8 percent of the users found the device useful and that only 20 percent of the trainers thought

that their own training was inade- Kederis , C. J., Nolan, C. Y., & quate . Morris, J.

The use of controlled exposure devices to increase braille reading rates. Kay, L. International Journal for the Educa- Sonic glasses for the blind—A prog- tion of the Blind, 1967, 16, 97-105, ress report. This is a report of 2 studies American Foundation for the Blind, using controlled exposure techniques Research Bulletin No. 25, 1973. to increase braille reading rates. In one study, familiar words, phrases, Sonic glasses are similar in ap- and short sentences were read at di- pearance to optical glasses but work minishing time intervals on a rapid- by the emission of VHF sounds which exposure device. In the other, are "echoed" by reflection from the literary material was read at in- objects in the arc covered by the creasing word rates on a pacing device,

254 .

Both studies used an experimental and (a) normal seeing adults, even with control group of 15 matched subjects eyes closed, and those blinded after divided into reading rate levels birth saw a compound figure as con-

(fast-slow) . The experimental groups sisting of 2 or more simple, rela- practiced reading one half hour/day tively large figures — called a type for 20 consecutive school days. A segregation--while congenitally blind monetary regard was used to control subjects perceived it as consisting for motivation on the criterion tests. of many fragmented parts--called non No significant effects of reading o type segregation. (b) Young chil- practice were found on pre- and post- dren with normal sight perceived fig- reading tests. Motivational effects ures with various types of non a apparently accounted for the 2 4 per- type segregation, indicating that cent average reduction of reading normal sight and visual experiences times in both studies. were not sufficient for attaining a type segregation. (c) The non a type segregation of the congenitally blind was weakened when size stimulus con-

Kenmore , J text, plus figure completion were varied. Helping blind persons to find and prepare for employment.

The New Outlook for the Blind, Klimasinski , K. 1973, 67, 111-115. An attempt to test the personality of Four kinds of job market surveys the blind using the MMPI. are recommended for discovering possible occupations for blind persons. American Foundation for the Blind, The role of schools in promoting the Research Bulletin No. 24, 1972, success of blind youngsters when they 65-74. enter the job market is fully discussed. Where schools fail, many The Polish version of the MMPI countries start intermediary centers (550 items) was modified (342 items) to prepare blind young people for the and administered to a group of 30 world of work. The role of the place- blind men tested in groups of 4 or ment officer is emphasized as are the less, with the order of item presenta- roles of agencies and government. Ex- tion changed for each group. Half of amples in each area are drawn from the subjects had visual losses greater many countries around the world. than 20/200 but did not differ on any Based on a speech given to a biennial MMPI scale from subjects with visual conference of agencies and schools losses less than 20/200. All subjects for the blind in France. were compared with MMPI norms based on 100 men with normal sight; statistically significant differences were obtained for 2 control keys plus 4 clin- Kimura, M. ical scales. The difficulty in spec- ifying the extent to which these dif- Perception of the compound figures by ferences reflect loss of sight or tactile motor-scanning. subject's social situation is noted.

(P. A. Vol. 49, No. 4, April 1973 from) Japanese Journal of Psyahology, April 1972, Vol. V3(l) , 1-12. Knight, J. J.

Investigated the mechanism by Mannerisms in the congenitally blind which the congenitally blind tactu- child. ally perceive compound figures in which 2 or more simple, geometrically The New Outlook for the Blind, 1972, contoured figures are overlapped. A 66, 297-302. series of 11 experiments were performed with congenitally blind adults, Describes mannerisms in the con- normal seeing adults with eyes closed, genitally blind child as non-instru- persons blinded after birth and normal mental behaviors used in infancy as seeing children. Results show: a means to reduce tension and

255 . — . , maintained afterwards when the child i.e. that teachers who have never has not been able to substitute non- taught blind children do not neces- instrumental gross motor activities sarily have more negative attitudes by acquisition of learned instru- toward blindness than teachers who mental behavior. The sighted infant, have had experience with blind chil- during his normal development, learns dren. to abandon non-instrumental gross motor activities and begins to acquire learned instrumental behavior. During the first two years of life, the Kumpe, R. blind child should receive special help; otherwise, some basic instru- Preparation of blind prospective col- mental coping behaviors, especially lege students. reaching and crawling skills, may not be developed to their fullest poten- International Journal for the Educa- tial. tion of the Blind, October 1968, XVIII.

One hundred two blind students, preparing to enter their first year

Kuhn , J at college or university, were divided into 2 groups. One group was given

A comparison of teachers ' attitudes 10 weeks of special preparation for toward blindness and exposure to college; the control group received blind children. no special training. The training for the first group comprised prac- The New Outlook for the Blind, 1971, tice with several skills, both aca- 65, 337-340. demic and personal/social, which would lead to greater success in the In order to determine whether university setting (e.g. writing term regular classroom teachers with ex- and research papers; taking notes; perience teaching blind children orientation and travel; discussions would express more positive attitudes of personal and social problems; toward blindness than would teachers special resources for blind students without this exposure to blind chil- etc.). The evaluation of the program dren, the "Attitudes to Blindness by the experimental group itself in- Scale" was given to 46 elementary dicated that 39 percent felt it Very school teachers. Twenty-eight of Helpful; 44 percent felt it of Some the teachers taught in a school with Help; and 17 percent felt it was No a resource room for blind children Help. The control group, asked to and had had a blind child in their • consider the value such a program regular classroom for at least 1 hr. might have been to them, felt that per day; the other 18 teachers it would have been Very Helpful taught in a school located in the (43 percent); of Some Help (29 per-

same district but which had no blind cent) ; or of No Help (28 percent) children enrolled. A comparison of Comparison of academic performance the mean scores showed no significant of the 2 samples indicated that the differences between the 2 groups, al- special training had significant ef- though the mean scores of both groups fect for the experimental sample, as (43) were approximately 10 points 17 percent of them, were in the first lower (thus indicating a more posi- quartile academically as against tive attitude toward blindness) than only 7 percent of the control group. the mean scores of college students There was no significant difference

tested by Cowen , Underberg , and Ver- between the groups in continuance rillo (1958). A chi square analysis in college to the end of the first made on each question revealed no year, but both groups showed an ad- significant results. The author con- vantage in this area over the gen- cludes that exposure to blind children eral first year population. No does not necessarily mean that the significant difference in grade point teacher thus exposed will have a more average was found between the groups. positive attitude toward blindness than those teachers without the experience of teaching blind children. However, it seems apparent that the reverse conclusion is also indicated

256 .

Kurzhals, I. W. the child's ability. These mothers became aware of their attitudes dur- What is "readiness" for the blind ing interview and responded readily child? to guidance leading to spectacular improvement in the child. (3) Autono- International Journal for the Educa- my high, posture normal, sociability tion of the Blind, October 1968, XVIII. and language low, sensory-motor very low. Associated with mothers who had This article discusses "readi- a long depressive period on discovery ness" in terms of four general areas: of blindness and never really got physical readiness, mental readiness, over it, though on the surface they social-emotional readiness, and psy- appear to. They over-react and, de- chological readiness . It is suggested termined that the child should be that the blind child should manifest "normal," encourage activities like "readiness" in each of these areas riding, walking. But they never play before entering an educational set- with the child or talk with him— these ting. Once having entered a school, mothers do not respond to guidance however, the instructional program since it arouses guilt feelings. must be geared to maintain the prog- (4) Very low overall score--genuinely ress already made by the child and multi-handicapped children. continue his growth and development. The program for blind children in the state of Utah (U.S.A.) is described; it is a developmental program for Lairy, G. C. & Harrison-Novello, A. primary-aged children and is divided into learning areas: orientation and Family attitudes and congenital visual mobility; music and rhythm; explora- deficit. tory time; language development; numerical development; and reading. The Psychological Abstracts , August 1973, aim of such a program is to help the 50 (2) , No. 3256. child to develop as a total person and to prepare him for later educa- Suggests that appropriate parental tional experiences. reactions to handicaps of a child have important consequences for the psychological development of the child. De- velopmental problems are reviewed.

Lairy, G. C. , & Harrison-Novello , A. Characteristics of blind or nearly blind children examined by the authors

The blind child and its parents : Con- are described. Results on a modified genital visual defect and the reper- version of the Maxfield-Bucholz social cussion of family attitudes on the development scale, observations of early development of the blind child. parent child interactions, and discussions with parents suggest a classi- American Foundation for the Blind, fication of these children into four Research Bulletin No. 25, 1973, 1-25. groups with different developmental profiles A study was made of legally blind children under 6 years old in Paris. (French: "Interpretation," Vol. 5(2-3), The Maxf ield-Bucholz Scale of Social April 1971, pp. 157-186.) Maturity was administered and the child observed in the presence of its parents, during extensive interviews. Performance was found to be related Le Zak, R. J., & Starbuck, H. B. to family background and four patterns seemed to emerge. (1) Overall score Identification of children with speech normal, heterogeneous scatter between disorders in a residential school for the five behavior categories : associ- the blind. ated with accepting parents and, usually, younger siblings. Children International Journal for the Educa- encouraged to use residual vision to tion of the Blind, 1964, 14, 8-12. best advantage. (2) Speech very good; posture normal; autonomy, sensory- The 173 children studied ranged motor, sociability about half the in age from five to 21 years. The normal score. Associated with over- oral stimulation technique was uti- protective parents underestimating lized and descriptive ratings were

257 , for normal and deviant responses of each subject. The same order of (a) articulation, (b) voice charac- presentation was used to make a teristics, and (c) rhythm. Speech weighted trial, in which the pencil disorders were evident in 50 percent used by the subject to record his of the children with articulation response was weighted with lead-shot. disorders in the highest percentage, Equal numbers of subjects began with then voice, nonfluency, and multiple weighted/unweighted, equal numbers disorders in decreasing order. Find- started with preferred/non-preferred ings suggested the need for formal hand. Arms were alternated every 10 speech therapy programs to develop trials. Two scores were recorded: earlier speech proficiency. These (1) position score — the magnitude of programs would be supplemented by error in locating the target posi- classroom speech improvement programs tion; (2) movement score — the distance to facilitate maximum communicative moved relative to the length of the potential of blind students. trajectory. It was found that early blind subjects were significantly more accurate than sighted subjects in reproducing movement length, with

Levitt, A. L. , Willerman, L. I., no difference between the groups on Rosenbaum, A. L. , & Levitt, M. position scores. Weighting the pen- cils made no difference to the re- Intelligence of retinoblastoma pa- sults. There was a significant loss tients and their siblings. of accuracy over the 40 trials--possibly because there was no feedback Child Development , 1972, 43, 939-948. on the accuracy of the responses made. The only view which can explain how Compared 25 sighted and 19 blind early blind are better on arm move- retinoblastoma patients with their ment, but doesn't lead to expectation normal siblings (n = 59) on the WAIS of superior performance in other Wise, Stanford Binet Intelligence parameters of kinaesthetic percep- Scale or Williams Intelligence Scale tion, seems to be one involving tim- for Children with Defective Vision. ing, blind people characteristically Subjects blinded from the disease reporting that they judge distance by average 10 IQ points above their sib- timing. At any rate the evidence sug- lings while the sighted subjects with gests that an adequate, or even su- both eyes affected by the disease were perior, model of personal space can inferior to their controls. Unilater- be developed without vision. ally affected subjects did not differ from their controls . It is concluded that while retinoblastoma per se is not associated with intellectual su- Linn, M. periority or inferiority, retinoblastoma associated with blindness An experiential science curriculum may result in selective cognitive for the visually handicapped. superiority.

Exceptional Children , 1972, 39, 37-43.

Describes the adaptation and Lindlay, S. evaluation of a materials-centered experiential curriculum. The adapta- Kinaesthetic perception in early- tions of both the physical and the blind adults. life science units can be used in classes with one or two visually im- American Foundation for the Blind, paired students and in classes of all Research Bulletin No. 25, 1973, visually impaired students. The pro- 175-193. grams involve firsthand experience with objects and live organisms. Twenty early-blind subjects were Evaluation measures were designed to matched by age and sex with 20 sighted assess the major objectives of each subjects, aged 16-45. Subjects were unit in the program. Classroom trials asked to repeat arm-movement trajecto- (4 classes N = 48) of two of the adapt- ries, presented to them as drawings. ed units revealed that visually im- Each subject had 40 trajectories, the paired students made significant gains order of presentation being random for in understanding both content and process objectives of the units.

258 ,

Linvill, J. G., & Bliss, J. C. because retational difficulties are present in these doubly handicapped A direct-translation reading aid children. A concrete adaptation of for the blind. L. Gorman's "Patte Noir" test is demonstrated, suggesting new per- Stanford Electronics Laboratories, spectives in the exploration of Stanford Univ., Stanford, Calif. blind children's personalities. (SU-SEL-65-055, Technical Report No. 4819-1).

A reading device for the blind Mayadas , N. S. is proposed in which a facsimile of ordinary printed material is present- Role expectations and performance of ed tactually. The tactile image is blind children: Practice and implica-

presented by a dense array of pins tions . which can be made individually to vibrate through perforations in a Education of the Visually Handicapped plate on which the finger of the 1972, IV, 45-52. reader is rested. The role performance of 56 blind In the arrangement proposed, the adolescent subjects were compared image of ordinary printed matter is with the role expectations of (1) sig- focused on an array of photocells nificant others, (2) the blind child's which are coupled one-to-one to perception of significant-other ex- piezoelectric reeds which drive the pectations, (3) blind child's self- image-producing pins. Preliminary expectations, and (4) expectations of designs of this arrangement have been persons who were "strangers" to blind- tested, and engineering considerations ness. All blind subjects and their of the design are presented. significant others, plus a random

sample of strangers , were administered Successful reading tests with questionnaires to score the variables blind subjects are reported in which in expectation categories. In addi- a computer controller simulates the tion, performance behaviors were optical portion of the system. The scored from a daily count of actual tactile images presented on a field behaviors occurring over a 2-week of 96 piezoelectrically driven pins period. Results indicated that role have been readable by the three sub- synchrony existed between performance jects tested. of blind subjects and (1) the expectations of significant others, (2) the subject's perception of significant other expectations, and

Lissonde , B., & Porot , M. (3) the subject's self-expectations; role asynchrony was indicated between Visual deficiency and mental defi- expectations of persons who were ciency. strangers to the blind and the performance of blind subjects. Thus

Review de Neuropsy ahiatrie Infantile role expectations , except for random et d'Hygiene Mentale de I'Enfance^ stranger expectations, were found to 1971, 19, 1-19. be significant partial predictors of

performance . Analyzes elements common to children who suffer both visual and mental defects. Those with relatively high intelligence generally have a visual acuity of less than 1-10. These subjects display obvious difficulty in motor tests, and are usually of calm, passive and introverted character. It is noted that differences exist between those who are blind from birth and those who later become blind, the former being better able to cope with their handicap. Parents' attitudes play a primary role in this context.

259 . . ,

Mersi, F. blind and partially sighted subjects were compared. However, when levels Braille for partially seeing. of anxiety for those students in grades 9 and 10, and 11 and 12 were The Visually Handicapped , 1972, compared, it was found that the anxi- 1, 32-33. ety level of 11th and 12th grade subjects was significantly greater

Braille-shorthand is an important (P = 0.05) . An examination of spe- aid for partially seeing pupils , stu- cific responses was also made, and it dents and office workers. In indivi- was revealed that the majority of dual cases braille longhand writing anxiety responses centered around can be a means for mentally retarded the areas of social competences, partially seeing to develop a self personal appearance, and adjustment decipherable writing. The condition to blindness. in each case is however, that braille writing is adapted to the special requirements of the group in which it is applied. The best solution would Miller, W. H. , & Porter, J. E. be embossed braille print with contrast in color, which for example Read it. Say it fast. — The use of could be accomplished by means of Distar Instructional Systems with microencapsuled film. It would meet visually impaired children. especially the need of the "mix tech- nicians" and the correctors. In the Education of the Visually Handicapped majority of cases a braille writing 1973, V(l) , 1-8. machine , that stamps only visually well perceptible holes, would suffice. Distar is a teaching program Visually adapted braille is a help and with three sequenced levels in read- advantage in the case of the preven- ing and arithmetic, and two levels in tive learning of the braille system. language , with aids , books and games As the braille writing machine is for each program. The major premise easily handled, even some manually of Engelmann et al. — the originators handicapped persons could thus be of Distar—was that any child can helped eventually. The issue of all learn if he participates in logically publications in punched braille with sequenced learning experiences, Distar color contrast would be advantageous was tested on visually handicapped for many readers. Being dependent on children in a non-graded primary braille is not in all cases identical school after a sequence of seminars with the suitability of applying teach- and workshops with the teachers. ing methods for the blind. Adaptations made: (a) for Partially Sighted-- (1) Original print stimuli were enlarged to 4x and gradually reduced to 2x over 40 lessons. Simi- Miller, W. H. larly with spacing. (2) Visual em- bellishments to attract pupils atten- Manifest anxiety in visually-impaired tion in the standard text were ex- adolescents cluded. "Check-test stimuli" were simplified. (3) Hand signals (cues Education of the Visually Handicapped, for response) were exaggerated. 1970, 11(3) , 91-95. (b) for Totally Blind— (1) Standard texts using varying sizes and contexts Hardy's Anxiety Scale for the randomly distributed over the page Blind was administered to 50 pupils were changed to standard braille and at a residential school for the blind the page divided into four quadrants who were placed in grades 9-12 and in (2) Standard texts use small letters special classes. The purpose was to for "silent" letters. For the blind examine relative levels of manifest the words were spelled as they sound anxiety between students in grades 9 and were accompanied in class by and 10, 11 and 12, and special classes, verbal explanation. (3) Auditory and between totally blind and partial- rather than visual cues, were used, ly sighted students. No significant e.g. clicking fingers instead of differences were found when special raising hand as cue to start. The classes were compared with grades 9 program was concluded to be success- and 10, 11 and 12, nor when totally ful and extended now to multiple handicapped as a result.

260 , ,

Milstead, J. R. Mishchenko, M. N.

Modification of echolalic speech in Peculiarities of voluntary movements a blind, behaviorally deficient of the facial muscles in blind peo- child using parental contingency ple. management.

Sovetskaia Nevropatologia , 1935, Journal of Communications Disorders 7, 121-132 (in Russian);'

1972, 5(3) , 275-279. Psyohological Abstracts , 1936, 10, Item 246. Details the case history and treatment of an 8 year old, blind, "The peculiarities of voluntary behaviorally deficient boy whose movements of the facial muscles of echolalic speech was modified through blind people, investigated by Quint's a contingency management program us- method, are as follows: facial motor ing the subject's parents as the activity in the blind is reduced be- primary therapists. The frequency low that of seeing people; asynchronic of nonecholalic responses increased movements are more difficult for the as a result of treatment, and the blind than are synchronic ones." change appeared stable.

Moore, M. Mishchenko, M. N. Development of number concepts in Quint's method of the investigation blind children. of the facial motor phenomena of blind children and youth. Education of the Visually Handicapped 1973, 5(3) , 65-71.

Nevropatologia i Psikhiatria , 1936, 5, 1379-1386 (in Russian); Blind children have often not

Psychological Abstracts , 1937, 11, developed the logical cognitive proc- Item 1254. esses on which mathem.atics is based, before they enter school. Basic to "The action of the loss of the an understanding of number is the visual receptor on facial motor phe- understanding of the conservation of nomena was investigated, using Quint's number, in which classification and method, consisting of 26 motor tests. seriation skills are involved. Blind Comparing blind children with normal children tend to be "experientially ones , a decrease of facial activity deprived" —missing out on the "help- was observed. The visual receptor is ing" tasks at home which develop ba- the main factor in the development sic concepts. Many experiences, and formation of the facial motor formal and informal, must be pro- experience, the other factors being: vided for the blind child in school age, time of the loss of vision, and to develop skills of matching, class- constitution. The loss of the visual ification and seriation before he is receptor excludes the possibility of ready for the development of mathe- forming facial movements by imita- matical concepts. A hierarchy of tion. As a substitute for vision skills has been developed such that there are used the proprioceptive mastery of a particular task in the motor formations. The correlation hierarchy presupposes mastery of between facial motor activity and previous concepts and skills. Fail- constitution is the same as between ure to learn may not be attributable general motor activity and constitu- to poor teaching or inability to tion. " learn but to choosing a task which is at a higher level of conceptual development than the child has reached. A summary is presented of a hierarchically organized sequence of tasks adapted for the blind leading to an imderstanding of numbers and the number system.

261 . . ,

Morris, J. E., & Nolan, C. Y. Naddeo, C. L. , & Curtis, S. W.

Minimum sizes for areal type tactual Some effects of the disorientation symbols of tactual-kinesthesia.

International Journal for the Educa- International Journal for the Educa- tion of the Blind, 1963, 13, 48-51. tion of the Blind, 1968, XVIII.

Purposes of this study were to The missing parts of 14 objects establish minimum outer dimensions of the Haptic Intelligence Scale were for areal symbols and to determine identified by 2 groups of sighted whether grade differences exist in blindfolded Ss . The Ss in one group the minimum sizes at which students received the objects with their hands can recognize them. positioned normally in front of them. The Ss in the other group received Minimum outer dimensions were the objects with their hands posi- established and a notable grade level tioned in front of them, crossed at inequity in identification was evi- the wrists and inverted with thumbs dent as symbols diminished in size. -• down, to provide a condition of dis- Children in grades four through seven torted tactile-kinesthetic perception. require large symbols for easy re- The Ss in the control group required cognition. Graphics for the blind less time and made fewer errors in should be constructed according to identification than the Ss in the ex- grade level use. perimental group. The differences between the groups were statistically significant. Such results have implications for both practical testing

Myers , W. A. " procedures and philosophic considerations regarding the appraisal of in- Color discriminability for partially telligence, where sensory and percep- seeing children. tive channels are distorted in quality rather than reduced in quantity (from Exceptional Children, 1971, 38, threshold) 223-228.

Thirty 8-12 year old myopic children without color vision problems Napier, G. (acuity 20/70-20/200 in the best corrected eye) viewed Snellen "E"s in A writing vocabulary study relative five different colors against dif- to braille contractions to be mas- fering backgrounds of the same col- tered by primary level children. ors. A black on white combination served as a control condition. "E" Education of the Visually Handicapped measured the distance at which sub- 1973, V(3) , 74-78. ject could identify the direction of the "E" as well as the distance at Outlines the difficulties in which the subject reported greater learning braille —especially in writ- clarity; the results of both measure- ing, with so many rules and contrac- ments were comparable. Low contrast tions to be kept in mind at each step. combinations were poorer than the This study hopes to find (1) those

control (p < 0.05) . A few combina- words that have the highest functional tions (e.g. yellow on black and blue) value in meeting the child's needs; were non-signif icantly better than (2) whole word contractions and their the control. The relative ranking sequence of introduction among the of all combinations is reported. words of (1) above; (3) those part- The study is significant in that it word contractions with the highest used controlled colors (Miinsell) and frequency (to determine the most fxonc- statistical techniques. tional) . Method: Used 5 published vocabularies to determine which words occurred most frequently and which contractions among those words occur most frequently. Vocabularies used were relevant to primary graders rather than adults, with an emphasis on

262 . .

writing rather than reading. All control. Learning was tested by a words containing braille contractions 36-item, four-alternative multiple were noted. Those appearing in all choice test. Numerous methodological 5 were considered the most crucial factors were identified as well as ones to learn first — since they are significant variables which appear deemed important by 5 authorities to influence the relative effective-

(Category (1) words) . Next important ness of the two modes of learning. are those occurring in 4 vocabularies The possibility that an' overemphasis (Category (2)) and so on. 7560 words on braille reading presents an obsta- were studied. Of these 4472 contained cle in the education of blind children contractions. When duplications are is a relevant implication. removed 215 8 words remain--these constitute the basis of the study and have been designated the Composite Vocabulary. Of 190 braille contrac- Nolan, C. Y. tions only 169 occur in the Composite Vocabulary; 129 are whole word signs, Blind children: Degree of vision, 40 are part-word signs. The first 11 mode of reading--a 1963 replication. part-word signs constitute 74 percent of all part-word signs. A teacher The New Outlook for the Blind, 1965, who uses these 11 part-word signs and 59, 233-238. the 6 3 whole-word signs from Category (1) words has the core of words with Summary of differences: Compari- which to launch initial instruction son of 1960-1963 data. in writing braille. 1. The total nximber of legally blind children registered with the American Printing House increased National Committee for Research by 2,536 children during the in Ophthalmology and Blindness. period. Of these, 2,072 were enrolled in local school programs Proceedings of the symposium on and 46 4 were enrolled in residen- research in blindness and severe tial school programs visual impairment 2. Between 1960 and 1963 the per- New York: American Foundation for the centage of students listed as Blind, 1964. 113 pp. braille readers decreased 5 percent and the percentage of stu- Attention in eye research has dents listed as print readers shifted increasingly from visual acu- increased 5 percent. This ap- ity to visual efficiency. The im- pears due primarily to a dramatic plications of the relevant factors increase in the proportions of involved for ophthalmologists and the residential school students in coordination of their work with other visual categories I-IV who are disciplines were the primary topics listed as print readers. of the symposium. Definitions, regis-

tration reporting techniques , coding 3. In 1963, 12 percent more residen-

procedures , and technological develop- tial school students possessing ment were other themes for discussion. object perception or better were registered as print readers.

4. The number of legally blind stu- Nolan, C. Y. dents classed as ungraded in 1963 was 1,348. This was a Reading and listening in learning by 50 percent increase over the the blind. number of students so reported in 1960.

Exceptional Children , 1963, 29, 313- 316. 5. The accumulation of braille readers in kindergarten and grade 1, In an effort to compare learning so evident in the 1960 data, does attained through listening and through not occur in the 196 3 data. The braille reading, 70 braille readers 50 percent increase in the number were divided into groups designated of braille readers classed as as listening, braille reading, and "ungraded" suggests that this

263 .

peak in the 1960 distribution Nolan, C. Y., & Morris, J. E. for grades was partly the result of an accumulation of children The Japanese abacus as a computational of limited academic aptitude. aid for blind children.

Exceptional Children , 1964, 31, 15- 17. Nolan, C. Y., & Morris, J. E. Forty-two junior high students Development and validation of the were pretested for skill in computa- roughness discrimination test. tion of whole and decimal numbers by

any method they chose . Continuous International Journal of Education instruction and practice in the use for the Blind, 1965, XV(1) of the soroban was given for eight months, and post-testing followed. The Roughness Discrimination Statistically significant increases Test (RDT) was developed as a reading in computing speed and accuracy were readiness test for braille after pre- demonstrated which verifies that the liminary research revealed the growth soroban is a practical and efficient of ability to make tactual discrimina- approach in overcoming the previous tions. The final version of the RDT computational problems of blind chil- is composed of a set of 69 cards upon dren. each of which is mounted 4 pieces of sandpaper. Of these, 3 pieces are alike and one is different, being rougher or coarser than the others. Packard, B. The task for the student is to find the sandpaper that feels different A technique for developing perceptual from the others. The test was found materials for the blind. to have a split-half reliability coefficient of 0.94. Only items having American Foundation for the Blind, a difficulty index between 0.10 and Research Bulletin No. 25, 1973, 0.85 and an item-test correlation 253-255. significant at the 5 percent level or higher were retained in the final Describes the translation into version. No significant differences tactile kinesthetic tasks of selected were found between the sexes in their pages of the Marianne Frostig Program ability to make tactual discrimina- for the Development of Visual Percep- tions and early research indicated tion by construction of raised line that there is only a moderate rela- work sheets. Greater use of existing tionship between discrimination knowledge on learning disabilities scores and IQ. Both predictive and and perceptual handicaps in sighted concurrent validity for the test were children is recommended for workers determined. Predictive validity was with the blind. determined using first-grade students by testing them with the RDT at the first of the school year and then testing their reading at the end of Pape, R. the year. It was found that the RDT correctly predicted whether these The effect of visual defects in edu- students would be in the upper or cation and at work. lower half of the group, 70 percent of the time when reading errors was The Visually Handicapped, 1972, 1, the criterion and 75 percent of the 24. time when reading time was the criterion. Concurrent validity was de- Recent studies and inquiries termined for second grade students by have shown that even small visual correlating RDT scores with reading defects have a considerable effect time and reading error scores. Defi- on education and vocation. By re- nite but small relationships signifi- sults of experiments with partially cant at the 5 percent level of seeing and one-eyed persons factors confidence wer e found for both cri- of impediments are made explicit, teria. and they are applied to the manifold group of the partially seeing in the narrower sense of the word.

264 .

Physiological and optical mechanisms to measure self-image more objective- are discussed, as well as the limited ly. No clues for a specific self- possibilities of optical aids. It is image in visually handicapped stu- pointed out that in each individual dents was found in the sample. The case an experienced oculist had to be use of control groups is suggested called upon to explain the degree of in further research. impairment and the applicableness of optical aids.

Postel, J., Caillon, N., & Neu, C.

Patton, W. E. Oral language of the blind child.

Research on criteria for measuring Hevue de Neuropsy ahiatrie Infantile mobility readiness of adventitiously et d ' Hygiene Mentale de I'Enfance^ blind adults. 1971, 19, 21-32.

The New Outlook for the Blind, 1970, Distinguishes blind children's 64, 73-80. speech from that of sighted children in 2 ways: (1) at the articulation This research on the dynamics of level disturbances are most frequent readiness for mobility service had 2 because imitation of phonatory move- major research goals : to attempt to ments are missing, particularly those identify specific dynamic factors of which play a part in facial expres- readiness of adventitiously blind sion; (2) and at the expression and adults, and to construct an instru- content level, verbalism is found in ment for measuring total readiness. 50 percent of blind children. This Fourteen criteria were found to have verbalism decreases in proportion as, a positive relationship to both the during language learning, reliance on readiness score and the actual subse- concrete tactile examples increases. quent performance score. A readiness The importance is noted of affective instrument composed of 8 criteria was development, especially the child's fovind to have a positive correlation early relationship with his mother, with performance, significant at less whose attitudes regarding the blind- than the 0.01 level. This research ness of her child may have a greater study appears to have identified im- influence on his language development portant specific factors related to than the sensory handicap itself. mobility readiness. Although the relationship between the resulting research instrument and performance does suggest its usefulness in assessing Postel, J., G'stell-Jeannot, M. C, readiness, further research is indi- Krief, J., & Postel, M. cated. Study of an initial ten cases of tactile dyslexia in children blind from birth. Pfeiffer, H. Review de Neuropsy ahiatrie Infantile Investigations of the self-image in et d'Hygiene Mentale de I'Enfance, visually handicapped youths 1972, 20, 733.

Heil Padagogik , 1972, 41, 266-272. Notes that tactile dyslexia has characteristics similar to visual dys- Studied the self-image of 24 vi- lexia. The distinguishing features sually handicapped high school stu- are only those specific to blindness dents by asking them to describe "how and tactile perception. Tests used I am" and "how I would like to be." for sighted dyslexic children had to Results were evaluated by categorizing be adapted for blind children. Graphic 12 basic elements contained in the es- recording of the movements of digital says, e.g. mentioning of interests, reading showed that these were dis- handicaps, satisfaction, or plans for organized with numerous vertical and the future. Evaluation indicates that reversing movements, indicating major the method of assessment should be difficulties in motor coordination. strengthened by questionnaire scales It was confirmed that dyslexia can

265 . affect the reading and writing of lin- person's pocket and vibrates when it guistic signs whose material basis receives a signal from the induction and perception have no link with the field. The length of vibration can sphere of vision. be varied according to a pre-selected code. The TAC-COM was evaluated as (1) a fire alarm, (2) doorbell system, (3) end-of-line indicator on Priest, H. F. typewriter or Perkins machine, (4) mobility aid device, and (5) general Preliminary report on research com- communication systems, e.g. time- pleted in 1966 using the Kay Ultra clock, office call, telephone ring. Aid for the blind. All these were found effective and plans are under way to investigate American Foundation for the Blind^ other uses. Research Bulletin No. 17, 1968, 83- 90.

This experiment involved the use Randall, T. J., & Springer, D. of the Kay Ultra Aid by 24 subjects over a period of about 13 weeks. For Optical aids: An interdisciplinary the purposes of comparing different prescription sorts of training, the sample was split into three groups: one received The New Outlook for the Blind, training with movement, another re- 1973, 67, 12-19. ceived training with no movement, and the third group received no training. The roles of the optometrist who After the training sessions all sxjb- prescribes low vision aids to par- jects made approximately 2,000 judg- tially sighted persons and other pro- ments, spread over 12 to 20 half-hour fessionals who provide training and sessions, of the size and distance of follow-up services are related to the test objects. In general, distance successful use of such aids. The nec- judgments were much more precise than essity of acquiring general, non- size judgments and they also were less optometric information concerning the affected by changes in the total stim- prescription and the uses to which it ulus situation. In particular, per- will be put by the patient is empha- ceptual constancy, a main interest of sized, in an informal study involv- this study, was found to operate more ing 25 patients it was found that 8 with distance than with size judg- had stopped using their prescribed ments. Differential training had aid. Factors affecting success of little or no effect, probably because the aid prescription: (i) vocational the training period was so brief. change; (ii) vision change; (iii) un- met vision needs; (iv) motivation; (v) type of prescription; (vi) training. When an optical aid program in- Proscia, V. A., Silver, S., & Zu- corporates an effective training pro- malt, L. E. gram and follow up care scheme , these factors become less significant Joint enterprise undertaken between hindrances to the successful utiliza- two centers for development and tion of optical aids in vocational evaluation of a tactile communication training and/or educational programs. aid for deaf -blind persons.

American Foundation for the Blind, Research Bulletin No. 26, 1973, 205- 211.

The TAC-COM is a vibro-tactile communication system developed by the M.I.T. Sensory Aids Evaluation and Development Center. It consists of a sender and receiver, utilizing induction principles as the communication link. Induction loops are set up in the location (house, factory, etc.). The receiver is worn in the

266 Rogow, S. M. involved in this performance are examined. It is concluded, by implica- Language acquisition and the blind tion, that subjects with monocular retarded child. vision and a similar combination of factors could develop similar phe- Education of the Visually Handicapped, nomenal reading abilities on appro- 1972, IV, 36-40. priate materials.

The author reports on a 3-year study of the speech patterns of a nonverbal nine year old blind girl. Schiff, W., & Isikow, H. These speech patterns, consisting of, for exanple, refusals, demands, or Stimulus redundancy in the tactile repetitions of nonsense syllables, perception of histograms. indicated that while language was comprehended by the child in terms International Journal for the Educa- of its syntactic and semantic aspects, tion of the Blind, 1966, XVI(l). it was devoid of its social function as a means of communication. Most of An attempt was made to evaluate the child's speech concerned an ac- the relative effectiveness of several tion or a need; her concerns were re- different ways of presenting tactile lated solely to herself and her pre- histograms for interpretation by occupation with her own feelings. blind persons. The materials varied Subjects about which the child had in the size of differences in histo- intense feelings were never referred gram bar-length, regularity of histo- to directly; when after 2 years she gram, tactile quality of information was able to verbalize these, they indicating bar- length (mode) , and were disguised by expressing the op- redundancy of stimulus information. posite of what she actually felt. Forty blind high school students The author concludes that failure to matched for IQ across conditions speak, or failure to use meaningful served as subjects. speech, does not indicate that a child has not acquired linguistic Except for tactile quality and competence. Rather, it reveals so- stimulus redundancy, all variables cial isolation and experiential de- had significant effects on response privation, which robs language of time. Most interactions of the vari- its function as a means of social ables also proved significant. Tac- interaction and communication. tile quality (mode of presentation) and redundancy had no independent significant effects on error scores but was significant in interaction

Schale , F. C. with "difficulty" variables, with the most redundant presentation providing Exploring the potential of the mon- fewest errors when size differences ocular blind for faster reading. were small. When size differences were moderate or large, different Academic Therapy, (Sum.), 1972, textures or outlines were quite ef-

7(4) , 401-410. fective means of indicating different areas. The time and error measures Tested two predetermined gifted were relatively independent of one speed readers on their ability to another, as indicated by a non- read and comprehend a page at a significant correlation of +0.23. glance with one eye covered as a Practical considerations of the find- means of evaluating the same poten- ings bearing on preparation of tac- tial for the monocularly blind. The tile materials for blind students phenomenal abilities of the two sub- were discussed, as were implications jects are described in detail and for perceptual theory, and research the relationship to eidectic memory techniques with tactile materials. is discussed. Both subjects, using only the right eye, were able to scan non-fictional articles of general interest at page-at-a-glance rates of less than one second per page with

excellent conprehension . The factors

267 . , ; ;

Singer, J. L., & Streiner, B. F. Moreover, this decrease was sufficient to enable appropriate learning Imaginative content in the dreams behavior to occur. Because svibject and fantasy play of blind and sighted was successful in this task she be- children gan to shine--this was reflected in her work, behavior and self-image. Perceptual and Motor Skills, 1966, Class members began to accept her. 22, 475-482. When it was suggested to her that her rocking could also be reduced this In order to test hypotheses grow- occurred without reinforcement. ing out of a consideration of the role of vision in the development of a differentiated capacity for imagination, 20 pairs of otherwise matched Smith, B. F. blind and sighted children (aged 8 to 12) were studied. Interviews and The social education of deaf -blind spontaneous accounts of their play, children at Perkins School for the fantasies, and dreams were recorded Blind. and rated independently by judges for

Imaginativeness . Results indicated The New Outlook for the Blind, 1966, that sighted children proved more 60, 183-186. imaginative in all 3 areas , with blind children showing generally a concrete The author describes the methods and limited fantasy content except for by which deaf-blind children at Perkins their greater reliance on imaginary School are helped to develop independ- companions. The author suggests that ence and social awareness, by examin- in view of these findings special ing the roles played by the adults in- attention to the training of the blind volved in the children's training. in imaginative story-telling and prac- These adults are: (1) the Attendants tice of fantasy play would be desir- (involved in direct care of the deaf- able. blind children; act as companions; help the children to interact mean- ingfully with the blind children among

whom they live) ; (2) the Housemother Slar, M. J., & Rampulla, J. (supervises the work of the attendants; directs household routine of the chil-

Decreasing inappropriate classroom be- dren) ; (3) the Liaison Officer (helps havior of a multiply handicapped blind resolve any conflicts arising among student. the team of adults on the staff) (4) the Principal (supervises and co- Education of the Visually Handicapped ordinates activities of attendants and

1973, 5(3) , 71-74. housemothers; reviews problems with adult staff and suggests solutions) Immediate token reinforcement, and (5) the Orientation Committee coupled with peer pressure was used (responsible for the continuing in- to decrease inappropriate clapping be- service training of attendants and havior. A 20-year old blind subject housemothers). A deaf-blind child's with learning disabilities, mental re- progress through the school is de- tardation, social and emotional prob- scribed with attention to the parts lems. Due to her desire for attention played by each of the staff in the she had not learned to function as a child's social development. contributing member of a group in the class. Tokens were earned, first, for each 20 minute period in which she re- frained from clapping her hands. Three tokens earned five and each member of her class a bag of sweets. Next the period was increased to 30 minutes. Finally tokens were given at various intervals with a bonus of three tokens if subject did not clap at all during the 2-hour observation period. Inap- propriate hand clapping decreased considerably and was virtually eliminated.

268 ,

Snyder, T., & Kesselman, M. learning has not been determined; (2) day programs for abulatory deaf- Teaching English as a second language blind children with basic self-care to blind people. skills and a determined learning potential; (3) residential programs The New Outlook for the Blind, 1972, for those who are ambulatory, possess 66, 161-166. self-care skills and show signs of learning potential that can best be A method of teaching English as developed through concentrated resi- a second language to the visually dential teaching; (4) custodial pro- handicapped, using initially the grams for severely impaired children aural-oral method (all training con- who have not demonstrated any learn- cerned with mastering the sound sys- ing potential. Realization of these tem, of English before attempting to goals is dependent on interdisciplin- read or write it) is described. Ex- ary coordination of the medical pro-

tensive use is made of tape recorders, fessions , special education systems tactile stimulation, and lessons and state level agencies. whose content encourages the learning of the skills of daily living (mobility, gestures, social relations, handling money, etc.). It is observed Stephens, B. that: (1) blind people learn a second language best in intimate, monolingu- Cognitive processes in the visually al settings; (2) vocational motiva- impaired. tion stimulates the most incentive to learn English; (3) both educated Education of the Visually Handicapped and uneducated blind students can 1972, IV(4) , 106-111. demonstrate exceptional ability to learn a new language; and (4) lack An outline of the Piagetian the- of visual aids is not a handicap to ory of development. Cognitive devel- learning, but the blind individual is opment occurs as a result of inter- aided by tactile stimulation during action with the universe. In the the learning session. Five case his- norm.al child, visual activity plays tories are included. a very important role but the blind child must depend on ears, nose, touch, smell. The origins of intelligence lie in the sensory-motor per- Stein, L., & Green, M. B. iod. This stage is not replaced by subsequent stages but is extended and Problems in managing the young deaf- incorporated into the later, more com- blind children. plex thought processes. The blind child is greatly hampered in this Exceptional Children, 1972, 38(6), stage in development of sensori- 481-484. motor schemas and in general coordination. Verbal coordination cannot com- Attempts to develop some general pensate for this. Action learning is concepts regarding the early manage- vital before the child can progress ment of deaf-blind children and the to conceptual, concrete operations help that can be offered their par- and formal operations stages of ents. Looks at the medical prob- thought/cognitive development. Un- lems--especially of the victims of less blind children have experienced maternal rubella, and the difficult- the concrete world and involved their ies the parents encounter in dealing bodies in acting on it, true concepts with their children. Emphasizes the elude them. A need is stressed for need for early and long-term psycho- tactual, perceptual knowledge of ob- educational management of such chil- jects in order to counterbalance the dren and considers 4 programs for negative effects of visual depriva- meeting their needs: (1) an early tion. management program emphasizing psycho- educational management co- jointly with medical care should be designed to identify and train the child who is non-ambulatory, delayed in self- care skills and whose potential for

269 , . ,

Streitfeld, J. W., & Avery, C. D. reading distance, lighting, and optical or non-optical reading aids. The The WAIS and HIS tests as predictors tests were administered twice, once of academic achievement in a residen- using standard print and once using tial school for the blind. large print. Analysis of the data obtained (utilizing a two-tailed t- International Journal for the Educa- test and analysis of variance) indi- tion of the Blind, 1968, XVIIlO) . cated that the Ss performed as well in standard print as in large print This study is concerned with de- on measures of comprehension and read- termining whether verbal achievement ing speed. Large print, however, was as measured by the Verbal Scale of apparently less fatiguing to read than the WAIS (Wechsler Adult Intelligence standard print, possibly because of

Scale) , or performance IQ as measured the students ' greater familiarity and by the HIS (Haptic Intelligence practice with large print materials. Scale) , is more predictive of aca- Visual acuity was not related to per- demic success for the blind. The Ss formance on the measure of comprehen- were 31 students, 17 male and 14 fe- sion, but those students with high male, all students at a residential acuities were able to read both stand- school for blind children. The WAIS ard and large print faster, and ex- and the HIS were administered to all perienced progressively less visual Ss ; a rating scale (1 to 14) based fatigue on large print. upon teachers' ratings and grades was devised to rate each Ss ' academic achievement, for each individual school subject and for overall per- Sykes, K. formance (including the 4 previous years). Results indicated: (1) The Print reading for visually handi- mean IQ for the present sanple was capped children. slightly below standardization groups — for the HIS the difference was sig- Education of the Visually Handicapped nificant at the 0.01 level. (2) For 1972, IV(3) , 71-75. the totally blind, the WAIS and the HIS are equally good at predicting Most pupils who are legally blind grades. (3) For the partially sight- do have some residual vision which ed, the WAIS Verbal Scale appears to tends to be ignored. It is very im- be a better predictor, perhaps be- portant, with the present trend to- cause those with some sight are less wards integration of blind into ordi- likely to develop their tactile and nary schools, that pupils are encour- kinesthetic skills and are thus at a aged to read ink-print wherever possi- disadvantage when blindfolded for the ble, especially standard ink-print so HIS test, thus confounding the result- that the pupil can use the same materi- ing score. als as his peers. Each child should have a desk of the right size, to minimize general fatigue; his own lamp which he can alter in intensity Sykes , K. C. and position; and the optical aids appropriate (plus training in their A comparison of the effectiveness of use and frequent re-evaluation) standard print and large print in fa- Readiness for reading is very im- cilitating the reading skills of vi- portant--many visually handicapped sually impaired students. children are developmentally delayed and need training in verbal/auditory Education of the Visually Handicapped skills, mobility and generalized body

1971, III{4) , 97-105. movement. They need social involvement and wide experience so that they In order to determine whether can relate what they are reading to standard print is equally or more ef- their own life. Reading material must fective than large print in determin- be relevant. Use of the Barraga Scale j ing reading ability, 41 visually im- suggests development of visual skills ' paired students, 2 8 boys and 13 girls, and these have transfer value to other ranging in age from 13-20 years, were learning situations. Reading materi- given 2 of the forms of Series 2 of als must be drawn from a wide variety the Davis Reading Test. The Ss were of sources to ensure children learn free to choose their reading position. to handle printed matter other than

2 70 , . ,

in one standard form, e.g. in science development of sighted and blind chil- it is necessary to handle graphical dren. The literature points to the and illustrative material and techni- difference in play and its role be- cal vocabulary. Skills of skimming, tween sighted and blind children. scanning and reviewing must be taught Blind children lack the desire to and research into rapid reading is play and their play pattern is lim- urgently required. The teacher must ited. Form and content of their sto- be taught use of scales and tests ries, dreams and play suggest a which can provide information into greater concreteness and lack of flex- visual learning. Finally all avenues ibility or associational variety. The of input must be explored and teacher- role of play in the intellectual de- materials produced. velopment of the blind child differs in regard to acquiring knowledge of reality of objects and situations, language and abstract functioning, Tait, P. and of inner life. Since play is a vital element in educational develop- A description analysis of the play of ment, educators of blind children may young blind children. need to foster spontaneous play activities before any meaningful academic Education of the Visually Handicapped learning can be accomplished. The

1972, IV(1) , 12-15. blind child must be taught to engage actively, creatively, and independently In order to determine whether or in spontaneous play activities. not the play of blind children dif- fered from that of sighted peers during a controlled play situation, 29 legally blind children, 4-9 years Talkington, L. W. of age, were presented with special play materials during individual 15- An exploratoiry program for blind- minute free-play sessions. Analysis retarded. of the narrative case records indicated that the blind children dif- Education of the Visually Handicapped

fered from the seeing children in the 1912 , IV(2) , 33-35. following ways: (1) the blind children

had an inability to comprehend the Twelve blind-retarded residents ' space around them or spatial bounda- of a public institution for the men- ries in general, as shown by their tally retarded acted as subjects in concern with the partitions in the an intensive stimulation/activity room, the roles assigned by them to program designed to increase activity a cardboard box, etc.; (2) blind and skill in communication, socializa- children showed less flexibility in tion, sensory stimulation, and mobil- their approach to play materials ity. Basic reinforcement techniques

(e.g. , a hat was always used as a hat were utilized throughout, including by blind children, while sighted chil- use of both food and positive verbal

dren used hats as beds, bowls, etc.) ; responses (praise and encouragement) (3) blind children asked more ques- Pre- and post-measures, using the tions than sighted children and, in Verbal Language Development Scale and general, attempted to establish con- the Vineland Social Maturity Scale, tact with the observer more frequently revealed a 1.9 year increase in than did the seeing children. The language age and a 1.3 year increase author discusses the role of play in in attainment age by the end of the the education of the blind child. program.

Tait, P.

Play and the intellectual development of blind children.

The New Outlook for the Blind, 1972, 66, 361-369. Reviews research on the use of play activities for the intellectual

271 .

Thier, H. D. a non-solution might also contain a solution as part (subsystem) of the Laboratory science for the visually mixture. This is done by allowing handicapped blind child. the filtered liquid to evaporate. If the liquid is only water, little The Hew Outlook for the Blind, 1971, or no residue is left. If it is ac- 65, 190-194. tually a solution, a great deal of residue remains and the blind child A "Direct Experience" science can determine that the liquid was program is described and two examples also a solution. are discussed fully. The first is a biological science unit entitled "Or- Part of the development of the ganisms." In the direct-experience, program is the design of evaluation laboratory approach of the program, activities. This is as yet uncom- children learn about such concepts as pleted, but will involve developing birth, death, reproduction, habitat, tests to determine the intellectual by experiences with classroom aquaria development of visually handicapped and the fish and plants found in them. children, the success of the adapta- Blind children are able to discover tions made in the program, and the the characteristics of fish directly effectiveness of the program as a by utilizing a special aquarium set- whole. up--a second duplicate plastic aquarium with holes in the bottom, is placed in the first and the water comes through the holes from the Tillman, M. H. first into the second aquarium. The fish is then put into the second The factor structure of verbal abili- aquarium, the blind child tips the ties of blind and sighted children. second aquarium so that almost all the water runs out and the fish is Paper read at 1967 AERA Annual Meet- trapped in the remaining water at ing. the bottom. The blind child can then put his hand in and feel the movement The purpose of the present in- of the fish, its shape, action of vestigation was to describe the per- fins, etc. It has been found that formance of blind and sighted chil- the child can even lift the fish out dren on the Verbal Section of the of the water to explore it further Wise. Performance was evaluated in without damage to the fish . The terms of factor comparisons between blind child is given a piece of clay sub-tests using items as variables. from which to develop his own model Specifically, the emphasis was placed of what the fish looks like. Similar on developmental aspects of perform- explorations can be made with snails, ance. Do blind and sighted children plants, and filamentous algae. "Food- of similar age and sex perform equal- web" concepts are taught by placing ly well on the WISC Verbal sub-tests? guppies in a tank with goldfish. The Do both groups have comparable factor guppies ' sxibsequent disappearance structures? leads the children to develop hypotheses for the cause of the disappearance Results for each sub-test in- and after experimentation they can be- clude test statistics, tetrachoric come aware of the various links in the inter-item correlations, and princi- chain ple components factors with varimax rotation. The obtained factor struc- A physical science unit on liq- tures were quite sim.ilar; however, uid solutions is also described. the communalities for the blind on Blind children are able to determine all sub-tests except Arithmetic were whether or not a liquid substance is low. The application of factor anal- a solution (a mixture which is clear, ysis to the tetrachoric inter-item not cloudy) by the use of filters. correlations for each of the WISC Any evidence of a residue left in the Verbal sub-tests can most parsimoni- filter paper after the liquid is ously be described by a proximity- poured through, is an indication that difficulty pattern. In terms of the liquid is a non-solution. A educational objectives, there are further test is made to determine several conclusions to be made about whether the liquid thus identified as the blind. First, there appears to

272 be a lack of integration among educa- correlations for each of the WISC tional experiences with the result verbal subtests can be described by that each bit of knowledge is iso- a proximity-difficulty pattern. lated and cast into a separate frame Across groups , the obtained factor of reference. Second, verbal abili- structures were similar on Arith- ties focus on a basic vocabulary with- metic, Information, and Comprehension out much elaboration. Third, the but less so on Similarities and Vo- blind tend to approach abstract con- cabulary. Fewer factor loadings and ceptualization problems from a con- weaker communalities for the blind on crete and functional level and conse- all subtests except Arithmetic sug- quently lag behind the sighted chil- gest a greater specificity in the dren. And finally, the blind are organization of abilities sampled quite comparable to the sighted in by these tests. In terms of educa- numerical ability. tional objectives, there are several conclusions to be drawn about the blind: first, there appears to be a lack of integration among educational Tillman, M. H. experiences with the result that each bit of knowledge is isolated and cast The performance of blind and sighted into a separate frame of reference; children on the Wechsler Intelligence second, verbal abilities focus on a Scale for Children: Study I. basic vocabulary without much elaboration; third, the blind tend to ap- International Journal for the Educa- proach abstract conceptualization tion of the Blind, 1967, XVI(3). problems from a concrete and functional level and consequently lag The purpose of the present in- behind sighted children. Finally, vestigation was to describe the per- the blind are quite comparable to the formance of blind and sighted chil- sighted in numerical ability. dren on the Wise. Performance was evaluated in terms of item-difficulty curves, t-tests , and subtest relia- bilities. Results indicated that Tillman, M. H., & Williams, C. blind children score about the same as sighted children on Arithmetic, Word associations of blind and sight- Information and Vocabulary, but did ed children to selected form classes. less well on Comprehension and Simi- larities. When IQs were pro-rated A.E.R.A. , 1968. from Arithmetic, Information and Vocabulary Scores, the resulting IQ More than thirty years ago, Payne was shown to be higher than an IQ (19 33) studied free associations of based upon Arithmetic, Information, blind children to a modified version Vocabulary, Comprehension, and Simi- of the Kent-Rosanoff word list. She larities. The depressing effect pro- reported that blind children gave more duced by Comprehension and Similari- idiosyncratic responses, more per- ties was shown to be offset by a severative responses , and fewer com- spuriously high Digit Span score. mon responses: consequently, Payne Moreover, if desired, the 6 subtest suggested that blind children had a Wise IQ could be accurately predicted restricted field of association. The from the 3 subtest IQ based on Arith- present investigation, also using word metic, Information, and Vocabulary. association techniques, was designed to examine levels of syntactic development in blind and sighted children. Since association on the basis of syn- Tillman, M. H. tactic function, that is, homogeneous or paradigmatic responding, is The performance of blind and sighted highly related to the development of children on the Wechsler Intelligence syntax in children, word association Scale for Children: Study II. responses in the present study were scored correct when the response mem- International Journal for the Educa- ber was the same part of speech as tion of the Blind, 1967, XVI (4). the stimulus word.

The application of factor- analy- Thirty-five blind children, seven sis to the tetrachoric inter-item at each age level from seven through

273 . , eleven years, were obtained from three point, and, on the basis of a con- Southeastern residential schools. A servation of substance experiment sighted group comparable to the blind with 189 blind and partially-sighted in age, intelligence, number, and children, it is inferred that while socio-economic background was ob- the best of them are performing on a tained from three communities in par with the best of their sighted Georgia. Two tasks were involved: peers, the age range in which con- a word association (WA) task (Brown servation is attained is more extended and Berko, 1960) consisting of 36 for the visually handicapped. It is high-frequency stimulus words from suggested that Piagetian-type tasks six parts-of-speech categories , count may be of direct practical use to nouns, mass nouns, transitive verbs, teachers for diagnostic and teaching intransitive verbs, adjectives, ad- purposes verbs; a word usage (WU) task consisting of nonsense syllables introduced within a syntactical context such that the six form classes of interest were Tuttle , Dean W. represented; in the latter task, the child was asked to supply a meaning A comparison of three reading media for the nonsense syllable. for the blind.

Results (Groups X Level Anova) Education of the Visually Handicapped indicated that on WA tasks both 1972, IV(2) , 40-44. groups scored about the same level, that rate of development was about One hundred four subjects took the same, and that the sequence of ' three eauivalent forms of the Reading development was highly correlated. Versatility Test, intermediate level-- On WU however, a ceiling effect was one in braille, one in normal reading, obtained for the blind group but not and the third in compressed speech. for the sighted group. Several pos- Comprehension scores, both of braille sible explanations were considered: and normal recording, revealed 4 lev- biased low estimate of the intelli- els: (1) low braille readers, high gence of the blind group; effects of listeners; (2) high braille readers, heterogeneity of variance; a possible low listeners; (3) low braille readers, confounding of subject and task vari- high listeners; (4) low braille read- ables; and a face-value hypothesis. '. ers, low listeners. Analysis of vari- It was suggested that the use of ance over the four levels and overall younger children would clarify the indicated the following conclusions: nature of differences obtained in (a) for total sample, no difference the WU task. In short, while it is in comprehension among the three me- possible to infer specific language dia; (b) for the total sample, reading deficiencies based on research litera- by braille took twice as long as by ture, the results provided no support listening to recording, and almost for a developmental lag in the blind three times as long as by compressed group's acquisition of syntax. speech; (c) for the total sample, reading by listening to compressed speech was more efficient than reading by listening to normal recording Tobin, M. J. which, in turn, was more efficient than reading by braille. Conservation of substance in the blind and partially sighted.

British Journal of Educational Psy- Umsted, R. G.

chology , 1972, 42, Part 2, 192-197. Improving braille reading. Previous investigations of the effects of visual impairment on the de- The New Outlook for the Blind, 1972, velopment of perceptual and cognitive 66, 169-177. skills are discussed and related to the practice of making separate educa- In order to determine whether in- tional provision for the visually struction for increasing the speed and handicapped. The alleged developmental accuracy of the recognition of one- lag is examined from a different view- cell whole word and part-word signs,

274 ,

the two-cell contractions , and the Wachs, T. D. short-form words in English braille would improve the braille reading Personality testing of the handicapped: process, 72 high school students were A review. divided into high, medium and low levels of reading by scores on a si- Journal of Projective Teahniques

and Verbal TAT) , it is concluded that The New Outlook for the Blind, 1973, there is no one major test upon which 67, 161-167. the clinician can rely. It is suggested that best results would be The sequential orientation and obtained from use of a battery of mobility curriculum in use at the tests consisting of the Rotter ISB Iowa Braille and Sight Saving School the Insight Test, and the MMPI. consists of 127 lessons divided into 5 units of instruction (indoor travel, campus travel, residential travel, basic business travel and advanced bus- Weiner, L. D. iness travel) . In addition to the motivational possibilities built into The performance of good and poor the curriculum, the instructional braille readers on certain tests in- staff have also experimented with a volving tactual perception. mobility quiz contest and with awards and cash prizes for achievement and International Journal for the Educa- extra-curricular travel experience. tion of the Blind, 1963, XII(3), 72-77. The specific skills and tasks involved in each unit are listed. Two groups of 2 5 children given 6 experimental tests of tactual perception. Three tests were simple in form; 3 required complex tactual preparation. The 2 groups divided into

275 , . . good and poor braille readers — Group the interpretation of human inter- I and Group II. Both groups per- action in general and for the under- formed similarly on the simple tactu- standing of personal and world ex- al perception tests but Group I per- perience of visually handicapped formed significantly better on the persons in particular. After a short complex tests. No definitive rela- discussion of the concept of "atti- tionships found between tests and tude" and the methodological limita- the variables of mental age, chrono- tions of research concerning atti- logical age, IQ ' s , and reading tudes in this special field, several achievement. Concluded that other studies dealing with attitudes towards factors might be related to tactual the visually handicapped are reported. perception, e.g. increased neural This report is arranged in four fields sensitivity in fingertips of good of problems: A, studies concerning the readers , high development of gross determination of attitudes of the see- and/or fine motor coordination , and ing public towards the blind give possession of this skill in isolation only few indications of difference or of influencing factors. negative attitudes. The dependence of results from research methods was already pointed out; B, some studies concerning the attitudes of persons

Weininger, 0., Rotenberg, H . , & having daily intercourse with visu- Henry, A. ally handicapped people could state a wide spread of opinions; C, based Body image of handicapped children. on studies of Sommers , Cowen and others the attitudes of parents of Journal of Pers onality Assessment blind children are reported, and 1972, 36, 248-253. mechanisms are shown, which parents apply to be able to put up with the In an investigation of the body lot of having a blind child. The im- image of handicapped children, eight portance of parental attitudes for institutionalized children with spina the behavior and self-concept of the bifida were compared with eight simi- child is accentuated; D, by the ex- larly afflicted children residing at ample of the two personal testimonies home and attending a special day- the inconsistency of attitudes to- school for handicapped children. wards blindness among the blind them- A control group of eight normal chil- selves is illustrated. Inconsistency dren were matched for age and sex, as and uncertainty in front of one's own were the two handicapped groups. All position could be found by several subjects were within the normal range studies. Finally two theoretical con- of intelligence. A make-a-person task cepts which are frequently met in was individually administered. The literature are briefly described and three-dimensional figures created by discussed. They are arranged within the subjects were duplicated in exact a more general model developed by proportion on graph paper to facili- Allport tate measurement and to obtain pertinent data. Results indicate that the institutionalized group has a more distorted body image than that of Williams, M. either the group attending day-school or the normal group. Braille reading.

~~The Teacher of the Blind, 1971, LIX(3), 103-116. Weinlader, H.~~

"Over 600 Ss , aged 10 to 18 years. in so- "Attitude" as a basic concept the Ballard Silent Reading handi- were given cial psychology of the visually Test, transcribed into braille. It capped. was found that about 4 percent of those from 10-16 years were unable to 93. The Visually Handicapped , 1972, 1, complete the test sati sfactorily Several factors were s uggested as always borne As human behavior is responsible; it is likely that addi- and molded by social factors , this tional physical or men tal handicap pointing out the article begins with to blindness accounted for a large importance of social psychology for

276 . proportion, while the position was development. The disturbance is mo- aggravated by factors such as late tivated by frustration, generating admission to blind school, absences hostility and guilt. These emotions from school, a negative attitude to- distort communication and learning wards the learning of braille, etc. and threaten family happiness. Cor- For the 60 percent who completed the rection of the problem must deal with test a reading rate of about 100 the entire disturbed constellation." words per minute was found to be average. Two small groups of sighted children, of 11 and 14 years of age, were given the same reading and com- Z ah ran, H. A. S. pletion test. It was found that when blind Ss of comparative ages had mas- A study of personality differences tered the mechanics of braille read- between blind and sighted children. ing, their ability to comprehend the test narrative was equal to that of British Journal of Educational Psy- their sighted peers, but that their chology , 1965, 35(3), 329-337. reading rate was only about half as fast." In the literature there are two different points of view regarding the possible personality differences between the blind and the sighted. Winkley, W. The first considers that blindness leads to compensatory behavior, pos- Public high school or residential sibly accompanied by personality high school for blind students? maladjustment and introversion. The second implies that in regard to ba- Education of the Visually Handi- sic personality variables, the proc- capped, 1972, IV(4), 120-121. ess of adjustment in blind persons is not significantly different from Subjects — 28 college students that of the sighted. who spent last 3 years of high school in a residential school. Ninety-five The writer carried out an ex- college students who spent last 3 periment with matched groups of blind years of high school in a public and sighted children to offer quan- school. All legally blind. Ques- titative data on this problem. The tionnaires were sent out to reha- Williams' Intelligence Test for Chil- bilitation counselors to find out dren with Defective Vision was used whether there was any demonstrable for assessing the IQ's. For person- difference in academic success at ality assessment, the writer prepared college between the 2 groups. Re- a Blind Children's Structured Inter- sults showed that both residential view, a Sentences Completion Test and and public schools were doing equally a Sem.antic Differential. The Junior well in preparing students for aca- Maudsley Personality Inventory was dem.ic work in college. The samples also used. were found to be very much alike in variance, too. Conclusion--the Study After a pilot study and admini- suggests that residential/public stration of the tests, the data were schools equally good on academic analyzed and almost all the test re- side, now need additional research sults agreed with each other in giving in the areas of social competence and statistically non-significant differ- social adjustment. ences in favor of the sighted, thus supporting the second point of view.

~~Yamprey , N~~

~~Problem of the defective child.~~

~~Psy chologiaal Abstracts , 1964, 38, Item 3007.~~

~~"Every defect, whether physical or functional, major or minor, has an effect on the family nucleus which has a further effect on the child's~~

~~277 Zweibelson, I., & Barg , C. F. which was significant at the 0.05 level, the blind subjects choosing Concept development in blind children. primarily concrete and functional~~

responses , and the sighted subjects The Neh} Outlook for the Blind, 1967, choosing more abstract responses. 61, 218-222. This supported the prediction that blind children would not use abstract Two groups of 8 subjects each, concepts to the extent of children one blind, the other sighted, well with normal vision. In view of the matched in terms of age, sex, intel- findings of this and of several other ligence, and socioeconomic background, studies, the experimenters suggest had their responses to each word in that the educational techniques used the similarities and Vocabulary sub- in teaching the blind should be re- tests of the Wise rated on the basis evaluated and modified to take into of whether it was a concrete, func- account the blind child's deficiencies tional, or abstract definition. The in abstract concept formation, espe- results indicated a difference be- cially in the integrated school situa- tween the 8 matched pairs of subjects tion .

~~278 . : , ,~~

~~ABBREVIATIONS OF ORGANIZATIONS~~

Editor's Note. Several of our ACM Association for Computing readers have indicated confusion Machinery about the use of abbreviated names ACM Committee on Professional of organizations in manuscripts Activities of the Blind, published both in the Research Bulle- The Mitre Corporation, tin and in certain other publica- 433 North Circle Drive, tions. Authors may also quite natu- Colorado Springs, Colorado 80206 rally use abbreviations of organizations within their own countries AFB American Foundation for the without realizing that not all read- Blind ers will share this common core of 15 West 16 Street, knowledge New York, New York 10011

We shall try to alleviate this works in cooperation with problem by publishing the following AFOB American Foundation for Over- compendium of abbreviations for or- seas Blind ganizations we have come across in 22 West 17 Street our own materials. New York, New York 10011

Additions and corrections will AFIPS American Federati of Informa- be welcomed. tion Processing Societies 211 East 43rd Street, Asterisk indicates that publica- New York, New York 10017 tions are also available from: AOA Administration on Aging Superintendent of Documents U.S. Department of Health, U.S. Government Printing Office Education, and Welfare, Washington, D.C. 20402. 330 Independence Avenue, SW, Washington, D.C, U.S.A. AAIB American Association of In- structors of the Blind APH American Printing House for now known as the Blind 1839 Frankfort Avenue, Association for Education of AEVH Louisville, Kentucky 40206 the Visually Handicapped 1604 Spruce Street BCAB British Computer Association Philadelphia, Pennsylvania of the Blind 19103, U.S.A. BCAB, BCM, Box 9 50,

AAWB American Association of Workers London, WC2 , England for the Blind 1511 K Street NW, CCD Central Council for the Dis- Washington, D.C. 20005 abled 34 Eccleston Square, of Blind and Par- ABAPSTAS Association London, SWIV IPE, England tially Sighted Teachers and Students CEC The Council for Exceptional T. Moody (Secretary) Children Department of Political Economy, 1411 South Jefferson Davis Glasgow, University of Highway 8RT , Scotland. Glasgow, G12 Arlington, Virginia 22202 ACB American Council of the Blind 106 N.E. Second Street Oklahoma City, Oklahoma 73104

~~279 , ,, : , , , :~~

CIE Centre International de ICTA ISRD Committee on Technical 1 'Enf ance Aids Chateau de Longchamp, Housing and Transportation Bois de Boulogne Fack Paris 16®, France S-161 3 Broirana 3 Sweden *DBPH Division for the Blind and Physically Handicapped lEE Institution of Electrical U.S. Library of Congress, Engineers 1st Street and Independence Savoy Place

Avenue , SE London WC2R OBL, England Washington, D.C. 20540 IEEE Institute of Electrical and DES Department of Education and Electronics Engineers Science 34 5 East 4 7th Street Elizabeth House, York Road, New York, New York 10017

London, SEl 7PH , England ILO International Labour Office DHSS Department of Health and CH 1211 Geneva 22 Social Service Switzerland Alexander Fleming House, Elephant and Castle IRCS International Research Com- London, SEl 6BY, England munications System

U.K. : P.O. Box 500 DIG Disablement Income Group St. Leonard's House Queens House Lancaster, England 180-182 Tottenham Court Road, and London, WIP OBD, England University Park Press *ERIC Educational Resources Informa- Chamber of Commerce Building, tion Center Baltimore, Maryland 21202, U.S.A. U.S. Department of Health, Education, and Welfare, IRIS International Research Informa- Office of Education, tion Service 400 Maryland Avenue, SW American Foundation for the Washington, D.C. 20202 Blind 15 West 16th Street GIHP Groupement des Intel lectue Is New York, New York 10011 Handicapes Physiques for the 8 , allee des Myosotis , 8 ISRD International Society

~~54500 Vandoeuvre , France Rehabilitation of the Dis- abled *HEW U.S. Department of Health, 219 East 44th Street Education, and Welfare New York, New York 10017~~

330 Independence Avenue , SW Washington, D.C. 20202 MIT Massachusetts Institute of Technology HMSO Her Majesty's Stationery Office 77 Massachusetts Avenue London, England Cambridge, Massachusetts 02139

IBSH Institut fur Blinden- und *NAE U.S. National Academy of En- Sehbehindertenpadagogik gineering 6804 llvesheim/Heidelberg 2101 Constitution Avenue, NW Schlossstr. 23 Washington, D.C. 20418, U.S.A. West Germany *NAS U.S. National Academy of Science ICEBY International Council for the 2101 Constitution Avenue, NW Education of Blind Youth Washington, D.C. 20418, U.S.A. now known as NBA National Braille Association ICEVH International Council for the 85 Godwin Avenue Education of the Visually Midland Park, New Jersey 07432 Handicapped Perkins School for the Blind 175 North Beacon Street Watertown, Massachusetts 02172

2 80 *NEI National Eye Institute RNIB Royal National Institute for U.S. National Institutes of the Blind Health 224 Great Portland Street 900 Rockville Pike London W.l, England Bethesda, Maryland 20014 SAEDC Sensory Aids Evaluation and NFB National Federation of the Develooment Center Blind Massachusetts Institute of Randolph Hotel Building Technology Des Moines, Iowa 50309, U.S.A. 77 Massachusetts Avenue Cambridge, Massachusetts

*NIH U.S. National Institutes of 02139 , U.S.A. Health 9000 Rockville Pike SIGCAPH Special Interest Group on Bethesda, Maryland 20014, Computers and the Physic- U.S.A. ally Handicapped Association for Computing *NINDS National Institute of Neuro- Machinery logical Diseases and Stroke The Mitre Corporation U.S. National Institutes of 433 North Circle Drive Health Colorado Springs , Colorado 9000 Rockville Pike 80206 Bethesda, Maryland 20014 *SRS Social and Rehabilitation NRA National Rehabilitation Asso- Service ciation U.S. Department of Health, 1522 K Street, NW Education, and Welfare Washington, D.C. 20005 Washington, D.C. 20201

*NSF U.S. National Science Founda- TNO Nederlandsche Centrale tion Organisatie TNO (Central 1800 G Street, NW Organization for Applied Washington, D.C. 20006 Scientific Research in the Netherlands TNO) NSPB National Society for the Pre- The Hague vention of Blindness P.O. Box 297 79 Madison Avenue Juliana van Stelberglaan 148 New York, New York 10016 The Netherlands

*OEBEH Office of Education Bureau for *VA Veterans Administration Education of the Handicapped U.S. Government U.S. Department of Health, Washington, D.C. 20420 Education, and Welfare Office of Education VIDPI Visually Impaired Data Proces-

400 Maryland Avenue, SW sors , International Washington, D.C. 20 202, U.S.A. P. Bryne (Secretary) 1130 S. Michigan, Apt. 716 RCEVH Research Centre for the Educa- Chicago, Illinois 60605, U.S.A. tion of the Visually Handi- capped VOS All Russia Society for the 50 Wellington Road Blind Edgbaston, Birmingham, Novaja Plostchad 14

B15 2EP , England Moscow, U.S.S.R. WCWB World Council for the Welfare RCSB Royal Commonwealth Society for of the Blind the Blind 14 Rue Daru 46 Victoria Street Paris 8, France London, S.W. 1, England WHO World Health Organization RFB Recordings for the Blind The United Nations 215 East 58th Street Secretariat Building New York, New York 10022 833 United Nations Plaza New York 10017, U.S.A. and: Palais des Nations Geneva, Switzerland

~~2 81~~

~~RESEARCH BULLETIN SUPPLEMENT~~

~~Name: AmBiChron Pitch Changer~~

~~Source: Richard F. Koch, Consultant 67 Smith Street, Lynbrook, New York 11563~~

~~Availability: In serial production.~~

The AmBiChron is an electronic analog of the rotary-head pitch changer. It divides the incoming signal into segments. Then, it discards or repeats part of each segment, and splices together what is retained. To lower pitch, it discards and stretches the remnant, thereby increasing wavelength. To raise pitch it repeats so that the output signal is — so to speak — squeezed in time, thereby decreasing wavelength. In either case, no gaps or spurious signals are generated. The AmBiChron is all electronic, and uses all off-the-shelf components which are available from a multiplicity of sources. (U.S. Patent 3,816,664, foreign patents pending.)

~~Name: Audible Digital Electrical Test Meter~~

~~Source: Audio Perception 212 Professional Building 150 North Center Street '"'... Reno, Nevada 89501~~

~~Availability: 60 days after order.~~

A digital multimeter for electrical testing by nonsighted persons: results appear both in lighted digits and audible tones corresponding to the reading on the lighted digits. An alternate feature is the availability of a voice reading to correspond to the digits as well as, or in place of, the tones. The meter allows hands-off operation and continuous reporting of variations in the readings being obtained.

~~Name: Audible Electronic Calculator~~

~~Source: Audio Perception 212 Professional Building 150 North Center Street Reno, Nevada 89501~~

~~Availability: 60 days after order.~~

An audible electronic calculator, which can be operated by nonsighted persons: the results appear in lighted digit form; and in audible tones corresponding to the lighted digits. An alternate feature is the availability of a voice reading to correspond to the digits as they are lighted. Standard model available in normal functions together with memory. Upon special order virtually all calculator functions can be included.

~~283 ,~~

~~Name: Audible Multi-meter~~

~~Source: G. P. Roberts lOA Mason Avenue, Cheltenham 2119~~

~~N.S.W. , Australia~~

~~Availability: One off model. Instructions for building further instruments can be obtained from Mr. Roberts.~~

The instrument is battery powered and completely self-contained. The front panel contains a pair of input terminals, a row of six pushbuttons for range selection and the read-out potentiometer. In use, the input terminals are connected to the circuit under test, a suitable range of voltage or current is selected and the read-out knob is slowly rotated until the point is found at which the whistle from an internal oscillator ceases. The read-out knob is fitted with a pointer approximately 2 in. long, the position of which can be readily felt in relation to an embossed scale marked in braille from to 10. It is found that with little practice, blind students can thus obtain measurements with a degree of accuracy comparable to that of a conventional moving-coil meter.

~~Name: Automotive Electronic Engine Timing Instrument~~

~~Source: American Foundation for the Blind, Inc. 15 West 16th Street New York, New York 10011~~

~~Availability: Laboratory prototype~~

The instrument uses a pressure transducer or gauge, in place of one of the spark plugs. The transducer electrically transmits cylinder pressure information, which is electronically converted into a "top dead center" position of the piston travel. This computed piston position is compared to the spark plug firing from the distributor.

A brailled multiposition switch which can be set to conform to the car manufacturer's timing specification is used to set the engine timing by turning the switch to the specified number of degrees before or after "top dead center." The mechanic has an audible or tactile indicator from which he adjusts the distributor for a null.

~~Name: Digital Volt-Ohmmeter for the Blind~~

Source: T. C. R. S. Fowler, B.Sc. University of Bristol, H. H. Wills Physics Laboratory, Royal Fort, Tyndall Avenue, Bristol, BS8 ITL, England. Availability: Experimental prototype. Significant features of this instrument are the use of 12 rocker switches arranged in three groups of four to give 3-decade voltage settings and readout (a 4, 2, 2, 1 weighting code is used), a spring-loaded bar switch to add one further least significant digit when pressed, and a high/low frequency audible output controlled by a voltage comparator. Two rotary switches and a multiplier lead enable one of seven dc voltage ranges or five resistance ranges to be selected. Total component cost is about &15 (U.S. $36). In operation, after a suitable range has been selected (by experiment if

~~necessary) , the measurement is made by switching all rocker switches to the~~

~~284 . .~~

"off" state (left hand side down), and then switching them on in sequence, starting with the most significant digit, each switch being left on if it does not cause the audible tone to change, but switched off if it causes a (high-to- low) change in the output frequency. When the process is complete pressing the bar switch will cause the tone to change. The states of the twelve rocker switches are then sensed by touch giving a reading in 3-digit decimal form. The measurement typically takes about 20 seconds.

The readout requires no extra encoding, and is readily used by sighted people and blind people with no knowledge of braille. The blind man for whom the instrument was developed has commented enthusiastically on the ease of use and unprecedented accuracy which it makes available to him. It would appear worthwhile to develop, for production, a complete digital multimeter on similar lines,

~~Name: Electromechanical Braille Line~~

~~Source: Arbeitsgemeinschaft fur Rehabilitationstechnik e.V. an der~~

Universitat Stuttgart ...... D-7000 Stuttgart 1, Schloss Solitude, Haus 3 West Germany and Institut fur Konstruktion und Fertigung in der Feinwerktechnik (FWT) (Institute for Construction and Production of Miniature Device) Universitat Stuttgart

~~Availability: Preproduction prototype.~~

~~ART and FWT have developed a braille line that is an improvement in braille reproduction. The device takes electrical signals in any form whatever; and~~

~~produces tactile braille symbols . without paper. ^~~

The braille symbols are produced by movable pins arranged in a braille matrix of standard dimensions . The pins can be lowered to erase the line and then raised to form a new line. The pins are held stationary while being read. An advantage of this system is that large quantities of information can be presented tactually without expendables (for example reading of books via a reading machine or video-display computer or telecommunication terminal)

~~The device is a modular display, the module being a six-dot braille cell such that braille lines of any desired length can be made.~~

~~Technical Data Module width (intra cell) 7.3 mm. (0.287 in.)~~

Pin spacing (inter cell) 3.0 mm. (0.118 in.) ' Pin diameter 1.5 mm. (0.059 in.) spherical Writing Rate 30 characters-per-sec Applications Reading Machine Electronic Calculator Display Braille terminal (e.g. teletype) Digital display for measuring instruments Instructional system for teacher-students Instructional system for programmers and many others

~~285 Name: Electromechanical Numeric Braille Display~~

~~Source: George F. Dalrymple , Acting Director, Sensory Aids Evaluation and Development Center Massachusetts Institute of Technology~~

~~Availability: Pre-production prototypes and a one-off working unit.~~

The device can be interfaced with almost any electronic instrument with digital binary coded decimal (BCD) outputs. Several variations on the basic design have been developed for use with various electronic equipment. The concept can be extended to modules that display the full braille cell, providing a complete alphanumeric display.

~~Name: Environmental Sensory System with Audio Output~~

~~Source: Richard G. Beschle, Associate Professor of Life Sciences, Worcester Polytechnic Institute, Worcester, Massachusetts~~

~~- * '. ' ^ Availability: Laboratory prototype.~~

The system scans objects with a Robot Research Model 80 TV camera coded for edge detection (tone sounds only when a change from light to dark or vice-versa is encountered). The camera feeds into a number of hard-wired circuits, which provide a tone to either of two headphones. A voltage-variable oscillator generates the tone and the voltage in the TV oscillator determines the pitch. Frequency of sound indicates position along vertical, higher frequency for higher position, lower for lower position. Variation in amplitude of sound delivered to each ear indicates position along the horizontal. Subjects have been able to identify household objects with the device.

~~Name: Liquid Level Detector~~

~~Source: General Electric Research and Development Center, Schenectady, New York~~

~~Availability: Production prototype.~~

~~Battery-operated electronic device to detect the level of liquid in a cup or glass. When liquid rises up to the two sensors by which the device is sus-~~

~~pended, (about 1/2 in. from the rim) , the sensors begin to act as electrodes transmitting current through the liquid. A buzz is then emitted by the device.~~

~~2 86 :; :; : :;~~

~~Name Man-Machine Tactile Communication Device~~

~~Source Dr. A. Michael Noll, Bell Telephone Laboratories, Inc., Murray Hill, New Jersey~~

~~Availability: Laboratory prototype~~

The device presents the user with three-dimensional tactual-kinesthetic sensory input. The user experiences the shape of the object programmed for display through the device by means of a hand-held movable ball fixed to a shaft. The position of the ball is monitored by means of three potentiometers. The force required to move the ball in any direction is controlled by three motors. Each pair of motors and potentiometers thus controls movement along one axis of three dimensional space. For the user the effect of moving the hand-held ball is that of exploring the shape of the programmed object tactually, using the ball as a probe. FORTRAN-compatible software has been written for controlling the motors in the device. Other programs have been written to simulate objects and surfaces.

~~Name: Paper Currency Identification Device (Revised listing)~~

~~Source Southwest Research Institute 8500 Culebra Road San Antonio, Texas 78284~~

Availability The device is now being manufactured under the name "Paper Money Identifier" by Marchak Engineering and Manufacturing Co. Austin, Texas, and distributed by Applied Rehabilitation Sys- tems, 3902 Idlewild, Austin, Texas. It is also manufactured by Delta Meter Service, 1717 West 9th Street, ,Los Angeles, California 90015.

~~Name Spellex Talking Typewriter~~

Sources Dr. Michael P. Beddoes Department of Electrical Engineering Vancouver 8, B.C., Canada or Dr. Ching Y. Suen Department of Computer Science Sir George Williams University Montreal, P.Q., Canada H3G 1M8

~~Availabi lity Experimental Prototype.~~

A digital instrument which gives the spoken sound of the character each time a key of the electric typewriter is pressed. Designed for both beginner and professional typists, it includes an editor for correcting typing errors. Available for demonstration in Vancouver and Montreal.

~~287 :: :; : ,~~

~~Name: Spelltalk (Audio output for reading alphanumeric.)~~

~~Source: R. L. Longini Medical Systems Engineering Laboratory, Carnegie-Mellon University, Pittsburgh, Pennsylvania 15213, USA~~

~~Availability: Laboratory prototype (fourth generation device built)~~

Electronically produced output is triggered by character recognizer or by machine readable record. Sounds are voice like enough so that 100 percent phonetic output can be learned as a dialect. Forty hours of training (high IQ) permits 150 words-per-minute of English to be understood. Price is estimated at about $300 with suitable input if built in quantity.

~~Name Tactile Digital Readout (Revised listing)~~

~~Source American Foundation for the Blind 15 West 16th Street, New York, New York 10011~~

~~Availability Experimental prototype.~~

The earlier "Digi Tac" and "Digital Voltmeter" prototype has been super- seded by a version which provides a braille display instead of a binary coded tactile readout. The present version uses a four-dot solenoid matrix with electrical encoding. The encoding is done using a "read only" memory from a standard seven segment display device and converting the numbers to braille. A working prototype has been completed using a single braille cell with electronic "digit" switching so that the numbers are read sequentially. A model using 8 or 12 braille cells for an in-line display is under development. The main thrust for this project has been towards providing an electronic calculator readout. The next phase will be to apply these circuits to various digital display instruments.

~~Name Talking Desk Calculator for the Blind~~

~~Source Priv.-Doz. Dr. H. Bebie Institut fur theoretische Physik der Universitat, Sidlerstrasse 5 CH-3000 Bern Schweiz and Prof. Dr. F. Fankhauser Univ.-Augenklinik Inselspital CH-300 8 Bern Schweiz~~

~~Availability: Production prototype. Manufacture and distribution by INTERZEAG AG, CH-8953 Dietikon , Switzerland, scheduled for 1975.~~

Spoken output is in words from zero to nine with a buzzer to indicate the decimal point. A starting button has been added to the regular calculator keyboard to initiate the spoken output. Output can be elicited repeatedly. Also figures typed into the calculator can be elicited by means of the same button, for verification before the calculating process begins.

~~288 :; :: ; ) .~~

~~Name Three-channel vibrotactile speech-reception aid for the deaf~~

~~Source James D. Miller, Central Institute for the Deaf, St. Louis, Missouri 63110~~

~~Availability: Laboratory prototype.~~

Envelopes of signals from (1) an accelerometer attached to the talker's nose, (2) a microphone near his mouth, and (3) an accelerometer attached to his throat have been multiplied by a 100-Hz square wave and the products used to drive three vibrators. The vibrators have been placed on separate fingers of one hand, or on a single finger.

~~Name Transcriptor~~

~~Source Arbeitsgemeinschaft fur Rehabilitationstechnik e.V. an der Universitat Stuttgart D-7000 Stuttgart 1,~~

~~Schloss Solitude, Haus 3 . • West Germany~~

~~and --•'~~

~~Institut fur Konstruktion xind Fertigung in der Feinwerktechnik (FWT) {Institute for Construction and Production of Miniature~~

~~Device . Universitat Stuttgart~~

~~Availability Laboratory prototype~~

~~A laboratory model to demonstrate the functions of an inexpensive portable reading machine for the blind.~~

The Transcriptor developed at the ART and FWT of the University of Stutt- gart represents the first step in the development of reading machine for the blind that can be used equally at work or at home. It has shown the feasibility of building a portable reading machine using a serial optical correlator

~~(OCR) . . ,~~

The characters are scanned by a movable reading head with mask. The characters are compared, correlated and recognized. The recognized signs are transformed into braille and embossed on a strip of paper.

~~Present State of Development~~

~~Type Faces : Courier (IBM) 29 upper case letters on.~~

~~Helvetica 26 upper case letters~~

~~Reading Speed: 6 characters per second~~

~~Error Rate 3 percent~~

~~289 .~~

~~Present State of Development (continued)~~

~~Line Feed: Manual~~

~~Page Size: DIN A4~~

~~Development Goals~~

~~The goal is to develop a reading device about the size of a portable electric typewriter.~~

~~Type Faces: Recognize all normally used letters, both upper and lower cases, numbers, and special signs~~

~~Reading Speed: 15 characters/sec.~~

~~Error Rate: 0.5 percent~~

~~Page Size: Din A4~~

~~Automatic focusing and magnification.~~

~~Automatic line positioning and following. -~~

~~Output: Braille line display Braille strip printer Braille page printer Magnetic tape or spoken word, depending on application, form desired and available equipment.~~

~~Development time: Estimated 2 years.~~

~~Name: Transicon (Revised listing)~~

~~Source: The Scientific Research Foundation Jerusalem, P.O.B. 375 Israel~~

~~Availability: First commercial model is now available under the name "TEXTOBRAIL" from the Israel Electro-Optical Industry, Ltd., P.O.B. 1165, Rehovot 76110, Israel.~~

~~290 ,~~

~~PUBLICATIONS OF NOTE~~

American Printing House for the Research, Rutgers University, Blind. Educational research, August 19 74. 315 pp. (Report to development, and reference group: the National Science Foundation, Report on research and develop- Washington, D.C., Research Ap- ment activities--Fiscal 1974. plied in National Needs, 73-7,

~~Louisville, Ky . : American Printing Research Grant No. GI-39420, Pro- House for the Blind, 19 74. (un- posal No. P312625) published report) Berla, E., & Murr, M. J. Searching~~

~~Aposkitis , C. Le role des oligo- tactual space. Education of the elements dans 1' apparition de la Visually Handicapped , May 1974,~~

cataracte senile. Ana. Oculist , 6 (2) , 49-58. (Paris), 1970, 925-930. 10, A tactile pseudomap was designed and tested with 144 braille read- Aranda, R. R. OILS: A model for the ers enrolled in schools for the social rehabilitation older of blind. Sixteen subjects from each persons with severe visual im- grades 4 through 12 partici- pairment. New York: Center for of pated. The pseudomap used point, Independent Living, New York In- line, and textured areal symbols. firmary, March 19 74, 50 pp. Bliss, James C. Summary of three T. McCrady, R. E., & Ball, S., optacon-related cutaneous experi- Hart, A. D. Automated reinforce- ments in F. A. Geldard (Ed.), ment of head posture in a cerebral Conference on vibrotactile com- palsied retardate. Perceptual & munication, Austin, Texas: The Motor Skills, 1974, in press. Psychonomic Society, 1974, 84-9 4.

Beddoes , M. P. Experimental study of Bliss, J. & Moore, M. W. The the Spellex editing talking type- C, Optacon reading system. Education writer. In Proceedings the of the Visually Handicapped 1973 International Conference on of December 1974, 6(4), 98-102. Cybernetics and Society . (Boston, Massachusetts, November 5-7, 1973) Brothers, R. J., & Huff, R. Manual IEEE Systems , Man and Cybernetics activities the Society, 290-291. of suggested for development of sound localization skills. Louisville, Ky.: IMRC for Beddoes, M. P., Suen , C. Y. , Dix- Visually Handicapped Children, on, B. A., & George, R. G. American Printing House for the Spellex: An automatic talking Blind, August 19 72. 17 typewriter for the blind. In pp. Digest of the 10th International Cox, C. H., Blesser, B. A., & Eden, M. Conference on Medical and Bio- A letter structure grammar, with Engineering, Dresden, logical applications to multifont charac- 1973, 121. ter recognition. Quarterly Progress Report No. 114, Research Laboratory Berkowitz, M. , Englander, V., Ru- of Electronics, MIT, 233-241. bin, J., & Worrall, J. D. An evaluation rehabilitation of Craig, J. C. Pictorial and abstract research. New Brunswick, N.J.: cutaneous displays. In F. A. Gel- and Health Economics Disability dard (Ed.), Conference on vibro- Bureau of Economic Research, tactile communication. Austin,

~~291 , .~~

~~Texas: The Psychonomic Society, Quarterly Progress Report, No. 113. 1974, 78-83. Cambridge, Mass.: MIT Research Laboratory of Electronics, April Curtis, W. S., Donlon, E. T., & 1974, 163-167.~~

~~Tweedie , D. Communicative behavior of deaf-blind children. Geldard, F. (Ed.) Cutaneous communi-~~

Education of the Visually Handi- cation systems and devices . (Con- capped, December 1974, 6(4), 114- ference in Monterey, California, 118. April 17-18, 1973) Austin, Texas: The Psychonomic Society, Inc., 1973. Czeslaw, Luszczynski, Mieczyslaw, 112 pp. Su\vinski. Zatrudnienie niewido-

~~mych w pracy umyslowej (Employ- Genensky , S. M. Closed circuit TV and ment of the blind as white-collar the education of the partially sight-~~

~~workers) , Zakland Wydawnictw crs ed, Santa Monica, California: Rand Warszawa, 1972. Corporation, March 19 70.~~

~~Daugherty, K. M. Listening skills: A Genensky, S. M. , Petersen, H. E.,~~

review of the literature. The New Clewett, R. W. , & Moshin, H. L. Outlook, October 1974, 68(8), 363- A double X-Y platform for Rand- 369. sight-type instruments. Santa Monica, Calif.: Rand Corp., Decem- Erber, N. P. Visual perception of ber, 1974. 43 pp. speech by deaf children: Recent This report describes a mechani- developments and continuing needs. cal platform called a double X-Y Journal Speech and Hearing Dis- of Platform. When this platform re- orders , May 1974, 39, 178-185. places a single X-Y Platform as part of a closed circuit TV system Flanagan, J. L. , & Rabiner, L. R. for the partially sighted, it per- (Eds.). Speech synthesis. The mits a user to read printed or hand- Journal the Acoustical Society of written material and to take notes America, April 1974, 900. of 55, on or copy from that material using pen or pencil. The platform has Evaluation lenses a Fonda, G. of contact two rectangular working surfaces for central vision in high myopia. whose motions in the left or right British Journal Ophthalmology of , direction (x-direction) are com- 141-147. February 1974, 58(2), pletely coupled, but whose motions toward or away from the user (y- Fonda, G., & Thomas, H. Correction of direction) are completely inde- low visual acuity in achromatopsia. pendent of one another. This de- Archives Ophthalmology , January of sign feature allows the user to 1974, 20-23. 91, change from reading (writing) to writing (reading) without having to Food and Drug Administration, Depart- search for the line that he was last ment of HEW. Laser products: Pro- writing (reading) posed performance standard. Fed-

~~eral Register , September 4, 1974, Other techniques for handling this~~

39(172) , 32094-32109. important visual information transfer problem have been explored and Freiberger, H. (Ed.). Sensory aids. are also described in this report. Bulletin of Prosthetics Research However, they do not appear to be BPR 10-20, Fall 1973, 347-364. as cost-effective as the double X-Y Platform. H. Presented Freiberger, Mobility. A series of experiments was per- the Seventy-Eighth Annual Meet- at formed with the cooperation of four of the American Academy of Oph- ing partially sighted people who dif- Dallas, September 16- thalmology, fered in age and ocular pathology. 1973. Transactions the 20, of These experiments were an attempt American Academy Ophthalmology of to determine whether copying from and Otolaryngology , Sept. /Oct. 1974, unbound and bound materials using a 78, 713-721. CCTV system equipped with a double X-y Platform can be expected to be M. M. , & Miller, P. L. Locally Fund, more rapid than copying from similar organized parser for spoken input. materials using a CCTV system equipped with a single X-Y Platform, or

~~292 , . .~~

using neither an X-Y Platform nor Hermelin, B. , & O'Connor, N. Spatial a CCTV system. The results of coding in normal, autistic and these experiments were not conclu- blind children. Perceptual and sive. However, three of the four Motor Skills, 1971, 33, 127-132. sxobjects reported that they could

copy more easily with a CCTV sys- Hermelin, B. , & O'Connor, N. Right tem equipped with a double X-Y and left handed reading of braille. Platform than they could using Nature, June 18, 1971, 231(5303), either of the other two techniques. 470.

Genensky, S. M. , Petersen, H. E. Huggins , A. W. F. More temporally Yoshimura, R. I., von der segmented speech: Is duration or Lieth, J. B., Clewett, R. W., & speech content the critical vari- Moshin, H. L. Interactive class- able in its loss of intelligibil- room TV system for the handicapped. ity? Quarterly Progress Report The Rand Corporation, June 19 74, No. 114, Research Laboratory of 54 pp. Electronics, MIT, 185-19 3.

~~Genensky, S. M. , Petersen, H. E., Hviid, J. Ikke-handicappedes reak- Yoshimura, R. I., von der tioner over for handicappede~~

Lieth, J. B., Clewett, R. W. , & (The reaction of the non-handi- Moshin, H. L. An interactive CCTV capped towards the handicapped.) system for educating partially Denmark: Gylendal, 1972. 335 pp. sighted and some other types of This study is based on inter- The Rand handicapped children. views with handicapped and non- Corporation, March 19 74, 36 pp. handicapped persons to explore emotional and psychological re- M. , & James, G. A. Mobility Gill, J. actions of the non-handicapped maps : The choice of symbols . The toward physical disfigurement. (Royal National Insti- New Beacon An extensive bibliography follows tute for the Blind, London) , Feb- the text on pages 30 2 to 30 7; 1974, 58 (682), 35-38. ruary an English summary appears on Presentation of a set of mobil- pages 323 to 332. ity map symbols adopted after extensive evaluation for touch dis- ICTA Information Center. Disabili- crmination and ease of interpreta- ties and housing requirements tion. Bromma 3, Sweden: ICTA Informa- tion Centre, September 1974.

Gill, J. M., & Tobin , M. J. European (Disabilities include: Allergy, register of research on visual im- Cystic fibrosis; Epilepsy; Ex- pairment. Coventry, England: De- tremely short stature; Impaired partment of Engineering, Univer- hearing and deafness; Impaired sity of Warwick, May 197 4. function of the kidneys; Impaired sight; Psoriasis; Retarded mental

~~Greenberg, G. L. , & Sherman, J. C. development; Motor disabilities; Design of maps for partially see- and Enterostomy operations.) ing children. In International~~

~~Yearbook . Gutersloh, West Germany: Jackson, John S. (Ed.) Proceedings Kartographisches Institut Bertels- of the 1973 Carnahan conference on mann, 1970, 111-115. electronic prosthetics . Lexington,~~

~~Ky. : Office of Research & Engineer-~~

Handikappinstitutet . Report on ing Services, College of Engineer- European conference on technical ing, University of Kentucky, Novem- aids for the visually handicapped, ber 1973. 133 pp. March 19 74. Bromma 1, Sweden: Handikappinstitutet, June 1974. Jacobs, K. Sehbehinderte auf dem Weg 302 pp. zum Beruf: Probleme und Entwick- lungstendenzen. (The visually im- Hermelin, B. Locating events in space paired in search of a profession: and time: Experiments with autis- Problems and developing trends.) tic, blind, and deaf children. Sehgeschardigte , Internationales Journal of Autism and Childhood Wissenschaf tliches Archiv, No. 3 Schizophrenia, 1972, 2(3), 288- (November 1973), S. 64-95. 298.

~~293 ) , : , ,~~

James, G. , & Armstrong, J. An evalua- Kirman, J. H. Tactile perception of tion of the Silva braille compass computer-derived formant patterns 1 - Static tests. The New Beaaon, from voiced speech. Journal of September 1974, 58(689), 225-229. the Acoustical Society of America, January 1974, 55(1), 163-169. James, G. , Campbell, D., & Arm- strong, J. An evaluation of the Lassen, P. Barrier free design: A Silva braille compass: 2 - Field selected bibliography . Washing- trials. The New Beaaon, October ton, D.C.: Paralyzed Veterans of 1974, 58(690) , 253-255. America, Inc., 1974. 103 pp. Contains references lander 22 James, G. , Armstrong, J., & Camp- headings, bell, D. Verbal instructions including standards and used by mobility teachers to give specifications, human factors, navigational directions to their planning, housing, transportation, clients. The New Beacon, April and bathrooms and kitchens.

1973, LVII (672) , 86-91. Mehr, E. B., Frost, A. B., & Ap- ple, L. E. Experience with closed Jensen, O. , & Truelsen, H. T. Maend circuit television in the blind Med Synshandiaap . (People with sensory handicaps.) Denmark: rehabilitation program of the Vet- Mentalhygie jnisk Forskningsinsti- erans Administration. American Journal Optometry tut, 1973. (Danish with English of , June 1973, summary. 50(6), 458-469. {Bulletin of Prosthetics Research , Spring 1974,

10(21) , 54-68.) Kafafian, H. Study of man-machine communications systems for the handicapped. (Second Report, Miller, D., Zuckerman, J. L. , & Rey- nolds, G. O. Volumes I and II, prepared for Holographic filter the U.S. Office of Education, to negate the effect of cataract. Arch. Bureau of Education for the Handi- Ophthalmology , October 1973, capped, Project No. 18-2003.) 90, 323-326. Washington, D.C.: Cybernetics Re- National Academy of A search Institute, February 19, Sciences. preliminary evaluation the 1970. 92 pp. of Bionic Instruments-- Veterans Administra- tion Kafafian, H. Study of man-machine C-4 Laser Cane. Washington, D.C. : communications systems for the Advisory Panel for the handicapped. (Final Report, Vol- Evaluation of the Laser Cane, Sub- ume III, prepared for the U.S. committee on Sensory Aids of the Office of Education, Bureau of Committee on Prosthetics Research Education for the Handicapped, and Development, Division of Medi- Project No. 18-2003 and No. 7-0533) cal Sciences , National Research Washington, D.C.: Cybernetics Re- Council, National Academy of Sci- ences, September 1973. 211 search Institute, 1973. 266 pp. pp.

Nelson, J., & De I'Aune, W. Kakalik, J. S., Brewer, G. D. A quantitative assessment of blinded Dougharty, L. A., Fleischauer, P. D. veterans' comprehension of dis- Genensky, S. M. , & Wallen, L. M. Improving services to handicapped crete time compressed speech. children. (R-1420-HEW) Santa IRCS (Research on: Biomedical Monica, California: Rand, May Technology; Eye; Psychology; 1974. Social and Occupational Medicine) 1974, 2, 1440.

Brewer, G. D. , & Kakalik, J. S. Im- Newland, T. E. Le d' aptitude a proving services to handicapped test 1 'apprentissage pour aveugles children : Summary and re commenda- tions. (R-1420/1-HEW) Santa Mon- (BLAT) . (Manual: The blind apti- ica, California: Rand, May 1974. tude learning test.) Bulletin de Psychologie , 1973-1974, XXVII (310), (5-9), 398-402. manual Kay, L. Orientation for blind per- Test and sons: Clear path indicator or en- embossed material available from vironmental sensor. The New Out- T. Ernest Newland, 70 3 South Race Street, Urbana, Illinois. look, September 1974, 68(7), 289- 296.

294 Nizetic. B. Prevention of blindness: in obtaining a copy of this report Potentialities of a systems analy- are referred to the: ERIC Document sis approach. Acta Ophthalmologica, Reproduction Service, P.O.B. 90, 1974, Vol. 5, 134-144, Arlington, Virginia 22100.

~~Noll, A. M. Man-machine tactile com- Rowell, D. The binaural sensory aid~~

munication. SID Journal , July/ for the blind. In the Proceedings August 19 72. of the Conference on Engineering Devices in Rehabilitation, Copies of this paper are avail- Boston, May 2-3, able upon request to IRIS. 1974. Cambridge, Mass . : Massachusetts Institute of Engineering. 37-41. O'Connor, N., & Hermelin, B. M. The reordering of three-term series Rowell, D. problems by blind and sighted Auditory factors in the design of a binaural sensory aid children. Br. Journal of Psy- for the blind. In the proceedings chology , 1972, 63(3), 381-386. of the 4th Annual Meeting, Bio- medical Engineering Society, Los O'Connor, N., & Hermelin, B. Seeing Angeles, 29- and hearing and space and time. California, January 30, 1974. Perception & Psyohophysics, 1972,

ll(lA) , 46-48. Rowell, D. , Lawrence, P. D., & Per- ron, A. G. Volitional control Office of Demographic Studies. Addi- tional handicapping conditions for a scanning sensory aid for the blind. In the Proceedings of among hearing impaired students : United States 1971-72. Data from the Conference on Engineering De vices the annual survey of hearing im- in Rehabilitation, Boston, 2-3, paired children and youth. Series May 19 74. Cambridge, Mass.: Massachusetts Institute of En- D, Number 14. Washington, D.C.: gineering. 42-45. Gallaudet College, November 19 73. 49 pp. Sachs, R. M. Technology comes to telecommunications for the hear- Oomen , H. A. P. C. Vitamin A defi- ing impaired. Highlights , ciency, xerophthalmia and blind- 52(3), Fall 1973, 16-18. ness. Nutrition Reviews , June

1974, 32(6) , 161-166. Smith, G. C, & Mauch , H. A. Abstract of summary report on the develop- Otto, W. (Ed.). Research in reading: ment of a reading machine for the As reviewed in The Journal of Edu- blind (July 1973) . Bulletin cational Research 19 33-19 73. Spe- of Prosthetics Research , BPR 10-20, cial issue. The Journal Educa- of Fall 1973, 231-232. tional Research , May-June 1974,

67(9) , 386-420. Spahr, J., & Fankhauser, F. On automation of perimetry--problems and Pastalan, L. A. The simulation of solutions. Switzerland: Universi- age-related sensory losses : A new tats-Augenklinik Bern. approach to the study of environ-

mental barriers. The New Outlook, Summary : After a brief descrip- October 1974, 68(8), 356-362. tion of the components of the Octo- pus, an automated perimeter, two Rodabaugh, B. J., Weisgerber, R. A. mathematical strategies, leading Optaaon : Ink print reading for to correct threshold determination the blind. Palo Alto, Calif.: are compared. The efficiency of American Institute for Research, finding the true threshold value September 19 74. 44 pp. is evaluated by means of computer An artificial pa- The American Institutes for simulations. tient being represented by sub- Research study summarized in this simulate variable booklet has been more fully describ- programs which threshold fluctuations and variable ed in a final report submitted to degrees of reliability. It is shown the Bureau of Education for the that the efficiency measured in Handicapped, U.S. Office of Educa- percentage of correct answers ob- tion, Washington, D.C. Organiza- tained is best for one specific tions and individuals interested

~~295 ) ,~~

type of strategy, when a large educational system for the blind and range of mis judgments or uncer- partially sighted is discussed in tainties of the true threshold connection with special education is assumed. in general as well as the regular school system.

Suen , C. Y., & Beddoes , M. P. Devel- opment of a digital spelled-speech reading machine for the blind. Haas, P. Physical education at In IEEE Transactions on Biomedical schools for the partially sighted

Engineering, Vol. BME-2 , No. 6, in the Federal Republic of Germany. November 1973, 452-459. 175 pp. The first part of this paper , & B. Assessing Tobin M. J., Hunter, deals with the results of question- the manual dexterity of the visu- naires , assessing the situation of handicapped. The New Beacon, ally physical education at schools for the , 169-172. 56(687) partially sighted. More than 1000 questionnaires had been sent to teach- Veringa, F. Non-adequate stimulation ers and students (grades 5-9) . Some tool in visual physiology. as a of the results are the following: The Archives Internationales de Physi- preferred kind of sport is swimming, de Bioahimie , 1974, ologie et followed by soccer. Sixty-one per- 82 361-363. (2) , cent of the students would like to have more sport; 9 3 percent of the printing program Vinding, J. Braille teachers feel that advanced training system, description . Denmark: Dansk for teachers is necessary. Twenty- Blindesamfund, Computerized Braille four percent are specially trained Production, 1974. (Unpublished teachers ; only 10 percent of the report . teachers instructing in sport are specially trained in sport and educa- A. impedance Wilber, L. Acoustic tion of the partially sighted. evaluation. Highlights , Fall

~~1973, 52(3) , 4-6. In the second part a skiing course with 13 partially sighted students is described.~~

Abstracts prepared by the Educa- tional University of Heidelberg, Krug, F. & R. Physical aptitude of Department for Education of Blind partially sighted children. and Partially Sighted (Teacher College 198 pp. and Research Center) . Physical carriage, organic efficiency and motor coordination was Copies of the studies, in German, tested in 43 partially sighted chil- are available from the Institut fur dren aged 7-8 years, with a control Blinden-und Sehbehindertenpadagogik group of normally seeing children, 69 Heidelberg, Ziegelhauser Land- matched according to age, place of strasse 35, West Germany. DM o , 15 residence (rural/urban) , sex, and per page. socioeconomic status of parents. The results indicated that the partially Bossert, A. The educational system sighted more often have a poor carri- for the visually handicapped in age (not significant) , to have less Denmark. 187 pp. efficient organs (significant) , and In the context of several earlier have significantly lower scores in studies on the educational facilities their coordinative performance. of foreign countries, this paper describes the situation in Denmark in- tensively. In this country there has Schiller, H., & Siebertz , M. Social been an immense change in the organiz- background of the students in ation of the educational system for schools for the blind and partially the blind and partially sighted dur- sighted. 132 pp. ing the past ten years. The social status of 869 students The description based on an is of schools for the blind and 1713 stu- analysis of papers on special educa- dents of schools for the partially trips, tion, legal texts, educational sighted has been classified according interviews, and correspondence. The

296 . to the device of Kleining and Moore. After an introducing discussion Results show that 66 percent of the of the methodologic problems and the students of schools for the blind and description of the procedure to ob- 70 percent of those at schools for tain the material, the case studies the partially sighted belong to lower are reported in detail. social classes. These classes are significantly overrepresented among The results justify the method fully. the visually handicapped students in The paper gives a good accounting of the special school settings compared to problems of these persons cind raises the normal population in the GFR various questions for further research. (45 percent) . This effect is not due to the additionally mentally retarded children Possible reasons and consequences Thumm, I. Parents of partially sight- of this result are discussed. ed children: Statements on the possibilities of their children at school, occupation, and social Schleehauf, U. Adventitiously visu- contacts. 175 pp. ally handicapped professional The parents of 15 workers. 82 students of a pp. school for the partially sighted were Written biographic materials by interviewed concerning their opinion the author analyzed the problems of on special education of partially 14 visually handicapped adults work- sighted children, occupational possi- ing on academic level. bilities and social integration. The interviews are The description is concerned with reported in detail. The summary shows that the parents the individual anamnesis , visual handicap, direct consequences, search are well aware of the advantage and disadvantage for new possibilities, new beginning, of special schools. The later occupational rehabilitation, new profession, re- situation of their view. children, however, can't be judged by the parents in a realistic manner. Psychological problems of the children stemming from their impairment is mostly denied whereas all children have very few contacts to normally seeing peers.

~~297~~

~~HV1571 RESEARCH BULLETIN. C.l R No. 29, June 1975~~

~~Date Due~~

~~HV1571 c.l R RESEARCH BULLETIN,~~

~~No. 29, June 197 5~~

~~ISSUED TO~~

~~£^ ^.re^\r\r. ^^ ^g^~~