The gestalt line

by

Geert-Jan A. Boudewijnse

Department of Psychology

McGill University, Montreal

A Thesis subrnitted to the Facuity of Graduate Studies and Research in partial fulfiiment of the requirements of the degree of Doctor of Philosophy.

Al1 rights reserved

Geert-Jan Adriaan Boudewijnse, 1996 National Library Bibliothèque nationale I*(of Canada du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395. tue Wellington Ottawa ON KIA ON4 OttawaON KIA ON4 canada Canada

The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de rnicrofiche/~,de reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fkom it Ni la thèse ni des extraits substantiels may be prhted or otherwise de celle-ci ne doivent êeimprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. CONTENTS

RES UME

ACKNOWLEDGMENTS

TRANSLATOR'S NOTE

INTRODUCTION

The lineage

General Framework

Previous work on the history of gestalt theory

Overview

FRANZ BRENTANO

Aristotelean roots

Unity of consciousness

Intentionality

Imer perception

CHRISTIAN VON EMRENFELS

The article

Gestalt quality

Ehrenfels' contribution Gestalt qualities as building blocks

-4 doubtfid expianation

What does the fùture have in store?

EDGAR RüBN

The experiments

The explanation

Rubin and the school of Brentano

Contour versus gestalt quality

CARL STUMPF

Psychological origin of space perception

Substance

Learning

Tone psychology

Stumpf s scientific method

Tone füsion

Cornplex

Rubin's critic

Speech sound

Stumpfs dualism versus his students' monism

Stumpf and his students

Theory of Knowledge BERLirY GESTALT SCHOOL

THE GRAZ SCHOOL

Meinong

Production theory

Benussi

Witasek

KofRa's criticisrn

CONCLUSION

Brentano and the gestalt notion

Some other observations

Divergent opinions

Tasks for the tùture ABSTRACT

The thesis specifies how Franz Brentano inspired some of his students and how those midents, in their tum, infiuenced the next generation of psychologists. Mer outlining the essentials of Anstotie's psychology, the thesis explains some generai positions that Brentano borrowed hmAristotle. It goes on to relate Brentano's concepts of 'presentation,' 'unity of consciousness,' and 'dserence berneen the mental and physicaI,' as well as his cal1 for a

'science of the minci' to ideas of Christian von Ehrenfels, Car1 Stumpf and Alexius Meinong.

Ehrenfels thought that a mental element, which he named gestalt quality, explains why a string of presentations has a certain form. The thesis then iooks at a book of Edgar

Rubin, even though Rubin was not a student of Brentano. His experiments that demonstrated the figure-ground phenornenon, however, were weil known to the Berlin gestalt schwl. My analysis of Ehredels also sheds light on Rubin's theory, a theory that the

Berlin gestalt school seemed to have overlooked, perhaps because Rubin's findings fit so we1I into their own notions.

Stumpf developed his theory of how presentations fom a unity pady in rejection of Aristotle's notion of substance. His theory grounds his criticism of associationisrn.

Stumpf s students, however, would not accept his dualistic view, but they would benefit from the experimental methods that he developed and taught them. Stumpf adapted his theo~yin response to his students' work, and that version formed the basis of his objections against their explanations.

Meinong's theory of how presentations are united was inspired by the Scholastics,

1 and his midents would render it into a psychological format. It is as a review of that latter work that the Berlin gestalt school presented its gestalt notion for the fira tirne in mature forrn.

The conclusion very briefly reiterates Brentano's influence. it also contains Mme general observations regarding the diversity among the gesîalt notions, the wide scope of the gestalt authors, and their zeal for the pursuit of pure scientific knowledge. Cette thèse explique comment Franz Brentano a inspiré plusieurs de ces étudiants et comment ces étudiants, a leur tour, ont influencé la génération future de psychologues.

Après avoir fait resso~irl'essentiel de la psychologie d7Aristotle7cette thèse explique certaines positions que Brentano a emprunté à Aristotle. EUe relate les concepts de

"présentation", "unité de conscience" et "'différence entre le physique et le mentai" de

Brentano ainsi que son penchant pour une "science de l'esprit" avec des idées de Christian von Ehrenfels, Car1 Smmpf et Aiexius Meinong.

Ehrenfels pensait qu'un élément mental, qu'il a nommé qualité gestalt, explique pourquoi une séquence de présentations a une certaine forme. La thèse se tourne ensuite vers un livre de Edgar Rubin, même si Rubin n'était pas un étudiant de Brentano. Ces expériences qui démontraient le phénomène %g~re-ground"~néanmoins, étaient bien connues de l'école Gestalt de Berlin. Mon analyse de Ehrenfels lève le jour sur la théorie de Rubin, une théorie que l'école getalt de Berlin semble avoir négligé, peut-ètre parce que les résultats de Rubin correspondent si bien à leur propre notions.

Stumpf a développé, partieHement en rejetant la notion de substance d7Ariaotle,sa théorie à savoir comment des présemations forment une unité . Sa théorie fonde sa critique d'aSSOciati0nniSme. Les étudiants de Shimpf cependant, n'accepteront pas sa vision double mais bénificieront des méthodes expérimentales qu'il a développées et qu'il leur a enseignées. Stumpf a adapté sa théorie en parailde au travail de ses étudiants et cette version a formé la base de son objection pour leurs explications. La théorie de Meinong qui suggère que les présentations sont unies a été inspirée par les scholastiques, et ses étudiants avaient a le rendre dans un format psychologique. C'est lors d'une révision de travaux antérieurs que l'école gestalt de Berlin a présente sa notion gestalt pour la première fois dans un format complet.

La conclusion réitère brièvement la notion de Brentano. Elle contient aussi des observations générales par rappon a la diversité parmi les notions gestalt, la grande possibilité des auteurs gestalt et leur grande ardeur vers la poursuite du savoir scientifique pur. ACKNOWLEDGMENTS

A thesis is not the result of a solitary effon, although reading, studying and writing mean one spends a lot of tirne alone. Sa most theses are the result of CO-operationand this one is no exception.

A few years ago 1 approached professor John Macnarnara. Since my student yean at the University of Leiden in the Netherlands, where 1 obtained the quivalent of my bachelor's and master's degrees, 1 have wanted to write a thesis related to the history of pqchoiogy. In that nrst conversation with we didFranz Brentano, whom we both admired. When we realized the iink between Brentano and the later gestalt theorists, my research project took form. John was full of humour, had an enomous knowledge, was inspiring, and above all, an excellent teacher. in early 1995. he was diagnosed with a brain tumour. He undenvent therapy but to no avail. The cancer had, of course, a crippling effect on him. In the spring of 1995 he went to his office and asked me to corne by. He clearly summoned his concentration and we talked for an hour about my work. We had an idea what the thesis would look like, but there was not much on paper.

My thesis has not deviated fiom the plan laid out at that moment. I was able to say goodbye to John a few days before his death and thank hirn for being such a wonderful person, but

1 still regret that he could not see the fhai produa. 1 hope it would not have disappointed hirn.

1 notice in rny conclusion that the gestalt theorists were broadly educated, pursued wide interests and were not &aid to enter new fields. That is dso true for professor James Ramsay who volunteered to take over John's task as my adviser. He staned his studies as an undergraduate in English teaching, but has made a name for himself as a statiaician. 1, however. know him as a scholar in the history of ideas who likes to place the great minds in their contemporary contexts. Besides, his enthusiasm and encouragement significantly sped up my writing. I have the fondest mernories of our weekly coffee chats. lim's approach can best be described as 'management by objectives' and he is amatingly good at it. He was able to make something as simple and everyday as a coffee break an event to work towards.

His comments and remarks were always so fodated that they charged my batteries.

1 also want to thank here Albert Bregman and Don Donderi. They have read dr&s of the thesis and 1 learned much nom their knowledge of Aristotle, perception, and modem psychological research. Not ody the McGill Department of Psychology was supportive. 1 also found wiUùig help at the Department of Gennan Studies. 1 am very grateful to professor

Iosef Schmidt, whose conscientiously revised my rendering into English of quotes from

Stumpf and Rubin. He also brought to my attention a pmof Christian Morgenstern, which

1 have ùicluded in the thesis.

1 benefiaed much fiom my discussions with two fellow students at McGill, Dean

Sharpe and David Nicolas. David carefully read my manuscript and offered many helpful questions and comments. I also appreciate Perry Fuchs for reading a draft and helping me with his suggestions.

1 like to remind my Canadian fiiends how much English is indebted to htch

(incidentdy, my native language). Nowadays, English speakers do not like to see this tradition - which has brought so much good to their language - continued. So, 1 asked houk Hoedeman to edit the 'final' draft. Don't be fooled by her Dutch name - her English is impeccable, and if there is any 'Dutch' lefi, it must have been caused by a revision made afier her corrections.

Nowadays a cornputer is an indispensable instrument and I want to thank Moms

Eichler and Mehrdad Khosraviani for never giving up in explaihg its mysteries to me. The personne1 of hteriibrary Loans was fnendly and patient and always able to obtain the most strange texts with the minimum of reference. 1 thank Chantale Bousquet for the French translation of the abstract.

During my student years in Leiden, 1 was a research assistant of Rob Janssen. Rob spurred my theoretical interests and decisi-~elyiduenced my thinking about the human rnind.

My faniily in Holland has always been very supportive and without their assistance

1 would not have kenable to midy at ~McGdl.Elianna Beckrnan gave me the tirne and space to pursue academic interests. The thesis is dedicated to Our daughters Liette, Ianelle and

Kara Lynn They already jump more energetically, higher and farther than 1 do. I hope they wilI continue. TRANSLATOR'S NOTE

This thesis dixusses the thoughts of some German speakers, with the exception of one Dane. But, to stay in the Geman world, it is a German translation and not the Danish origuial that is used here. Some of the Gemian works are available in an English translation.

While working on this thesis, 1 have always used the original Geman texts. The quotes, however, are di in English and many are taken fiom existing translations. In those cases, the page number mentioned in my text refers to that English version and one can find al1 its details in the bibliography. 1 have also included the German original there.

Other works, such as those of Car1 Stumpf and Edgar Rubin, have not been

(completely) translated into Enghsh. 1 translated those passages myself and the page numbering refers then, of course, to the German text.

The authors, with the notable exception of that talent Franz Brentano, wrote an academic Gemm, dted and convotuted. That was the prevalent way of writing at that time and 1 do not want to render it into todayJssmooth and less complex English prose.

We ail how that sornetimes a literal translation does not capture the exact meanhg or, even worse, is deadmg. in these cases 1 chose to paraphrase. For example, I translate webaas 'inner life.' 1 use 'object' for since that English word indicates better than 'thing' that one is dealing with something concrete, something with a shape or form and that can be perceiveci by the senses. niing' in English refers also to more general and abstract entities and those are not what the German authors write about in the selected passages. Presentation

Quite ofien one wili encounter the word 'presentation' in my text. It is an English translation of an important 19th-century German psychological notion and that word needs an explanation. Franz Brentano used the term Vorstellua and that is rendered into English as 'presentation' or 'idea' by his translators. Brentano (1874) believed that when one is cunscious of an object, that "object is present Ui consciousness" (p. 201). That presence is denoted by 'presentation.' In his first major book, Car1 Stumpf also used Vorstellung and its meaning cornes close to 'experience' or 'perception.' mmvor- Literally means 'space presemtation' and Stumpf s book deals with how space is present in consciousness. Christian von Ehrenfels also used Vomto indicate that something, in his case mainly a tone or series of tones, e>osts in consciousness. Note that 'presentation' is a mental phenornenon and, according to these authors, a hdamentai one. A presentation is, one could say, the root of other conscious activities. For instance, one cannot like a rose, or judge Santa Claus to be non-existent, if one does not have a presentation of a rose, or of Santa. This is always vaiid, even for incompatible concepts such as a 'round square.'

If1 am to deny the existence of a round square, then 1 must be in a position

to think thk round square, i.e. to combine the two determinators roundness

and in some way or another (Ehrenfels, 1890, p. 103).

Another anaiogy of presentation is that of a building block. Brentano and his students believed that several presentations can form a new one. So, Stumpf believed that space presentation is made up of other presentations, for instance colour and extension.

Something that is made up of several elements can be described as a cornplex, and Stumpf

(1873) used the tenn Vorstellunpo~~le~(p. 109). In Stumpfs next major work

Toripsycholoe, one does not find the word anymore; instead Stumpf used m,which I tramlate as 'content.' Where presentation pomts to the existence of an object in consciousness, content indicates how it is experienced. Therefore the change in words does not signi5 a change in underlying theory .

The same view, namely that of building blocks, was also expressed by Ehrenfels

(1890). He hold that the presentation of a melody is grounded on individual presentations of tones.

It is then indubitable that the presentation of a melody presupposes a

compIex of presentations - presupposes a sum of presentations of successive

single tones with distinct and mutuauy exclusive temporal determinations (p.

85).

And a few sentences firrther Ehrerifels wrote:

.h analogous question can clearly be raised also in regard to spatial shapes.

Indeed, because in this case all the constituent parts of the presentational

complex fomiing the basis of the perceived shape are given sùnultaneously,

matters are considerably simpler (p. 85).

In other words, a melody is a presentation which contains individual presentations of tones; a visual shape, likewise, has individual presentations (colour, texture, and so on). Ehrenfels argued that more is needed to understand a melody or shape, and his idea will be explained in his chapter.

Em~finduneor sensation, is a special case of presentation. It is evoked by a basic, elementary stimulus such as a colour or single tone. It is triggered by an outside event, which is not necessarily the case for a presentation; a 'round square' cannot be seen by one's eyes.

There is an irnpiicit assumption in the idea that presentations can be units, building up more cornplex presentations. That assumption is that the unit, the (more) elementary presentation, remains unaitered when included in a cornplex. That view is challenged by

Wolfgang Kohler ( 19 13). Kohler held that it is an empirical fact that the relation between the outside stimulus and how it is present in consciousness is not constant. There is, in this view, no constant relation between stimulus and conscious experience. He and his fiends

Max Wertheimer and Kurt Kof3ca assateci thaî the expenence of a stimulus depends on the overall structure, on the gestalt in which the individual stimulus is embedded. The Linew

We start by drawing a "-y me"of the gestalt theonsts in Figure 1. That picture lads directIy to some of the questions answered by this study, such as ( 1) What is the role of Franz Brentano in gestalt psychology?, and (2) What did Max Wertheimer, Kurt Kofka and Wolfgang Kohier leam from their teacher Car1 StumpP . . ui 1890, Christian von Ehrenfels wrote a far-sighted article: On 'Cesdt QalitieSI.

Therein, he proposed a solution to the problem of the perception of gestalten. Gestalt is a

German word meaning 'outer shape' or 'a whole at the perceptuai level.' Gestalt is thus a word that expresses a perceptuaI property of a thing. Barry Smith ( 1988) explained this tenn as follows:

'CM& is itself derived rnetaphoricaily from the Germanic (a place

to stand), which has the same root as the English 'saIl.'There is a common

use of the term to mean 'external or visible fom' (e-g. of the de* or of a

ceramic pot), as also a famiy of uses allied to English expressions such as

'cut a figure,' 'Great Figures in Styrian Philosophy,' and so on. The term

signises also quite generaiiy a structure or cornplex, so that Clausewitz, for

example, can speak of the war as an 'absolute gestalt,' an 'indivisible whole,

whose elements (the individual victories) have value only in relation to the

whole' (p. 14). FRANZ BRENTANO (1838- 19 17) Calls for a science of mental phenomena. Consciousness is a unity made up of parts.

Graz school Carl Stumpf Alexius Meinong (1848-1936) (1853- 1920) Developed an experimental Christian von Ehrenfels Object theory. method to study mental (1859- 1932) phenomena. Argued that a mental element, the 'gestalt Stephan Witasek quality,' foms a unity (1870- 19 15) (melody) out of the Vittorio Benussi individual (tone-) (1878- 1927) sensations. A gestalt is the Berlin gestalt school result of an active Max Wertheimer ( 1880- 1943) mental proçess. Kurt Koffia (1886-1941) (Critical review of the Graz school.) Wolfgang Kohler ( L 887- 1967)

Rejected the view that there exists a constant relation be tween the individual stimulus and its experience. It is the whole that determines how the parts are perceived. Isomorphic relation between gestalt, brain processes, and conscious expenence.

Figure 1 the gestalt Line Ehrenfels believed that his solution was valid for dI kind of gestalten, but he predorninantly demonstrated his notion with the example of a melody. The problem of gestalt perception is more chalfenging than it appears at first glance. Ehrenfels reminded his readers that a melody cm be played in different keys and with different instruments.

Although at every performance dirent tones reach the ear, one hears, nonetheiess, the same piece of music every time!

Ehrenfels had ben a student of Brentano and he had written his doctoral thesis under Alexius Meinong, who had studied under Brentano. Ehredels wrote his article when teachg in Viema, at a time when Brentano taught there as weU. Later, Max Wertheimer attendeci Ehrenfels' lectures, when Ehredels was professor in Prague. Wertheimer wrote his

PhD under Külpe, but before he wrote his dissertation he spent two years studying at the

Berlin Instmite of Car1 Stumpf and wbiie working on his thesis he retumed often to Berlin for research and disaission. Stumpf had been a student of Brentano too, and for that matter an early and devoted one. Wolfgang Kohler and Kurt Koftka wote their PhDs under

Stumpc and when Stumpf retired from his post in Berlin in 1922, he made sure that Kohler succeeded him. So, aü three famous gestalt psychologist s studied in Berlin under Sturnpf and that is why they are known collectively as the Berlin school.

Ehrenfeis, however, was not the only student of Meinong who would study gestalt perception. Meinong's two collaborators at the University of Graz, Stephan W~tasekand

VdoBenussi, also came up with an explanation of gestalt perception. Their theories are very similar; in fact they are translations to the psychological domain of Meinong's highiy abstract object theory. Meinong and his collaborators are also named afler their town; they are cailed the Graz school.

Thus it was the students and students of student- of Brentano who thought deeply about the gestalt idea. This raises the question whether there is a special link between

Brentano's teaching and the gestalt notions of the later generations. That question was already raised by a student and orthodox foliower of Brentano, Oskar Kraus ( 192 1) and

Kraus took a very strong position in this matter. He went so far as to deny that the later gestalt ideas contained anythmg new. They could, he claimed, already be found in the works of Brentano and his immediate students, although not always described by the word 'gestalt.'

It is worthwhile to follow Kraus' line of thought - not because it is original or balanced, since it is not, nor because Kraus is a player in Our story, which is also not the case. Kraus, no doubt unintentionaiy, gave an insight to the world in which the later gestalt theorists matured to full-fledged psychologists. Kraus agreed that there are wholes that are more than the sum of their parts. His task was to relate that notion to Brentano and his students. He started with an argument going back to Ariaotle and also employed by

Brentano and Ehrenfels. The experience of sight by one person and that of hearing by someone else form two separate realities which cannot be brought together. The sight and hearing by one penon, on the other hand, are parts of one single phenornenon. One cannot sirnply add up the two experiences; they form either a unit or they do not.

in his next step, Kraus applied StumpPs teaching on space perception. Kraus, following Stumpfs Iead, argued that the sight of colour does not add up with the sight of extension. He meant that colour always occupies some space and that there is no space without wlour. One cannot have one without the other, according to these authors. Colour and extension thus form parts of one single phenomenon. Kraus then mentioned transposability as a charactetistic of a gestalt. Remember that a melody can be played in different keys. That again shows, according to kaus, that a gestalt is about relations and not only about the absolute values of t he elements. When one changes the absolute values of the elements, their relation cm remain constant.

Kraus was now able to deny Wertheimer's claim that modem psychology - read psychology at the beginning of this century - considered a mental cornplex as a bundle, a mosaic, or an association. Kraus was right: Neither the unity of consciousness, nor space perception, nor a melody was understood in his tirne by scholars belonging to the Brentano

Ncle as a bundle of elements. Wertheimer's claim was too strongely stated, since it was not true for an important group of theorists.

Kraus' conclusio~neveriheless - that there is nothing new under the Sun, since

Brentano and hk midents said it aii - does not hold. Kraus gave a partisan account of some notions developed in Brentano's circle. Remarkably, he did not address Meinong's gestalt school. Perhaps Meinong was too much of a dissident for Kraus' liking. Anyway, if one wants to conclude that there is a relation between B and G, for iristance that G(esta1t) equals

B(rentano), or that the gestalt notion was already formulated in one way or another by

Brentano and his -dents, then it is not enough to say sornethùig about B. Naturally enough, G has to be dealt with as weii. Only then cm one draw a conclusion about the relation between G and B. In this case, as we will see, the gestalt theorists were original theorists who made unique contributions to psychology. Kraus, it seems, because of his one- sidedness, overstated his case, although he was right to trace the gestalt notion back to Brentano.

One of the reasons that the gestalt notion became so influentid is that it spread beyond the borders of the Gerrnan-speaking wodd. And the IWO world wars were instrumental in its expansion. The gestalt experimentalist of Meinong's Graz school,

Benussi, was not an Auhan citizen; he carried an Italian passport. Thus, afler World War

1 he was forced to leave Austria for Itdy, where Cesare Musatti became his principal student. Musatti, in tum, taught Fabio Meteili and Gaetano Kanizsa. Stumpfs students influenced yet another country and again a forced emigration formed its basis. Kohler,

Kofb and Wertheimer lefi Nazi Gerrnany for the U.S.A.. Ail three got teaching positions in the U.S.A., where they wrote books and articles that did not go unnoticed. Koffka . . (1935), for instance, wrote a much studied book, of wofch

I.I. Gibson ( 197 1) acknowledged the importance to his own scientific development.

The study of gestalt psychology, therefore, is not only interesthg fiom the point of view of historical tmth. Researchers such as Kanizsa and Gibson show the relevame of gestalt psychology for modem psychologicd theones. And thmakes gestalt psychology an important part of experimental psychology surviving ffom the late 19th century.

The period in which the gestalt idea took form was one of strong economic growth,

Unpressive technological innovations and fhdarnental scient Sc breakt hroughs, until Wo rld

War 1 ended that burst of creativity. But it was also a p&od of nationalism, with its accompanying prejudices. Stumpf and Wertheimer unhesitatingly employed the tem

'primitive people' for other races, and Meinong would tum out to be an anti-Semite. The

German-speaking version of nationalism manifesteci itself more positively in a high regard

for Kantian p hilosophy and (German)music. Brentano, Stumpf, Ehrenfels, and Meinong were not only well-trained and deep thinkers; al1 were ais0 accomplished musicians.

Philosophy and music formed pan of theû [education]. But they were first of al1

university men. Psyc hology was developed at 19t h-century Gerrnan-speaking universities.

What follows is some general background on German and Aunnan University life.

The argument will not be developed here. Still, 1 hope to prove two things with this section.

One is that the role of the individual researcher's personality is worth studying (and 1 must

immediately add that 1 wiil not study that roie in my thesis). The other aim is to lay the groundwork for my argument that Wertheimer, Kohler and Koma did not revolt against

their teachers.

The &man scientific world at the beguuiing of this century was a small one. The

universities were hierarchicai, with the professors wielding enormous power, unlike today's

North Ametica's universities. In the Gexman-speaking world, then, up-and-coming men with a certain ambition, such as Wertheimer, Kohler and Ko££ka, had to accommodate the

penonalities and ideas of t heir teachers, not only long before they graduated, but well into their more meyears, Znce those professors had a great say in the scientific societies and

what was published in the scientific joumals. Furthemore, those professors advised the bureaucratie authorities on appointments for university positions. Clearly, the German universities in that period did not have an atmosphere conducive to a revolt of ambitious yung men against their elders.

In the case of Car1 SmrnpS one has to realize that he was not only the principal teacher of the Berlin school, he was du, one of t he reigning psychologias in Germany . He was a very powerful man, and it was thus not only his ideas, but his personality in general that set the fiamework for the generations that he taught.

Previous work on t he history of the -tait id=

An important contribution to the history of the gestalt movement was made hy

Mitchell Ash. Ashk (1 982) doctoral thesis, The Ewceof GedTheory? Expenmentd

Psvchology in &&1890 - 1920, is a historian's account of the ongin of the Berlin gestalt school. The dissertation is not published, which is surprising, since it is, by ail accounts, an outstanding mdy. However, in 1995 he published another book on the development of the gestalt psychology in Gerrnany, and the first two parts of his Gesu

Psycholopve1 890- 1967 are a revision of his dissertation. The two other parts are an extension in time of his earlier study. Ash (1995) wrote that his

book reinforces contexnialist accounts of science by locating experimental

psychology and gestalt theory in the changing social and institutional milieux

of German academic and cultural life fiom 1890s to the 1960s (p. 5).

In the contextuaiist approach lies, 1 think, the strength as weli as the weakness of

Ash's works. But first we need to know what it means to call an account 'contextualist.' This term was. so fa as 1 cm see, coined by hh. Unfortunately, Ash did not give a definitioa but he ( 1982) wrote thar his dissertation

sets the origin of gestalt theory in a broader social and intellectual context,

and thus shows how the structure of the German universities shaped the

development of experimental psychology in that country (p. III).

Ash infonned us of the university setting in which Wertheimer and his fnends matured to experirnentd psychologist S. Contextuai in Ash's hands means, thus contemporary cont ext .

Ash showed that expenmerrtal psychology in 1890 was a specialty within philosophy and it remained so during the time that gestalt psychology took form. Experimental psychology was a subdiscipline, and as such it had a certain independence. There were specialized journals, societies that met regularly, and there were laboratories, albeit weakly nipported, al1 catering to the needs of the communtty of experimental psychologists. Anyone who wanted to make a career in this field had to make sure, according to Ash that his works combined experimentai rigour with philosophicd depth. Remember that psychology fell under the heading of phiiosophy. Psychology students had to spend time and effon, naturaily enough, mastering srperimental techniques. And obviousiy, that effort could then not be used for a more classicai philosophical training. This brought about an angry reaction by the more ciassically inciined philosophers. Ash recounted how those philosophers argued, with some justification, that experirnental psychology was unable to solve philosophical problems. And those philosophers wncluded, not unselfisMy, t hat experirnental psychologists, who were competing for the sarne positions in the philosophical faculties, should not be appointed as professors. Ash situated the rise of gestalt theory between 1890 and 1920, and his arguments seem conclusive. In 1890 Christian von Ehrenfels published the first gestalt article, and in

1920, Kohler's book on physical gestalten appeared. The gestalt notion, with its implications, was by then weil thought out. The works of the Graz school fdin that same perïod, though pre-dating those of the Berlin school by about 10 years. The first article in which a member of the Berlin school clearly articulated his own position was w&en by

Kof£ka in 1915, and it gave a detailed exposition and a well-thought-out criticism of the

Graz school. The Berlin school developed, or at least presented itself, partly in contrast to the Graz school. This warraats a discussion of the Graz school. Ash mentioned the Graz school and told us sornething about their ideas, but did not go into detail.

Ash placed the genesis of the gestalt notion in its contemporary context. Ash's position is not only theoretically very defensible; the results of his study prove its validity.

Nevertheless, I feel that his approach needs to be complemented by a study of non- contemporary influences. 1 have two reasons for this addition, and both foliow fiom Ash's own research. Ash ( 1982) beiieved that

In the rn* however, they [the gestalt theorists] brought about their

"scientific revolution" not by irnporting new perspectives fiom elsewhere,

but by radically transforrning and reconstructing already available concepts

and methods (p. XXN).

Ifgestalt psychology is a transformation of Gennan psychology, then one must have some idea what is being transformed. The conclusion that a transformation had taken place can only be based on a stuày of the men who proposed the onguial concepts and methods, and one must also study how their teachings were transmitted to later generations. In other words, Ash's conclusion calls for a study of the iineage of the gestalt notion.

Secondly, to have an eye ody for conternporary events - they are relevant, of course

- misses an imponant aspect of scholarly life, namely the study of the great minds of the past. That is especially true for philosophically trained minds. Then again, the question of the influence of eariier scholars on the gestalt generation needs to be addressed as well. And that is what this dissertation sets out to do. It traces the lineage of the gestalt idea, and it supports Ash's conclusion with independent research.

The obvious person with whom to start a study of the lineage of the gestalt school is Car1 Stumpf, the principal teacher of the Berlin gestalt school. One cannot, however, appreciate Stumpf without a firm grasp of Franz Brentano. Ash did not closely examine

Brentano, but said only a few words in passing. Brentano, however, is a fascinating figure in his own right and he deserves a more subaantial treatment in a history of the gestalt school. He was. after ail, the foremost teacher and lifelong inspiration of Stumpf We will see that the study of Brentano wiil bring Stumpf into a better perspective and also highlight

Stumpf s contribution to psychology, namely an experimental method to study the imrnediate given.

Another important contribution to our understanding of the gestalt idea is The

Fou- of Gestalt Theory, edited by Barry Smith (1988). This book contains the

English translation of Ehrenfels' article and several other philosophically-oriented essays.

Two of those papers, one by Barry Smith, wtT-An Pinophy; the other one by Kevin Mulligan and Barry Smith, -1s- J3e Fo-ns of Gestalt Theow, are of particular interest to us. Smith wrote that his

essay is intended also, however, as a first rough historical survey of the

wider gestalt position - dbeit fiom a somewhat specialized philosophical

point of view (p. 13).

He then added in a note to this passage that his

paper cm therefore be seen as a philosophical supplement to Ash 1982, an

excellent historicai study of the origin of gestalt psychology in Germany (p.

72).

If Smith's paper is a rough meyand philosophical supplement, then this dissertation spells out in more detail the historical lineage of the gestalt notion and is a psychological complement to the excellent work already done. Smith's contribution, being more philosophicai, for instance, does not explain the method by which experimental psycholo@sts of the beginning of this century studied the expenentiaily given.

In FoUndiltipns of Gestalt Theory the influence of Brentano on his students is recognized. There is, however, no attempt to relate Brentano's notions to the gestalt idea.

Smith ( 1988) rnentioned some arguments frorn Stumpf s "posthurnous philosophical masterpiece on the theory of cognition" (p. 71). He acknowledged, however, that Sturnpf s last book "remains unread" (p. 7 1).

1 agree with Smith that it is worthwhile to study Stumpfs last book. Stumpf s influence on his students cannot be distilled, naturdy enough, fiom a book that appeared many years after his students had already formulated their programs. For that, Stumpfs earlier work should be studied, nohuithstanding that one finds many of Stumpfs earlier thoughts restated in his last opus. Indeed, Smith ( 1988) mentioned Stumpf s first book and said that in it, Stumpfbroke with atomistic psychological theorking. That thesis. clearly of relevance to the gestalt thought, deserves to be studied in much more detail and supporteci by teaual anaiysis.

Another Uisight of Smith needs, 1 beiieve, to be spelled out in more detail roo. Smith

(1 988) related Ehrenfels' essay with Aristotelean notions, or:

with the classical division between matter and form. The matter of

experience is conceiveci as king constituted by the data given in supposedly

simple sensory acts, al1 of which are discrete and independent, i.e. are such

that each can exist in principle in isolation âom all others. The fOnn of the

experience is conceived as being constituted by special categorial objects

given in non-sensory intellectual acts (p. 23).

This thesis complements Smith's ( 1988) essays in the sense that it gives flesh and bones to

Smith's observation. . . David Murray's (1 995) malt Psv&oloW- ve Revolutioa is a psychologist's contribution to our subject. Murray (1995) did not address the precursors of the gestalt psychology, although he aressed in his textbook of 1988 the Iine leading from

Ehrenfels and Meinong to Benussi's experirnental work. Murray (1995) argueci that gestalt psychology anticipates cognitive psychology. He believed that :

The overwheimuip impetus in late gestalt psychology was to un@ cognitive

psychology by stressing the m-random, nokchance operations ofthe brain,

a brain in which contextual relationships in the environment were mapped by isomorphic relationships between events in the neuron substance,

rnappings which in turn were the result of evolutionary processes that have

taken place over millennia (p. 52).

One can deduce from the above quotation that the gestalt psychologists took a monistic stance (a coilapse of the mental and the physical), to be distinguished from their teacher's dualism (a principle separation between the mental and the physical). Arguably, today's cognitive psychologists have a monistic point of view too. It therefore seems that in its change to a monistic perspective, while dowing mentaiist concepts such as expectation, decision and so 0% without losing sight of psychological phenornena, or without becorning biology, gestalt psychology anticipated cognitive psychology. Now, we have reached a better understanding of the transformation that Ash signalled.

This section wili end with an oveMew of the thesis. But before that, 1 would like to state the problem of gestalt perception as clearly as possile and also indicate some solutions t hat are theoretidy possible.

1s a melody out there, or, on the other hand, are only the tones real and do we then produce the melody in ou.heads? Does a gestalt exia on its own, or is it only the element s that reach one's senses? These two positions seem to demarcate the opposing sides in the debate on the perception of gestalten. And one finds psychologists of the Brentano circle on both sides. Ehrenfels considered a gestalt quality to be an existing entity, and the Berlin school held that a gestalt exists as well. Meinong, Ehrenfels' teacher, and Meinong's foilowers Witasek and Benussi on the other hand, argued that a gestalt is produced in the observer fiom the more elementary sense impressions. Meinong would interpret Ehrenfels' paper in a merconsistent wit h his own ideas and t hereby deny t hat Ehrenfels advocated a position that properly belongs to the other camp. Anyway, Brentano's work would lead to two opposhg views on the perception of wholes. That controversy would become clear in an article of Kof]Ekalshand directed againn Benussi.

lf one beiieves that a gestalt is out there, then it is not unreasonable to assume that one perceives a gestalt fkst and abstracts the elements kom the gestalt perception. ui this view, the gestalt is in a sense the bearer of the elements and the elements are subordinate to their gestalt. Perception is then a process in which the gestalt, somehow, rnirrors itself in consciousness. An other view sees perception as an active process in which one perceives first the elemerrts and then produces the gestalt from the sensations caused by the elements.

Note that these two sides agree that a gestalt is more than its elements. In the first view the elements are subordinate to the gestalt. But dso in the latter view a gestalt is more than a simple addition of the elements. Now it is the result of a production process.

We saw that Brentano occupies an important position Ui the gestalt lineage.

Therefore we will îkst disaiss some of bis thoughts. Since Brentano was a great interpreter of Anstotle, we will place Brentano's psychology in an Anstotelean context. Thereafter we will look at the first real paper on gestalt psychology, that of Ehrenfels. Then we will step briefly out of Brentano's circle to discuss an influentid book by the Danish psychologist Rubin. With Stumpf we are squarely back in Brentano's tradition. Then we will say some words on Wertheimer, Kohler and KoBa. We saw that al1 three studied in Berlin, at the same institute and under the same professor. But they did not study there together. Their friendship and collaboration started only in Frankfùrt around 1910, where they did experimental work at the newly founded Academy of Social and Commercial Sciences in that city (see Murray, 1995). After that we go to the competing gestalt school, t hat of Graz, where Meinong Witasek and Benussi worked. Then it is time to discuss Koffka's ( 19 15) il paper that highhghts the differences between the two schools, &ter which we will end with some conclusions.

Gemthinken like Kant, Herbart, Goethe, and G.E Müller (1 850-1934) afFected alço the thoughts of the gestaltists. Wertheimer studied also under Külpe, Ko&a wrote his

Habiiitationschrifi (a second thesis, leading to a postdoctoral degree) under Messer. while

Friedrich Schumann sponsored the Habilitationschriften of Kohler and Wertheimer. 1 will not. however, explore these influences. but restrict rnyself to the influence of Brentano and his direct students. FRANZ BRENTANO ( 1838- 19 17)

Brentano is an imposing figure, even in the 19th-century German-speaking world.

He is irnpressive not oniy in our eyes, but in those of his students. Stumpf saw Brentano for the first time when Brentano defended his Habiiitation thesis on Aristotle at the Catholic

University of Würzburg. Stumpf was then a tint-year student and Brentano a junior member of the scientific staff. Brentano defended his thesis with zest, and the erudition he showed on that occasion made a lasting impression on Stumpf If we may believe Stumpf - and why not? - Brentano outshone al1 the senior academic staff. It was then that Sturnpf decided to attend Brentano's lectures (see Stumpf, in Kraus, 1919, p. 85), which resulted in Stumpf becoming Brentano's first student .

Brentano devoted his life to his psychological and philosophical concems. He had a gifi for words and he comrnunicated his views very well. As a consequence. his lectures were always packed (see, for instance, Kraus, 1 9 19. p. 10) and his books are still a pleasure to rad. He also devoted much of his time and energy to the intellectuai developrnent of his students. What better mentor could one wish for? Perhaps one who is less overpowering.

Ehrenfèls (quoted in Fabian 1986, p. 5) payed due homage to Brentano's productive mind.

But Ehrdels remembered also how Unpossible it was to convince Brentano of anything that opposed his views, even when it concemed matters not related to Brentano's core philosophical and psychological concems. Brentano had a combative nature and was a formidable opponent. Students were, of course, no match for him. So his students were placed in a situation in which they had to ask themselves why put this great man on a wrong footing, why be laughed at? And the only sensible nrategy for them was to gratefùliy accept the many intellectual gifts Brentano had to bestow, but not to discuss controversial points of view.

We saw that Stumpf was a student of Brentano in the early days when Brentano taught the history of philosophy in Würzberg. At that time Brentano paid considerable attention to the Greek philosop hers (se Stumpf in Oskar Kraus, 1 9 19), especially Aristotle.

Indeed, Brentano was one of the great interpreters of Aristotle and he brought important

.Ariaotelean doctrines into psychology (see Boring, 1929). Athough he wrote several more books on Aristotle bis most important for psychology is Psycholow from an Empirical Point of View, which appeared in German in 1874. Franz Brentano received a thorough classical training when he was a priest. He wrote his masterpiece, however, in a couple of months, immediately afler he 6.naI.i~broke with the Cathoiic Church, left the priesthood and resigned from the University of Würzburg in 1873. Brentano considered Aristotle a genius whose thoughts could still advance research (se, for instance, his ( 1882) Aristoteles Lehre vom

Lksprune des rnenscliches Geistes). It is only natural that he tumed to Aristotle when writing his book.

To appreciate bener Brentano's interest in AristotIe, one has to situate it in a broader framework. The rationalist spirit of the 18th-century enlightenment lead to an objective research on fundamental texts, a research that was independent of, and sometimes even antagonistic to, the estabtished opinions of the day. An area where those studies had a deep effm was the interpretation of the Bible. In 19th-century Germany, the Bible was studied as any other text, with al1 the literary and historical methods Bible-study could master without cari- much for theological sensitivities. Brentano's study of Aristotle must be seen in this li@. He studied the original texts to detect what the great man really said.

Brentano taught courses in Greek thought. therefore his students knew Aristotle well. That, of course. does not imply that they accepted al1 notions of .kistotle; in fact they did not. Their midy of fisotle, however, influenceci their thoughts on the human mind and therefore their psychoiogy. That makes it advisable to devote some attention to the

Anstotelean roots of Brentano's thoughts.

Aristotelean Roots

The postdate that an object's fom could be a cause or explanation for cenain phenornena is very old. It is implicit in the answer of parents to such pualinp children's questions as. why do birds have feathers and not hair? Parents simply respond that is bird's nature - meaning that is how they are organized. The metaphor is, interestingly. based on the object's shape. shape or form standing for organization. Notice that the answer appeals not to a factor that produces a change or movement, but to a system that requires certain components, a syaem that is itself an interacting combination of those components. It is this insight that Aristode sought to capture in his theory of formal cause which he developed in co~ectionwith three other causes in his Phvsica. That book opens as follows:

When the objects of an inquiry, in any department, have pnnciples.

conditions, or elements, it is through acquaintance with these that

knowledge is attained. For we do not think that we know a thing until we are acquainted with its primary conditions or first principles, and have

camed out our analysis as far as its simplest elements ( 184 a9- 15).

Then in Physig, book II, chapter 3, Aristotle wrote: "Men do not think they know a thing till they have grasped the 'why' of it (which is to grasp its primary cause) ( 1 94 b 19)."

Then he described the causes, four in total.

The first or material cause refers to matter; it is: "that irnrnediate constituent of it which taken by itself is without arrangement" ( 193 a 10). Matter is the "materiai substratum of things" (193 a29). Aristotelean matter should, of course, not be understood with our modem rnind. It is a metaphysid primitive, used basically to handle the individuation of that which has form. This wooden bed is different From that one, because their matter is different. Although they can share the same form, they are still different beds. The next fùnction of matter is to deal with change. Matter stands for potentiality, allowing things to be altered. A wooden bed can be painted or bumt, not because it has a certain fom but because it is made of wood. Tne notion of matter has met considerable opposition, including

Brentano's. This is, however, not the place here to discuss that cnticism.

The second or formal cause is the most important one for our purposes. It gves a structural expianation of phenornena and is therefore a principal of unity. If matter is the potentiality of things, fom is their actuality.

The form indeed is 'nature' rather than matter, for a thing is more properly

said to be what it is when it has attained to fiilfilment than when it exists

potentidly ( 193 b8). Owens ( 1963 ) explained the formal cause as follows:

The form is the findamental Being of the thing, and that whatever else may

be in the thing denves its being fiom the form (p. 185/ 186).

So "the essence of natural organism is their fow not their matter" (Irwin, 1990, p. 235).

This does not mean that a naturai organism can be without matter. as for instance a mathematical circle. ln InMn's words: "natural organisms are essent id1y compounds of fom and matter, rather than pure immediate foms" (p. 246). That means that one has to take into account both matter and fom to explain a natural organism. A matenal explanation on its own will not do, since then one leaves out form as cause of that organisrn. On the other hanci, pointing oniy to the forma1 cause is not enough either; then one overlooks the causal role of matter. Aristotle differentiated between several causes. To explain rneans, then. to describe the relevant ones.

Note that whenever one separates form from matter. one can talk about objects which are not materially realized. We will see that Meinong developed that insight fùrther.

He pointed out that we can talk about the Golden Mountain, about a riot which did not take place, or even about a round square. Since we cannot perceive those "objects," we must have psychological mechanisms in order to deal with them.

The third or efficient cause explains, as Owens ( 1963) puts it : "sensible changet' (p.

193). This cause best fits our everyday understanding of cause and effect. Aristotle referred to it as:

the primary source of change or coming to rest; e.g. the man who gave

advice is a cause, the father is cause of the child, and generally what makes of what is made and what causes chanse of what is changed ( 194 b29-3 1 ).

The last cause is teleological. Aristotle claims there that the notion of purpose appiies not just in hurnan action, but generally to change. In fact he saw end or purpose as logically coming prior to function. In Ph* book II, chapter 8, .4ristotle proposed this view against, we would say, his pre-Darwinian opponents and concluded that there is a cause "that operates for a purpose" ( 199 b32). Here particularly it is important to interpret cause as meaning explmation.

By the way, we find a similar kind of insight in the Bible. The Gospel according to

John opens with: "in the beginning was the wu.There is, in John's ide* a u,a plan, a blueprint. or, as it is most often translated, a Word operating in nature.

We find in Aristotle not only a differentiation between 'fom' and 'matter'. but aiso one between 'substance' and 'accidents.' Accidents are things such as colour, shape, sound, and weight, properties of an object which cm Vary without the object losing its identity. A fnend of ours can cut his hair and even give it another colour. He can also stan exercising vigorously, leading to a strengthening of his muscles. A strict diet will ensure that he loses

çome weight. Notwithstanding aii these changes in his outer appearance, the identity of our fiend remains, of course, the same. It is important to note that both change as well as continuity are present here. Aristotle used the word 'accident' to indicate that which may be changed without affecthg the identity - the length and colour of his hair, muscle tone and weight, and so forth.

Another example is that of a dog. During its Mietirne, Freddie the French poodle will change in shape from a puppy to an old dog; his bark will alter: and the colour of his hair can become greyer as well. Nevenheless, the dog remains Freddie. We have to recognize two aspects in Freddie's life, that of change and that of the underlying constancy. Anstotle used 'substance' for that which continues; in our two examples: the identity of Freddie the

French poodle and that of our fiend who became such an health enthusiast Substance handles the identity of al1 the changing accidents; it individuates them and makes them belong to Freddie, or to our friend. In the Aristotelean tradition, the accidents inhere in the substance and are individ~atedby it. The substance is thus the substratum, the bearer. of the accidents. The accidents can change. ïhey will belong to the same object. however, as long as the substance stays the same. It rnakes the grey hair and sofl bark belong to the old

Freddie, and the muscle tone and short hair to our revitalized fiend.

The notion of substance, of course, was not left unchallenged. Brentano's view is best described in Kateeorienlehre (1933), which appeared years afler his death. Stumpf rejected the notion of substance altogether. Ehrenfels, however, proposed a solution to the gestalt problem which cm be seen as an innovation in the Aristotelian tradition.

Substance is related to fom but it cannot be equated with form. Two exarnples make that clear. A caterpuar metamorphoses into a butterfly. That animal thus has two strikingly different fomis in its Lifetirne. SûU, the caterpillar and the butterfly are the Mme animal The second example is taken from a children's story. A witch tums a prince into a frog and. oh the wonders of what the kiss of a lovely girl can do! It can tum the frog back into its princely giory. Even very young children have no problem realizing that the prince and the frog share the sarne identity, although their foms are totally different. Now we have to tum to another differentiation made by Aistoile and developed in

De Anima. Our perceptual systems register colours, shapes, sounds and so forth. They perceive the accidents, but not the substance. That irnplies that we have to infer the underlying principle of identity fiom the perceptual array.

Perception, accordiig to Aristotle. depends on a process of movement in a medium originating frorn outside the organism. ïhat movement, in its tum, sets the perceptual organ into motion. For instance, light rays fiom a coloured object reach the eyes through the medium of air. (Now, of course, we know that air is not the medium of transmission of light in the way that it is for sound. One can see in a vacuum but not hear. )

By percephiai organ, Aristotle meant "what has the power of receiving into itself the sensible Foms of things without rnatter" (424 a 17). In other words, Anstotle thought that the percephial system assumes the quality of that which is perceived. Thus when perceiving red Our eyes becorne rd. That rneans that the matter of which the sense organ is made must be capable of change. And it implies that the matter is structured in a certain way, or has such a form that the organ is indeed sensitive to stimulation fiom the outside. It is thus matter and form together that explain the functioning of an organ.

Each perceptual organ has its unique proper object: seeing for the eye, hearing for the ear, and so fonh. If different systems each perform a different fùnction, then the need aises to combine each one's specialty. Somewhere, sight and hearing have to be integrated, or in Aristotle's exampfe of seeing and tasting:

Since we also discriminate white from sweet, and indeed each sensible

quality from each other, with what do we perceive that they are different? (426 b 10)

There has to be a single agency that detects the sweetness as weil as the whiteness of sugar and also notes that whiteness and sweetness differ tiorn one another. Aristotle located this comecting agent in a second-level perceptual system, narnely in the 'sense that is common.'

Every perceptual organ sends its information to the sense that is common. where they get integrated. ïhe sense that is cornmon also registers where the incorning information cornes from. The sense that is cornmon notices thus whether one is 'seeing' or 'hearing.' Awareness of our seeing and hearing arises, according to Aristotle, at a late stage when the incoming information sets combined.

In the Aristotelean psychology the accidents are detected by the senses. whereas the substance is not directly given in the stimuli. We see a caterpillar and a bunerfly, but not their underlying pnnciple of identity. We see a dog and hear a bark, but nothing in the perceptual stimuli tells us that they belong together. One has to infer that sight and sound are related. for instance on the basis that they are correlated in time and space. But the visual stimuli themselves do not contain markers telling us which sounds belong to them, and neit her do the auditory stimuli give information on their corresponding sight . If one cannot directly perceive substance, then substance has to be derived From the stimuli. The perceptuai system, according to Aristotle, carmot do that since it only registers the stimuli and integrates them (in the sense that is cornmon). Then, in order to arrive at substance, there has to be another system workhg on the percephial data, and that is the cognitive one.

If there are two different systems, then the question aises, how are they comected with each othef? That task is perfomed by an agency, which in the Aristotelian tradition has been narned the active intellect. The active intellect takes the input frorn the senses and cornes up with their fom, or their structure. That is, how the parts are related to one another. Thus perception represents the particulars - these hairs, this bark - cognition provides us with the more abstract structure, or plan of an object: Freddie. Perception bnngs us into contact with the constantly changing stimuli, while cognition brings us in touch with that which remains constant.

How can those two structures be unambigiously differentiated from each other?

Ariaotle proposed the critenon of truth conditions to distinguish cognition From perception.

The thinking then of the simple objects of thought is found in those cases

where falsehood is impossible: where the alternative of tme or false applies,

there we always find a putting together of objects of thought in quasi-

unity.. . . For falsehood always involves a synthesis; for even if you assert

"that what is white" you have included not-white in a synthesis... . that which

unifies is mind (430 a26-b5).

The criterion of thcondition is not applicable to perception, according to Aristotle. One's sense organs will pick up their stimuli accurately since they are made to do so. What one rnakes of that inforrnatioq how one interprets it - that it is Jean or Pierre approaching - there one can en. It is at the level of cognition that the criterion of truth applies.

WendeU Garner did not refer to Aristotle, but Aristotle clearly anticipated Garner's

( 1974) formulation.

What then is the role of the perceMng organism? It is to add various rational

processes, so that the organism cm corne to know the reai world and to process information about it effectively (p. 1 8 1 ).

Garner differentiated a perceptual system from a rational one. The task of the rational sysrem is "to select structure from the available in the real woridt' (p. 182). In an earlier book Garner. (1962) had already stated that "the search for structure is inherent in behaviour" :p. 339). .And in 1974 he noted that "the subjects in our experiments act as though they expect the world to be structured, and that their task is to seek or select structure" (p. 183). Garner ( 1974) believed that

The world is real, and this reality includes things such as structure. But the

organism dx, interacts with r his reality to seek and select structure. Rarely,

however, does the organism create structure. There is no need. Structure is

everywhere to be found, and the infornation-processing organism need only

to look, find, and select (p. 186).

Garner believed that we are imately endowed to detect the structure of the world. a property which Anstotle tned to explain by what has become known as the active intellect.

Unity of Consciousness

We will now look at how Aristotle formed Brentano's thinking. Brentano ( l874), folIowing Aristotle, remarked, "We do compare colours which we see with sounds which we hear" (p. 159). Then he asked himself, "How would this presentation of their difference be possible if the presentations of colour and sound belonged to a different reality?" (p.

159). A little tùrther on he wrote: In fact. it would be like saying that, ofcourse. neither a blind man nor a deaf

man could compare colours with sounds. but if one sees and the other hem,

the two together can recognize the relationship. And why does this seem so

absurd? Because the cognition which compares them is a real objective

unity, but when we combine the acts of the blind man and the deaf man. we

always get a mere collective and never a unitary real thing (p. 159).

Only if sound and colour are presented jointly, in one and the same reality.

is it conceivable that they cm be compared with one another (p. 159).

Brentano was silent about the physical world; he limited hirnself to the analysis of the mental. He believed that mental phenomena are unified, that they form a whole This means that the mental phenomena present at a certain moment are grouped together and experienced as a whole. An example can make this clear. Someone is at a concert enjoying the music, while sitting in a most uncornfortable chair. The music and the discornfort are deîected by different percephial systems. Each percept on its own has an effect in the mental realm. Sdl, the psychological experience of that event is a unitary one. That experience has parts, of course, but the experience as such has a unified character. That is a central point in Brentano's psychology and he formulated it as follows:

Our investigation leads to the following conclusion: the totality of our

mental life, as complex as it may be, always forrns a real unity. This is the

well known faci of the unity of consciousness which is generally regarded as one of the most important tenets of psvchology (p. 163).

The unity of consciousness as we know with evidence through imer

perception consists in the fact that aü mental phenomena which occur within

us simultaneously, such as seeing and hearing, thinking, judging and

reasoning, loving and hating, desinng and shunning, etc., no matter how

different they may be, al1 belong to one unitary reality only if they are

inwardly perceived as existing together. They constitute phenomenal parts

of a mental phenornenon the elemenrs of which are neither distinct things

nor parts of distinct things but belong to a real unity. This is the necessary

condition for the unity of consciousness and no funher conditions are

required (p. l63/ 164).

Stumpf would understand Brentano's unity of consciousness as follows:

I now experience hunger and ionging for a rneal: therein lie sensations, ideas,

feelings and perhaps still something more that psychology singles out. But

it is a spontaneous experience, it becomes divided only through reflection

(Theory of knowledgg, p. 25).

And Stumpf taught this notion to his students. Koffka and Herbert Langfeld attended

Stumpf s winter semester of 1906- 1907 and Langfeld published his impressions of that course in 1937. Langfeld wote that Stumpf opposed the atornistic theory and that Stumpf held that: The unity of consciousness is not a sum of parts but a totaiity, the pans of

which are recognized only through abstraction (p. 55).

So. the inputs of several perceptuai systems are cornbined into a unity. This unity is constrained by its inputs, but it also leads to constraints on the perceptual systems. .An elaboration of Stumpfs example can make this clear. When we are hung-. an apple is perceived dserently than when we have just finished our meal. The unity of consciousness, to use Brentano's term, will influence our perception of an apple. The other way around is also true and in a much more self-evident way. A perceived apple, or any perceived object for that matter. influences the unity of consciousness.

One finds in these examples only reference to cross-modality integration, a combining of information anivinghg 6om Merent senses, such as sight and taste. It seems that Brentano did not consider the possibility of within modaiity integration, as when tones fonn a tune. Each tone has an effea in the psychological reah, but a melody is not the mere sununation of those effects. It was Ehrenfels, a student of Brentano and of Meinong, who wrote the seminal paper on within modaiity integration.

Brentano anaiyzed the notion of unity of consciousness funher. He differentiated hemtwo kinds, namely a collective and a red unity. Then he asked himself under which kind consciousness feu.

We cari, of course, group together a multipticity of objects and cal1 their surn

by one name, as when we say 'herd' or the 'plant world.' The objects thus

grouped, however, are not thereby one thing, but what we might cal1 a collective (p. 156).

A goup of cows grazing together is just a herd. They do not form a real unity and neither does a tavonomy of plants. Then he asked:

In the case of more complex mental States do we have to assume a collective

of things, or, does the totality of mental phenomena, in the most complex

states just as in the simplest, form one thing in which we cm distinguish

divisives as parts? (p. 157)

We saw that he opted for the last type. Remember the example of the concert, where one has a unitary experience of enjoying the music while sitting in that uncomfortabie chair.

Intentionality

Brentano is best known for his insistence on the importance for psychology of the notion of intentionality. Brentano traced this back to the Scholastics, who were, like

Brentano, inspired by .i\nstotle. Intentionality means that mental phenomena are about something, that they refer to something. In the case of perception it means that perception is perception of an object, that there is an 'aboutness' in our perception. When we see, we see sornething, say a flower. James I. Gibson's ( 1979) description of perception comes, as he acknowledged, close to Brentano's idea:

Perceiving is an achievement of the individual, not an appearance in the

theatre of his consciousness. It is a keeping-in-touch with the worid, an

experiencing of things rather than having of experiences. It involves awareness-of instead of just awareness. Ir may be awareness of something

in the environment or something in the observer or both at once, but there

is no content of awareness independent of that which one is aware (p. 239).

Intentionality is aich an important notion because al1 psychological phenomena and ody those, have it . Ni the phenomena of the minci, accordiing to Brentano, are characterized by an intentional relation with something as an object. Brentano pointed out that heanng does not occw without something to be heard, nor seeing without something to be seen, nor believing without something to be believed in, nor hoping without something to be hoped for, and so fonh. That object can be mental or physical; it cm even be an illusion or fiction.

One can remember yesterdafs anger, imagine Pegasus, and even believe that one will be nch or successfùl one day.

Intentionality distinguishes the mental ffom the physical. If al1 mental p henomena have intentionality and physicd phenomena lack it, then there ercist two kinds of phenornena, each with its own characteristic.

Every mental phenomenon is charactenzed by what the Scholastics of the

Middle Ages called the intentional (or mentai) inexistence of an object... .

This intentional inexistence is characteristic exclusively of mental

phenornena. No physicai phenomenon exhibits anything like it . We can,

therefore, define mental phenomena by saying that they are those phenomena

which contain an object within themselves (p.88/89). There are. according to Brentano. three types of intentional relations: those of presentation; judgement; and emotion, interes, or love. Thus intentionality is not exclusively intelIectual. This insight that non-intellectual functions play a causal role in humans' actions can also be traced back to Aristotle.

We observe that the possessor of medical knowledge is not necessarily

healing, which shows that something else is required to produce action in

accordance with knowledge, the knowledge aione is not the cause (&

Anima 433 al -6).

The intellecnial facuhy is not suficient to cause an organism to move. Appetite and desire. according to Aristotle, are also needed. But then he continued:

Those who successfully resist temptation have appetite and desire and yet

foliow mind and refuse to enact that for which they have appetite (&

Anima 433 a7-8).

Aristotle recognked that knowledge, judgement and desire are al1 involved in causing an organism to act or not to act. Since there is only one action, al1 the determinants must be united somewhere into one single decision for one single act.

Brentano was also in accord with Aristotle's notion of truth conditions as a fundamental characteristic of the mental, for truth presupposes reference. Acceptance as true or rejection as false are about something; they refer.

By 'judgement' we mean, in accordance with comrnon philosophicd usage.

acceptance (as true) or rejection (as fdse) (p. 198). Brentano also agreed with Aristotle that reference cm imply positive or negative feelings towards an object, giving him the third class, that of emotion. As in Aristotle's anaiysis, the intelled and non-inteiiectd, or emotional, are combined into one action (or non-action). Brentano's notion of unity of consciousness is thus not only cross-modal; it includes intellectual as well as non-intellectual fùnctions,

What did Brentano mean by objects that we perceive? fis answer cannot be understood as our cornmon sense notion of objects. According to Brentano (1874), perceived objects are mental objects; they have mental inexistence, they are present in the mental realm. What Brentano meant Ath inexistence or presentation will become clear soon. Let us first look at his definition of mental phenornena.

We then defined mental phenomena as presentations or as phenomena which are

based upon presentation, al1 the other phenomena being physical (p. 97).

What did Brentano mean by presentation? It should not be mistaken for the concept of gestalt. A presentation, according to Brentano, could be the colour red, a tone, warmth, and so forth. Brentano was not specific about presentations. It is, however, clear that he thought of them as simple, basic perceptual properties, an idea which we find also in

Aristotle. Brentano thought that a conscious state was a unity, made up of those elements.

Brentano also accepteci more complex structures. They are, in his eyes, derived from those elements (such as colour and sound). The elements are thus the building blocks of consciousness. This could explain why Brentano had syrnpathy for John Stuart Mill's opinion that the task of psychology was to establish the logic, or the laws for human thinking. Just as a chemin looks for the basic elements and how t hey are related. so the psychologist tries to detect the basic mental elements and the laws that relate them. Note that the laws of the mental realm are fundamentally difFerent, according to Brentano, from the ones that are valid in the physicai world. In the latter one finds laws of causality, while the mental has logical relations. David Lindenfeld (1980). for instacce. noted that in Brentano's philosophy

"judgements and feelings prenippose ideas, i. e. are impossible without t hem" (p. 52). There is not a causal link From ideas to judgements and feelings. An idea will not automatically result in one particular judgement or feeling. Instead, the relation is logical.

We saw that Brentano thought of mental phenomena as presentations or as based upon presentations. Thus one finds in Brentano the notion of basic presentations, the ones which form the foundation, and higher order ones, those which rest on the basic ones. Here is a iikely source for Ehrenfels' higher order gestalten and Meinong's objectives. They saw a higher order gestalt made up of more elementary units, but more about that later.

Brentano ( 1 874) noted that one cannot have reference without presentation.

Presentation is the necessary condition for the other, non-intellectuai types of reference. "It is impossible for conscious activity to refer in my way to something which is not presented"

(p. 198). Or in Aristotle's words: "When the mind is actively aware of anything it is necessarily aware of it dong an image" (De Anima 432 a7). There should be an image before the other references can occur. In an earlier quote, Brentano used the word inexistence to express that notion. He got that term fiom the Scholastics. Inexistence points out that there can be no intentionality without some presentation in the mind. Something must 'exist in,' namely in the mind. In one of the last quotes. Brentano even equated mental phenomena with presentations. We will see later that Brentano's students Stumpf and Meinong wrote extensiveiy about presentations. Their work is rooted in Brentano's ideas on intentionality, although the stress would be different. In Brentano's work the aspect of intentionality, the gasping, the aboutness is placed in the foreground, while his students put the emphasis on presentation, the prerequisite for intentiodty. Brentano's teaching of reference to an object led to an exploration among his followers of what the ground of intentionality must be.

Brentano called the study of the mentai phenomena descriptive psychology. This science was to airn at the generai understanding of human consciousness. discover its basic elements and reveal how these elements are unified. Hence, Brentano thought descriptive psychology to be scientific and its objective to be the sarne as that of every other science: that is. the discovery of the basic elements and the laws governing their combinations. Its method was to be determined by its subject, and Brentano advocated intuition as the main source of knowledge (see Macnamara, 1993). This method should not be confused with introspection, applied by Oswald Külpe and Edward Bradford Titchener, especially since

Brentano argued against introspection. We will first look at why he proposed the method of intuition for psychology. Then we will discuss his warning against introspection.

Brentano taught that mental and physical phenomena ciiffer not only in intentionaiity, but that they are also accessed dserently. Mental phenomena are experienced immediately in the consciousness of the observer Physical ones. on the other hand. are 'mediate' to the observer: there is no direct access to them. Therefore physical phenomena are a form of mental phenomena.

hother characteristic which al1 phenomena have in comrnon is the fact that

they are only perceiveci in inner consciousness. while in the case of physical

phenomena oniy extemal perception is possible (Psycholow from an

Em~iricalPoint of View, p. 9 1 ).

Since we experience mental phenomena immediately, or as unrnediated, we have sure knowledge of them and, as a result, we know those objects to be really existent.

Physical phenomena, however, exist only For us as they appear in consciousness.

[Mental phenomena] are those phenomena which alone possess real

existence as well as intentional existence. Knowledge, joy and desire really

exist. Colour, sound and wamth have only a phenomenal and intentional

existence (Psvcholow fiom an Empiricai Point of View, p. 92).

We know from Langfeld's account of Stumpfs course that Stumpf explained this notion to his students. Stumpf, according to Langfeld (1937), stated that:

The great advantage of psychology over the other sciences lies in the fact

that the objects of psychology are a matter of immediate knowledge, while

the innermost characteristics of the occurrences of nature remain unknown

(P 34). Langfeld also formulated Stumpf s teaching as follows:

The natural sciences start fiom the phenomena as does psychology. but the

former are not concerned with the phenornena as such. The natural sciences

ieave the phenomena and arrive at forces and objects, which are presumed

to cause the phenomena, while the psychologist is concemed with the

fùnctions as the material of his science (p. 34).

Stumpfs teaching is thus clearly based on Brentano's discussion of intentionality. For

Brentano, objects have a mental inexistence. So for hirn we do not have a direct, unrnediated access to. for instance, a flower out there; we only know that flower as a mental object, or a mental phenornenon.

Kohler ( 1947) would formulate this understanding as follows:

Since the world of direct expenence is the first I knew, and since al1 1 now

know about the physical world was later inferred fiom cenain events in the

experienced world, how cm 1 be expected to ignore the experienced world?

Mer dl, it still remains my only basis for any gesses about physical facts

(p. 24).

Kohler's remark was aimed at the behaviourists, and there is no doubt that Brentano and

Stumpf would have agreed whole-heartedly. One page further on, Kohler said "there is no direct access to physical and physiological facts" (p. 25). An important cnticism of the

Berlin school agauist behaviourism is thus grounded in Brentano's psychology as transmitted to them by Stumpf Stumpf was one of the principal students of Brentano and his interpretation of his teacher led him to believe that psychology was not one of the natural sciences. Kohler and Stumpf s other students would not hold the notion that psychology is a separate science. Innead, they would arye that psychology is one of the natural sciences.

Nonetheless, they would continue to use StumpPs expenmental method. This disagreement beween Stumpf and his midents seemed to be more a matter of words or theory, although. no doubt, of importance, than of the actual way of performing experiments.

We saw that Brentano thought physical objects to be less real in the sense that they have ody a phenomenai and intentional existence. Nonetheless he did not deny that mental phenomena could be caused by physical input.

Sensations are effects of physical stimuli. Their origin is thus a

psychophysical process (Psvcholopv from an Ern~incalPoint of View, p.

46).

The distinction between the physical and the mental sciences is not only of theoretical importance, but also has a bearing on psychological methods. We are conscious of what we perceive, not of that which enables us to perceive. We see a flower, not our eye. the visual cortex, or whatever physiological cause is related to that perception. Brentano taught that the basic function of the mind is the perceiving, the grasping, of objects. Stumpf would explain intentionality in the following manner:

One never says that a colour quaiity is lighter, a tone impression higher, or

a muscle impression more intense than another one, instead one says that a

glove, a type of paper, a flame is lighter, a body heavier than another one,

or that a string is out of tune. The aim of a person reflecting is neither knowledge of one's sense impressions as nich. instead it is knowledge of the

world throua one's sensations (Tone Ps~choloa,p. 3).

Stumpf did not deny that one can reflect on one's sense impressions, but t heir aim is to corne in contact with the world.

The implication t hat perception is perception of somet hing, is directed towards something, becomes clear when we look at Brentano's waniing against the use of introspection.

Psychology, Iike the natural sciences, has its basis in perception and

experience. Above all, however, its source is to be found in the imer

perception of our own mental phenomena.. . .

Note, however, that we said inner perception and not introspection, i. e. inner

observation, constitutes this pnmary and essential source of psychology .

These two concepts must be distinguished 60m one another. One of the

characteristics of inner perception is that it can never become inner

observation... .

If someone is in a state in which he wants to observe his own anger ragng

within him, the anger must aiready be somewhat diminished, and so his

original object of observation wouid have disappeared.. . . we can never focus

Our attention upon the object of imer perception (Psycholog from an

Ern~incalPoint of View, p. 29/30).

Brentano, writing in German, uses the tems innere Beobachtung and innerc Wahmehmung. So to appreciate the above passage. we need to know what these German terms mean. Perhaps Imere Beobachtung can best be translated as limer-inspection' (or introspection. the word used above). The translation for innere Wahrnehmunq could be

'inner-perception' in its aspect of experience. Brentano's translators used 'i~erobservation' for imere Wahrnehmunq .An example can dari@ what Brentano was hinting at. We can observe a rose; we can see it, walk around it, and even make al1 kind of measurements. That is possible because the rose will not disappear. This kind of observation. so typicd of the natural sciences. is clearly impossible with our experiences. We cm perceive pain, or anger and so fonh. Although we can experience the- they are not open for observation in the same way as the rose is. If we try to see our pain, or anger, 'walk around' and 'rneasure' it, the object of our observation disappears. So we can experience mental phenomena, but not inspect them. As a consequence, introspection, being the inner counterpart of the kind of observation carried out in the natural sciences, is largely impossible.

Introspection is "the supposed focusing of attention directly on mental phenomena

thernselves... " (Macnarnara, 1993, p. 1 19). This. however, is impossible. as everyone can verifi by trying to observe his or her own anger. The anger will disappear, as Brentano pointed out, as soon as it becornes the object of observation. Note that introspection is impossible for mental phenomena. So, we cannot introspect the mental phenomenon of thinking a problem through. One is conscious of one's thinking, but one cannot introspect it. The result of the thinking when expressed, be that a judgement, a conclusion, or a mathematical proof. is, of course, open to general debate.

Lntuition or the perception of inner events, gives rise to propositions that Brentano thought to be irnrnediately evident and true. This could only occur in inner perception. in contrat to outer perception, since the latter phenomena cm aiways be doubted on logical grounds.

Introspection is thus not feasible for psychology. One rnay not conclude, however, that psychology is not a science. On the contrary, the method open to psychology will lead to tnrthful statements. Therefore psychology is a science. Brentano's insight guided Stumpf and his students, as well as Meinong and his followers, in their experimental work. They never asked their subjects to use introspection. instead, their subjects just had to report their experiences. What did they perceive? One note, a figure?

Thus the differences between mental and physical events have far-reaching implications. The characteristics of mental life are, according to Brentano, so different frorn physical (or physiological) phenomena that neither one can be descnbed adequately in tems of the other. Both have to be studied separately and explained in their own terms. If two things are fundarnentaily different, the study of those things must aiso be different. Hence, physicai phenomena are analyzed by the physical sciences, while psychology proper studies the mental. LandefeId ( 1937) remembered that

The definition he [Sturnpq gave, however, is his well-known statement that

psychology is a science of the elernentaq psychic functions, those functions

which are directly given and which by analogy with ourseives we conclude

exist in other beings (p. 33).

Psychoiogy was for Brentano and Stumpf not a science like physics, chemistry, or even biology. However. that does not mean that it is less scientific. It is also in search of laws, namely the ones that apply to mental phenornena. Neither does the statement imply that there is no psycho-physical relation. There could very well be, but if one wants to say more about that relation then one should know more about the psychological side as well as the other side. Both. however, should be çtudied according to their own proper methods.

But psychology has rejected Brentano's waming. It wanted so desperately to be scientifc that it modelied itseifafler the natural sciences and tried to implement its methods, which Brentano argued against, unsuccessfùily. Since 1874, the date of Brentano's book the natural sciences had made much progress. Most psychologists would be more acquairi.ted with the technological application of the natural sciences. and those are indeed moa impressive. The technological revolution enhanced the prestige of the natural sciences and added to the lure of the methods of those sciences, making them irresistible to psychologists. Murphy (1950) noted that

As one dips into the work of psychoiogists in any period from the pre-

Socratic to the present, one may run upon phrases which deny the possibility

of explaining wholes by a study of their constituent parts... .

It is, however, worthwhile to note that relationships or modes or

organization are repeatedly stressed by early Greek thinkers (p. 284).

Murphy mentioned Pythagoras and Plato but, strangely enough, not Aristotle. He considered Ernst Mach and Ehredels to be the immediate forerumers of the gendt school.

We saw in the introduction that Smith (1 988) related Ehrenfels' essay to the classical division of matter and form. 1 agree with these authors and wiil corne back to this point.

Christian von Ehrenfels was an onguial and creative thinker. He grew up in the privileged milieu of the Austrian rural gentry. His On 'Gestalt Oualitieg is an early work that had an immediate impact on the small world of German psychologists. Nonetheiess, he did not pursue this topic any fùrther. When he was professor at the German University of

Prague, he wrote several books, including an opus on sexuality. His view of these rnatters, however, was controversial and would not be shared by his contemporaries. We saw that his article on gestalt perception has a unique place in Ehrenfels' oeuvre. But it also stands alone in that it was his ody paper that resonated. AU his other studies had no effect on psychology. Reinhard Fabian (1986) wrote a very readable biography on Ehrenfels, and 1 hope that it will eventuaily be translated into Enghsh. The Article

Ehrenfes demonstrateci an excellent grasp of the problem of gestalt perception and an onginaiity in its solution. He started it with a cryptic remark:

The discussion which follows has the aim of giving scientific expression to

a psychological problem often noticed in philosophy but not yet, as we

believe, made fÙUy precise (p. 82).

The aim of his article was to demonstrate in psychological tenns facts which were already known to philosophers. Ehrenfels, being a -dent of Brentano, was well acquainted with the works of Aristotle, and it is only reasonable that he had those in mind. Ehrenfels? however, did not narne Aristotle. But he informed us that his analysis was prompted by . . remarks made by Ernst Mach in (1886) Contnbuaons to the Amdvsis of the Senses. Therein

Mach stated that we experience directly, without a mediating step or an intellectual processing, space gestaiten and tone gestalten. That could very well be tme. Still, Ehrenfels' rernark should not obscure for us the dinerences Ween the two men. The comment surely does not imply more than it says, namely that Mach tnggered some thoughts in Ehrenfels.

Mach, for instance, proposed a "complete paraiielism of the psychical and physical"

@. 30). He recogruzed "no gulf between the two provinces (the psychical and the physical)"

(p.30). Strictly speaking, parallelism does mean that psychical phenomena cannot be coiiapsed on physiological ones. Nevertheless, in Mach and cenainly later in mainstream psychology it came to stand for reductionism. Brentano denied that mental phenomena could be reduced to physical counterparts, and Ehrdsshared ttns view. Neither Brentano, nor Ehrenfels was in agreement with Mach's budding logcal positivism.

Ehrenfels' article appeared in the Vierteljahrsschrift &r Wissenschafiliche

Philosow, which was edited by Richard Avenarius. Mach and Avenarius shared the same philosophicai stance, and Ehrenfels' reference to Mach may have been intended to elicit the editor's sympathy. Ehrenfels visited Avenarius a couple of years before he wrote his article

(see Fabian, 1986). Avenarius must have expiained the ideas of Mach, whorn Avenarius admired, to his young visitor.

Ehrenfels introduced the subject matter of his article through a question:

1s a melody (i) a rnere sum of elements, or (ii) something novel in relation

to this sum, something that certaidy goes hand in hand with, but is

distinguishable fiom the sum of elements? (p. 83)

We hdthe sarne question stated differently wo pages further.

Let us suppose, on the one hand, thaî the mries of tones t 1, t2, t3,.. . .t(n), on

being sounded, is apprehended by a conscious subject S as a tonal gestait (so

that the memory-images of al1 the tones are simultaneously present to him);

and let us suppose also that the surn of these n tones, each with its particular

temporal determination, is brought to presentation by n unities of

consciousness in nich a way that each of these n individuals has in his

consciousness only one single tone-presentation. Then the question arises

whether the wnsciousness S, in apprehending the melody, brings more to his

presentation than the n distinct individuals taken together. An analogous

question can clearly be raised also in regard to spatial shapes (p. 85). The similarity of Ehrenfels' train of thought to Brentano's rhetorical asking if one cm unite the hearing expenence of a blind man with the seeing experience of a deaf, is, of course, striking. Indeed, what seems to be original in Ehrenfels' thinking is precisely that the approach is applied to a single modality, such as audition.

Sometimes, to ask a question is to presuppose a certain type of answer, as is the case here for Ehrdels. A melody is different from an assemblage of its elements. Why does he think so? Because a melody cm be played in different tunes; it fuifils the critenon of transposability. One can play a melody in a certain key. Remarkably, the same nine is recognked when played in a different key and even with a different instrument. We hear the identical piece of music, although the tones are quite different in each presentation. The musical keys differ and therefore so do the tones, but not the melody. A melody, then, cannot be equal to the sum of the individual tones. Harry Helson ( 1925) formulated

Ehrenfels' position as follows:

A configuration possesses properties over and above its parts and not

denvable fiom them; the configuration is transposable since it does not

depend on any given set of elernents (p. 360).

Ehrenfels also noted that it is easier to reproduce a melody than an individual note.

If someone needs a hi& C, for instance, the singer will sing a piece of music that will Iead to that note, Uimead of singing that tone immediately, without the help of a tune. Directly hitting a high C is only possible for those who have perfect pitch. It seerns that we have easier access in our memory to a tune than to one of its elements. mer having demonstrated that the whole is not a mere sum of its elements,

Ehrenfels mut now corne up with an explanation: the gestalt quality. That is, a perceptuai property attached to a number of distinguishable perceptual elements.

Gestalt quaiities are given in consciousness simultaneously with t heir

foundations, without any activity of the mind specifically directed towards

them (p. 112).

Ehrenfels thought of a gestalt quality as a "new positive element of presentation" (p. 92).

Ehrenfels also gave the following description of a gestalt quality:

By a gestalt quality we understand a positive content of presentation bound

up in consciousness with the presence of complexes of mutuaily separable

(i.e. independently presentable) elements. That complex of presentations

which is necessary for the existence of a given geaait quality we cal1 the

foundation of that quality (p. 93). in Boring's ( 1929) words Ehrenfels

thought of the fonnquality as distinctly secondary to the Fundamente, as

variable independently of them, but not independently given (p. 43/44).

To understand aU of this, we have to keep in rnind Ehrenfels' principal exarnple of a melody. A melody is deup of distinguishable perceptual elements, the notes. Ehrenfels believed, however, that with those notes there is something else given to consciousness, namely the gestalt quality. Max Wertheimer ( 1924) explained Ehrenfels' solution as follows:

When a melody is made up of six tones, and 1 reproduce it, playhg six

totdy different tones, and it is recognised - what remains? Quite certainly,

these six elements are a surn there ...; but besides those six a seventh must be

assumed, that is the gestalt qudity. The seventh is that which makes it

possible for me to recognise the melody (p. 10).

Smith (1988) gave an explanation similar to Wertheimer's, aithough stressing the atornic aspect of Ehrenfels' notion. Gestalt quaiities "are additional unitary objects, existing alongside the unitary elements with which they are associated" (p. 1 7).

If a gestalt quality exists then, one wonders, how do we perceive it? 1s there a speciaiized organ, or is the gestalt quality of a melody perceived by the ear just as the reguiar tones? Ehrenfels did not believe that the senses perceive gestalt qualities. He said, however, that the perception of a gestalt quality wuld be the resuit of an interaction of brain

States, each caused by an individual note. A togethemess of stimuli could be the physiological base for the emergence of a new mental element. Togetherness of stimuli, however, does not autornaticaly resdt in a gestalt quaiity. At the end of his article Ehrenfels formulated another law conceniing the emergence of gestalt qualities, namely cont rast &om its airroundings. In other words, Ehrenfels believed that the brain functions in such a way that it evokes a new mental element, the gestalt quality, when the right kind of sensory information reaches it.

Wertheimer used the tenn 'extra element' to express Ehrenfels' notion of gestalt quality. No doubt that description points in the right direction. But I think that a gestalt quality cmot be considered as an extra note, a seventh one, instead it has to be distinguished f?om the elements. A melody is transposable, exactly because the gestalt quality does not depend on any given set of elements. That makes gestalt qualities distinguishable in a one-directional manner.

Kthere is this gestalt quality, then notes will be heard as that particular melody. The notes depend on the gestait quality to be recognized as that piece of music. The other way around, however, does not hold. If the notes are t here, but the gestalt quality is absent, t hen there is no meiody. Therefore a melody can be presented without that particular foundation or these tones, but not that foundation or those tones without this particular gestalt quality.

Consciousness perceives the gestalt quality directly and passively. A gestalt quality exists with the notes, and one experiences it without any effort. In some cases, such as looking at a huge painting, Our effort is oriented towards the perception of the underlying stimulus complex. Still, the gestalt quality itselfemerges without a conscious effort aimed at the production of the gestalt quality. The gestalt quaiities are mentally given with their perceptual elernents. The gestalt quality does not depend on consciously directed activity on the part of the perceiver. To support his last rernark, Ehrenfels could have pointed out that one cannot choose not to hear the melody if the sequence of notes is presented.

How do we know that this series of notes has a geaait quality and that one not?

How do we know if a mental content is based on a gestalt quality or not? The answer can be found in Ehrenfels' definition of gestalt quality. If the elements can be transposed without altering the mental content, then a gestalt quality is present. If, on the other hand, the elements cannot be transposed, then there is no such quaiity in the stimulus complex. So, if we cmplay a series of notes in another key, without losing its identity. then the notes are accornpanied by a gestalt quality.

Ehrenfels' Contribution

Ehrenfels used the word gestalt only once or twice on its own, without adding the word quality He points out that a gestalt, or a whole, is made up - paradoxically - of distinguishable, and even independently existable paris. Quality is a tem equally used for the perceptual level. We can have the quality of colour, or of tone. for instance. Both words belons properly in a theory of perception. To chri@ this remark, we have to go back to

Aristotle's distinction between the forma1 and material cause. In the Aristotelian tradition, the tùnction of the active intellect is to detect the blueprint, the structure, the fon. That is a cognitive fiuiction. However, at what level in Ehrenfels' theory is a melody detected from its constituent parts? Indeed, at the perceptuai level! John Macnarnara just before his fatal illness, and myself as CO-author( 1995) formulated this reading as follows:

What Ehrdels proposes is a peruform (or gestalt) the counterpart at

the percephial level of the Aristotelean form, which in hstotle's psychoiogy

is accessed only at the conceptual level (p 4 13).

That there should be a counterpart at the perceptual level is precisely the import of the terni

'gestalt quality .' Ehrenfels thus translated the Aristotelean image of fonn to the perceptual level. According to Ehrenfels:

C haractenstic of gestalt qualities is that they constitute a presentational

61 content dependent upon yet distinguishable from their foundation (p. 257).

.A gestalt quality is a presentationd content, a content which is experienced in consciousness. Therefore it stands on equai footing with colours and tones. Like Aristotle's fomn, Ehrenfds' gestalt quaiity is not noticed by a perceptud organ, but discemed higher up, albeit passively. 1 think that Anstotle's mode1 lies behind this sequence. In the Aristotelian psychology, however, form is detected by the active intellect. Ehrenfels' gestalt quality, on the other hanci. is discemed passively and therefore belongs not to the cognitive system but to the perceptual one.

. . Gestalt auai ities as building blocks

Ehrenfels also discussed the existence of higher order gestalt qualities, where he distinguished between several types. In smell, elements from two different fields, pressure and temperature, are cornbined. The taste gestalt is a mixture of three kinds: pressure, temperature and smeU. These higher order gestalt qualities are reminiscent of the operation of Aristotle's sense that is common. The sense that is cornmon receives information of the specialised sense organs and integrates it. The brain, of course, does the same. Another example of a higher order gestalt is ballet. Ehrenfels asked himself rhetoncally: "Whether a cornplex ofmelody and visible movernent may not yield gestalt qualities of a higher order?

(P. 107)

Another highly instructive kind of higher order gestalt is a unity such as animai species. Presumedly Ehrenfels has in mind that a species forms a higher order unity in that its members d share the same specific nature. What Ehrenfels may have been reaching for is that gestalten do not just exist in the world of immediate experiences, but also in the world of ideas. ..ay,this line of thinking clearly shows the Ariaotelean origins of the whole notion of gestalt quality.

A doubtful explanation

Ehrenfels recognized another fùnction for geaait quaiity, namely to apprehend

Mnilarities. For instance, we can recognize the composer of a new song because we heard his or her songs before. And at a party we are able to recognize as brothers two men who are complete strangers to us.

Thus we recognize the composer of a melody through its similarity with

other, farniliar melodies, through without our being in a position to specify

more precisely in what this sirnilarity consists. We recognize the relatives in

a family in a resemblance rnanifested in their whole physical nature and

bearing, a resemblance which often stubbomly resists analysis into relations

of identity between individual constituent parts (p. 106)-

It seems that Ehrenfels had in mind that we perceive gestalt qualities. In the case of the songs or of the brothers the gestalt qualities have a similarity which is noted. The elements of the cognitive operations are then not the tones or the physicai characteristics of the two men, but th& gestalt qualities. Similanty, then, is an immediate perceptual phenornenon for

Ehrenfels and not a conceptuai one. Ehrenfels could have referred here to the doctoral thesis of his teacher and fiend

Alexius Meinong. Meinong f 1877) noted in his Hume studies (see p. 6 1 ) that mernbers of a family or inhabitants of a nation can share certain characteriaics. When writing his thesis,

Meinong believed that we really observe those resemblances. He discarded the notion that those observations are the result of mechanisms of imer processing of informatioc a solution offered by the theory of associationism. Ehrenfels' gestalt quality is an explanation of that phenornenon that is in line with Meinong's earlier thesis.

It is doubtful that Ehrenfels' explanation suffices here, since he did not explain how one recognizes sirnilarities of gestalt qualities. In order to recognize a similarity, we need something that detemines what characteristics we have to look at; what are the relevant attributes that tell us that things are or are not the sarne. Gestalt qualities, obviously, will not do t hat .

Ehrenfels also believed:

That the larger part of bath our everyday and our scientific vocabulary

designates gestalt qualities (p. 108). .. .

Thus gestalt qualities comprise the greater part of the concepts with which

we operate (p. 1OS).

Here Ehrenfels linked gestalt qualities to Our concepts. His view seemed to be that we recognbz amilarities in the gestalt qualities and that we give those sirnilarities a narne that is a designation in words. Ehrenfels had corne a far way from the raw experience of notes and their melody from which he departed.

Ehredels came up with gestalt qualities to explain the recognition of melodies. Then he applied the same concept to our recognition of family rnembers, or nationalities. Thus,

we need a gestalt quality to note the similarity of pieces of music. Don't we then also need

a gestalt quality to note the sirnilarity of gestalt qualities? And does this not lead to an

infinite regress of needs of gestalt qualities? Ehrenfels assumed that the brain is stmctured

in such a way that it notices gestalt qualities. His way out of our dilernma must then be to

assume that the bmin is also capable of noticing similarities between gestalt qualities. Then,

however, an obvious objection arises. Sensory elements can iduence the brain. But how

can a gestalt quality, a non-sensory element, an element in consciousness, stimulate the bmin? It seems, therefore, that Ehrenfels could not explain how one detects similarities of gestalt quaiities.

at does fiture have

Ehredels ended his article with the thought that tones, colours, and so fonh could

very well be made up of more simple elements. He arrived at this conclusion with the

following argument: Wh practice one can discriminate the individual sounds in a chord. If

one continues this decomposition it wodd be at least theoreticaiiy possible to "finally arrive

at a single protw,or at lest a single quality continuum" (p. 1 15). The experience of

colour, tone, etc. would emerge out of these simple elements. Since our knowledge rnust

respect this chernistry,

it fioiiows that the derivation of al1 contents of presentation from a cornmon

proto-element would yield the possiiility of comprehending the whole of the known world under a single mathematical formula (p. 1 16).

Ehrenfels very likely had an analogue with chemistry and physics in muid. Water. wood, iron, and so furth cm be broken down into molecules, and tùrther into atoms, then in protons and electrons and so on. The wide diversity of matter is based on some common- proto elements. Those elements have relations to each other and those relations are mathematicaliy describable. The sarne, according to Ehrenfels, mua be tme for consciousness. This leads to the conclusion that the mental inexistence of objects cm be describeci as well in mathematical tems. In this view, mental objects are made up of proto- elernents (which I assume, are not conscious), elements, gestalten, and higher order gestaiten.

Smith (1988) pointed out that the language of this passage is fundamental atomistic.

"the world as a whole is ultimately atornic in stmcture" (p. 16). These atomic units are in relation to one another. Those relations, according to Ehrenfels, can be formulated exactly .

Psychology, the study of the mental, is then a science and will lead to scientific laws. With that conviction Ehrenfels showed himself a student of Brentano. EDGAR RUBIN (1886-195 1)

Ash (1982) wrote that "the importance of Rubin's work for Kohler's theory is diflicult to overestimate" (p. 462/463). That remark is true, although it may still underestimate Rubin., sigdcance. He had not only a considerable influence on Kohler, but on gestalt psychology in general (see, for instance, Murray 1988). Koaa wrote a review of Rubin's book in 1922 and devoted a whole a chapter to the figure-ground phenornenon in his ( 1935) Princi les of Gestalt Psvcholoq. A reflection on the ongin of the gestalt schwl must, therefore, look at Rubin's work in more detail., ahhough Rubin would propose an explanation that differed fiom the Berlin gestaltists.

Edgar Rubin was a Danish psychologist and a student of Harold Hoffding. He worked for two years in G.E. Müller's laboratory at Gottingen (çee Ash, 1982, p. IX) and published in 19 15 in Danish the results of those experiments (5 soplevende Fimer). This book was translateci into German in 192 1 under the title Visuel1 wahr~enommeneFimiren.

It is this translation that we will use. Michael Wertheimer made an abridged English translation of the German version (mreand C~QU~).

Rubin showed his subjects pictures or slides with two meaningless coloured fields.

Those fields were so situated that one field enclosed or, we could Say, encircled the other.

The expriment consisied of two phases. During the first or leamhg phase, the subject was s i F- i fi- = IO 10 show a series of slides and instmcted to direct his attention to one of the two fields. One could pay attention to the field that surrounded, or to the other one, which was enclosed.

In the next phase of the experiment, that of testing, the subject was shown the meslides again in random order and intermingied with new but similar slides. He was then asked some questions. Those could be on the field that was 'figure,' in other words the one paid attention to during the learning trials. But the questions could also be about the other field, or

'gound,' to which the subject did not pay attention. I do not know if someone gave Rubin the idea for his experiments, but 1 have not fourid anyone claiming that .

Rubin noticed that if the subject focused his mention on a particular field during the leamkg, the subject was inclined to perceive that same field as figure on subsequent exposures. Thus, when one was instructed to look at the wrounding (enclosed) field during the leaming trials and later had to report what one saw at the test trial, the surrounding

(enclosed) field was seai again as figure. There was thus no shifi in allegiance. Rubin denied that this phenornenon could be explained by the theory of associationism and he gave a simple reason for his rebuttal: there exists only one element, the figure, and not two between which a connection can be formed. This, according to Rubin, exciuded association as the mec hanimi responsible for the figure-ground p henomenon. An association presu p poses the existence of more elements between which a Link will be established, whereas here only one element was present .

hother experimental finding concerned recognition. The field that was best recognized was the one that was figure in both the leaniing and the test trials. Subjects performed slightly les weil on fields which were ground during both trials. The performance dropped dramatically, however, when the fields reversed their roles. Fields t hat were figure

(ground) during the learning trials and later tested as ground (figure) were ahost never

recognized. In these cases "these fields will not be recognized again during the resting" (p.

26). Rubin conciuded that:

The experiments show that recognition depends completely on how the

differentiation takes place [narnely the differentiation between figure and

ground during the 1e-g trials] (p. 27).

Rubin's experiments had now proven that, mentdly, a difference exias between fields expenenced as figure and as ground. experienced ground are two differently experienced objects, which can be evoked by one and the same objective, specifiable stimulus object [objektiven Gegenstand] (p. D(). This finding resulted in more questions, such as: Are there "probability rules that an area is perceived as a figure?" (p. 79). This is an ernpincal question and Rubin found the answer through empirical means. Not

surprisingly, he found that such rules indeed exist.

lfone of two homogeneous, dserently coloured fields is appreciably larger

than the other one, and if it surrounds that other field, then there exists a

great probability that the smaller enclosed field is perceived as a figure (p.

79).

Notice that in Rubin's experiments one field was nûmunded by another. Rubin detected that the enclosed field had a higher chance of being seen as figure than the other one. A subject, however, could overcome this tendency and treat the other, surroundhg field as figure. Rubin also found that with a repetition of motifs, al1 the motifs are experienced the sarne way, al1 as figure or al1 as ground. Perhaps the example of wallpaper with a repetition of several motifs can make this clear. One motif stands out as figure for the whole wall, while the others are background. One does not see this motif at this spot and that one somewhere else. The same is true for "a coherent, monochrome field" (p.83). The whole field is seen as figure, or as ground. It is thus not the case that a part of a field is seen as figure and the rest as ground.

Knowing now that a ditErence between figure and ground exists, and having a better idea what makes a field a figure or a ground, Rubin wanted to know "what does this difference consist of?" (p. 36). This question calls for a more qualitative answer. Rubin formulated it this way:

The experienced figure and the experienced ground do not both have form

in the same manner, since the experienced ground has in a certain sense no

form (p. 36).

Or, otherwise formulated: "The figure has more the trait of an object and the ground more the trait of a material" (p. 48). Another relevant quote reads:

That the expenenced object, when it is a figure, by and large is more vivid

than when it is a ground, and .. . it prwaiis or dominates in consciousness (p.

67).

Rubin also had this to say on the dflerence between figure and ground: "From a psychological perspective the experimental figure is so to speak more processed than the experienced ground" (p. 96). He noted as well that the figure has "a certain independence" (p. 70), since "it is of little importance for an experienced figure if it is seen against cne ground or another" (p. 70).

The Explanation

There exists thus a nuidamental ciifference between the field perceived as figure and the one perceived as ground. The figure is that which we see; the ground just that, background.

Rubin had established the figure-ground phenornenon through experimental means.

Now he had to offer an explanation. This brings us to the point where Rubin believed that experiments are no longer applicable. It is perhaps surprising for the modem reader to note that experirnents are absent in this phase of Rubin's project. That was, however, the practice in eariy German experimental psychology. One of the founders of experimental psychology was Car1 Stumpc and Rubin's expenmental method did not difFer fiom Stumpf S. Although

1 have to add immediately that Rubin did not wodc under Stumpf, but under Müller and that

Müller also carried out important psychologicai experiments, both Stumpf and

Rubin used experiments to demonstrate the existence of mental phenomena. Stumpf proved tone fusion, Rubin the figure-ground phenornenon. Both gave a theoretical explanation of their findings. Those explanations were thought to be subjective, based on the creative insight of the researcher. Facts were thought to be objective, not their explanation. In this view, explanations cannot be experimentally proven and neither Stumpf nor Rubin would put their arplanation to a test. Eariy Gerrnan psychological experiments were solely camied out to elicit mental phenomena not to veriS, or fiilsi@ a theory.

in order to understand Rubh, one has to keep in mind that he considered the mental different fFom the physical. That was also a common notion in early German psychology.

Let us now look at Rubin's explmation in more detail.

Rubin thought that the 'contour' between the figure and its ground played an important role in visual perception. Contour should not be confûsed with 'border.' The border between two countries is a no-man's-land between those countnes. It belongs to both, or, ifone prefèrs, to no one. l'ha& however, is not the case with Rubin's contour, even though not far into his book he defined contour as a border, or "as the mutual border of both fields" (p. 36). This definition, however, does not seem accurate. Rubin conceptualized contour as an active entity - it does something, it has "a moulding effect" [ein formendes

W&en] (p. 36). We, however, do not notice the moulding, but only its result. "The figure is especially afFected by this moulding effect, and not so much the ground" (p. 37).

Therefore the contour is the demarcation of the figure, not of the ground. A contour is thus not the rnutual frontier of both fields but oniy that of the figure.

The mutuai contour has significance for the field that stands out as a figure,

because it is that field's border, which says: here that thing cornes to an end,

this mutuai contour seems to have so little in common with the ground that

one can continue to exist freely past the contour (p. 39).

The contour denotes the ending of the figure; it belongs to the figure. As a consequence,

"the contour is more relevant to the field Bcperienced as a figure than to the one experienced as ground" (p. 41). Since the contour belongs to the figure, it cannot be the case "that the munial contour has a different rneaning for the experienced figure and the experienced ground" (p. 4 1). Here, Rubin denied that the contour has one meaning for the figure (the ending) and another one for the ground (whatever that may be). tt seems thus that Rubin believed that the contour has no meaning whatsoever for the ground.

Till now we have used quotes suggesting that the contour could be a real physicai object. Rubin's following remark, however, contradicts that. He said that a contour is "not a dace, it encloses an area, but it does not constitute one, it has length but no width" (p.

106). There are some obvious consequences of this view. Since a contour has no width, it also cmot have colour (see p. 106). But it has a length, so we can track it with Our eyes, and that, of course, can take time (see p. 1 Wl6O).

To recapitulate, Rubin used fields with different colours in his experiments. There was thus an edge where those colours met, where one field finished and the other started.

That, however, was in Rubin's eyes not the contour. A contour is a non physical - since it has no width - but nonetheles active entity demarcating the figure. It should therefore also not be confused with the black lines one finds in children's colouring books separating the figures from each other and the background.

Rubin's concephialization of what a figure is also goes agakt Our common sense notion. A figure, according to Rubin, is a complex made up of parts. So far, so good. An example will illustrate what Rubin meant with this statement. A hand has a paim and five fhgers. Rubin's parts, however, were not the palrn and those five fingers. He distinguished between the field (in the fom of palm and fingers) and its contour. Both entities, field and contour, play a role in forming the figure of a hand. There is a difficulty for the relation between figure and contour in that the

normal experience of figure consists of a complex of two kinds of objects,

the experienced surface figure and the expenenced contour; and it is a

difficult task to make clear that there are different objects here with which

one has to deal (p. 145).

We find no reference in Rubin's book to Franz Brentano. Brentano differentiated between mental and physicai phenornena. Rubin shared that opinion. His subjects were presented with two neighbouting fields. They saw, however, a figure against a background.

Physically there seerns to be no difference between the fields, but rnentaiiy there is one.

Rubin even proved experimentally that one and the same field can be perceived as figure or as ground. In his words there are

two different expenenced things, which, under different subjective

conditions can depend on one and the same objective stimulus condition (p.

101).

The distinction between figure and ground does not stem "fiomthe nature of things" (p. 94).

The only explanation, according to Rubin, is the fact " that the inner life is by and large so equipped" (p. 95).

Rubin also conternplated why "our inner Iife is so equipped" (p. 94). He conciuded, not unreasonably, that this arrangement rnust be benekial to the observer. But he did not explain why it came to be. Therefore Rubin cannot be considered, at least not from this text,

as a forenimer of today's evolutionary psychologists.

Another proof of Rubin's differentiation between mental and physical p henomena can

be found in Rubin's disaission about the perception of a he. Rubin's research led him to the conviction that :

Under certain circumstances the line on be experienced as having no width,

namely when the visual angle is suficiently smaii (p. X).

In his treatment of visual objects without expansion [Ausdehnungslose

Gesichtsgegenstiinde] Rubin wrote:

[t is a fairly often discussed question, if the contour or the line is an actual

sensory visuai shape or not . Considenng that, it is surely not emphasized in

the more recent psychological literature that the specific line dSers from a

stripe essentially in that the line can be one-dimensional. It is true, one has

spoken of lines without width, but one considered that as something that

cannot be seen, something abstract, what emerges ~oma kind of border

crossing. One has not noticed that iines without width, although unthinkable

as objects of nature, can be experienced sensorially (p. 193).

In other words, lines without width are physidy impossible but, nonetheless, they can be seen. Note that Rubin discussed here the perception of lines, his use of the word contour notwithstanding. This passage, therefore, should not be confused with his treatment of the role of contours in the figure-ground phenomenon.

The difference between the mental and physical is also evident in the following quote:

Since we are inclined to think the mental in analogy with the objective, an

inchation ennies to avoid that which belongs [exclusively to t hese [mental]

facts (p. 192).

Like Brentano, Rubin claimed that mental phenornena have certain characteristics that distinguish them from physicai ones. If certain characteristics are not shared with the physical ones, then we wiII lose them if we treat the mental as identicai to the physicd.

Rubin and Brentano would argue that a researcher who overlooks the difference between the two realms will as a result miss the phenornena belonging exclusively to the mental world.

Finally, I would like to point to the ending of Rubin's book. He finished his description of his experimental work with an excursion into philosophy. Rubic believing that his findings shed some light on Hume's work, made some cornrnents on Hume. Those could be inspired by Müller, under whom Rubin worked and who was also a Hume devotee.

Nonetheless, the ending is for the modern reader of experimental research in al1 respects surprising. Hume's analysis of the mental, according to Rubin, was unhindered by preconceptions derived fiom knowledge of the physical realm. Hume studied only the mentai, and Rubin approved of that approach. Rubin ( 193 O), in a small review article which

is of no Werinterest to us, repeated that stimulus and experience should not be confused with one another, but be kept distinct at ail times. 1 believe, therefore, that Rubin would agree that the mental and the physical must be studied separately.

There is no reference in Rubin's book to Car1 Stumpf. Rubin, however, argued against Edrnund Husserl about an idea regarding the perception of space and colour. Husserl was a student of Stumpf and it seems that Husserl reiterated an opinion of Stumpf. Rubin also mentioned some other scholars whose work was influenced by Brentano. He referred to Ehrenfeis' article On 'Geut Oualities', to publications of Meinong, and to Meinong's students Vittorio Benussi and StehWitasek. Rubin knew thus works of Brentano's circle.

Contour versus Gestalt Ouw

It is worthwhile to compare Rubin's book with Christian von Ehrenfels' (1890) article, where a theory of gestalt quaiities is laid out. One of the ways Ehrenfels described a gestalt quality was as an enfolding united band [Umschlingendes einheitliches Band] (p.

28 1). This is, of course, also a fitting description of Rubin's contour.

Before continuing, we should note some obvious differences between Ehrenfels' and

Rubin's writings. Ehrenfels started with a thought experiment on melodies; Rubin did experiments with visual figures. Ehrenfels used gestaiten that make sense, whereas Rubin employed nonsense figures. The rnethod as well as the material thus dEered between the two psychologists. Those differences had a bearing on the questions the researchers can legitimately ask themselves. Ehrenfels' material makes it feasible to discuss the results of changes in part(s) of the gestalt. That is not possible in Rubin's case. One can study what happens when one makes small changes in meaningful units such as melodies. However, it rnakes no sense to ask if a nonsense figure remains the same or not after a smdl alteration. important as they are, these dïerences do not hit the heart of both theories. Theoretically more important is the difference between the conceptualizations of a gestalt quality and a contour. A gestdt quality, according to Ehrenfels, can only be present in the case of several perceptual elements (p. 252). Indeed, there is no melody when there is only one note; melodies demand several notes. A contour, on the other hand, is the contour of ody one field. There is thus only one perceptual element present. This difference could explain another one. Ehrenfels contemplated the existence of higher order gestalten.

Rubin's research, however, did not easily lead to the idea of higher order contours. Rubin limitai himself to figures consisting of one element. The qualification 'higher order' implies the presence of several elements. If we start with figures made up of more elements -

Ehrenfels' departure - it is no big step to think of gestalten made up of even more elements.

If, on the other hand, one starts, like Rubin, with a one-element figure, the step to a figure made up of sevdfigures cornes close to a conaadction in terms. In any way, it is a bigger step than Ehrenfels had to take.

There is however, a case to be made that figures and grounds may be hierarchically stnictured. Rubin, as far as 1 know, did not make this argument and neither did he pursue experiments to demonstrate it . Imagine a piece of white paper with a green spot on it and on that spot there is a red dot. In Rubin's analysis the red dot would (most likely) be seen as figure. Now one can argue for the following structure of figures and ground. First, the green spot with the red dot are separated from the rest of the paper. The green spot, being the smaiier enclosed entity, is figure, while the rest, the surrounding paper, is fieid. At the second level, the entity of the green spot and red dot is decomposed, and the smder, enclosed, red dot is separated as figure from the larger, surrounding background of the green field. In this analysis, there exists a hierarchy not unsirnilar to Ehrenfels' proposal.

In any case7 the suiularities between the gestalt quality and the contour are striking.

Fim of 4 both concepts are intended to explain the perception of fon in a single modality.

Then there is a dialectic relation between the gestalt quality and the contour on the one hand and their perceptuai basis on the other. The gestalt quality and the contour play a role in defining the perceptual basis. They make a tune out of a senes of tones and a figure out of a field. The gestalt qualiîy, however, depends also on its foundation, and the same holds for a contour. A gestalt quality is present together with the perceptuai elements and the contour with its field. Furthemore, both the gestalt quality and the contour do not depend on our own inner powers of interpretation. A melody or a figure is not the result of our own productive powen; it ewists on its own. Ehrenfels' melody and Rubinfs figure are more than the sum of their perceptual elements. A gestalt quality is added to the notes, a contour to the figure-field, at least as far as theory is concemed. We have no empirical evidence that there is no act of interpretation involved.

The gestalt quality and the contour hold the same position with respect to their perceptual element(s), the tones or the figure-field. A melody does not need these special notes. It can be played in different keys, thus with dEerent tones. A melody is a combination of a gdtqua@ and the tones. The same is true for the contour and its figure field. A figure field can have different colours, but as long as it is nirrounded by the same contour, it is the sarne field.

Finally, the properties of the unity of a melody or a figure cannot be explained through theû perceptual elements, the notes or the figure-field. The gestalt quality or the contour has to be taken into account as well. They are, according to Ehrenfels and Rubin the cause of unity.

A gestalt quality is not a melody and a contour not a figure, hence an important difference fiom the notion of gestalt itself A melody is a compound of a gestalt quality and tones, and a figure of a field and a contour. The tones, or the field, is the aspect that makes the rnelody, or figure, a particular, unique, individual thing. It makes it this melody, played in this key, by this instrument, or this field, in this colour, with this hue. The gestalt quality or the contour provides the universai aspect. It makes these notes Beethoven's Fifth m,or this field a "face." Ehrenfels and Rubin believed that the universal aspect is not a creation of our minds. Instead they considered it a real entity. This could explain why a gestalt quality and the contour were described in physical terms. In a certain sense, they are out there. Both thinkers believed that we capture passively and directly both the individual and the universai aspects of the things in the world. As we hear tones. we hear its gestalt quality; as we see a field, we see its contour.

Both thinkers stressed that, Born the perceptual element(s), a gestalt or figure is formeci by something outside us, something that operates on the perceptuai elements.

Ehrenfels descrii it as an extra element, a kind of extra tone. Rubin described it as a line.

One can argue that neither description is weli chosen, since a gestalt quality performs fùnctions quite remote from the tones of a melody. Rubin acknowledged that a contour is not a real iine, and it too performs a fiinction different indeed frorn an ordinary contour. A gestalt quality foma melody and a contour a figure. Both thinkers used physical terms to point to that which organizes something into a whole. Perception of a melody or figure implies that we perceive the gestalt quality, or the contour in addition to the perceptual elements. Remember that the gestalt quality and the contour are emities "out there," not the result of processes "within us. " Although Ehrenfels acknowiedged that we have no specialised sense organ to perceive gestalt qualities, he believed that the brain is built so that gestalt qualities are noticed there. Remember that perception of gestalt qualaies takes place passively . Rubin believed that we see the contour.

Thar can take place in the eye or higher up in the brain. Wherever, that perception is also a passive process.

Cmit be that Rubin's contour is Ehrenfekfnotion of gestalt quality for the special case of oniy one elernent? Ehrenfels' description of an enfolding uniting band expresses that the gestalt quality separates the elements which fall within the gestalt from al1 other elements. A melody is separated from the noise we hear, albeit not by an audible envelope; a figure jumps out fiorn its background. Therefore, a gestalt quality could well be the contour of one element. The thesis that the contour is a special case of a gestalt quality, however, was challenged by Rubin himself! Rubin points to an example where the contour can be altered without affecthg the figure. The example is that of a stamp with a zigzag contour. One can change the zigzag line for a straight one without, according to Rubin, changing the figure of the starnp. If we assume that a contour is a special case of a gestalt quaiity, then the alteration of the border of the starnp would mean that we can change the gestalt quality without affecthg the experience of the gestalt! We would thus change the gestalt quality while hearing the same meiody. Suice that cannot be the case, a contour is not a gestalt quality for the case of one perceptual element, except, however, if we accept that in that unique case one may change the contour without aEecting the gestalt. That exception may be motivated on the grounds that the element will stay the same and therefore so will its (gestalt) perception. In the case of one elemem, the gestait quality needs only to separate the field f?om its background, while in the case of more elements, the gestalt quality has the additional task of uniting the elements. Although Barry Smith (1988) did not mention Rubh, he had something to add to the above discussion. "From Ehrenfelsl point of view, a gestalt quality (whole property) disappears when we isolate its parts" (p. 56). Smith argued that there is no special case of a one-element gestalt quality, and as a consequence a contour cannot be that special case. Note that one function of a gestalt qualitv, to isolate the constituent elements fiom the surrounding, must still be performed in the one-element case.

Let us return now to the example of a aamp and see where we anive when we assume that a hierarchical organization of figures and grounds exists. First we must

'translate1the stamp into this model. The stamp has a white zigzag border, with a coloured interior, and in that interior there is a figure (for instance the face of the queen) and at the top right corner a number representing the value of the stamp. At the first level of analysis the coloured interior is separated fiom the white zigzag border. At the second level the numesai and the fice are separated fiom the background. As far as the white zigzag border is concemed, there cm be a fiirther, second-level separation of the araight inner edge of the coloured interior from the outer ngzag edge. Now, the change of one element, in this case the zigzag edge, does not influence the other contours and therefore the overall 'feeling' of the stamp remains the same. A change in one of the elements (the zigzag edge) does not affect the gestalt quality of the whoie stamp. It seems that Rubin's mode1 cm be reformulated in hierarchical tems. Then 'contour' and 'gestalt quality' express the same function, namely the experienced unity.

Kob( 1922) in his raiew of Rubin's work credited Rubin with having shown that the figure-ground distinction is a fact on its own and cannot be explained through mechanisms such as observation or attention. Koaa also noted that this distinction is hdamental to psychology. Kohended his review by remarking that meanwhile - that is. between the Danish publication of Rubin's book in 19 15 and the review in 1922 - Kohler had shown that this distinction can be understood theoretically. The Iast rernark shows in rny eyes that Kofbcompletely Rissed Rubin's explmation. I do not think that KofEka ignored

Rubin's thesis. Ko& had shown in other reviews, which we will discuss later, to be fair and intelligent in treating other works. Kama was, no doubt, capabie of dealing with Rubin's theoretical notions. Since that did not happeq 1 believe that KoBa looked only at Rubin's experirnental work and the convincing way the problem of form was shown. Koffka simply overlooked or did not grasp Rubin's theoretical contribution.

That cannot be said of the poet Christian Morgenstern ( 187 1- 19 14) ~hme

The Fene anticipates Rubin's notions: ifform is made up of a field and a contour, then one can ask oneself What will happen if one removes the field and lave the contour? Here is

Morgenstern's answer:

Once upon a tirne there was a fence of palings with space between

them so you could see through them.

An architect, who noticed this, suddenly turned up one evening and took away the space between them and built a large house with it.

The fence meantirne felt rather foolish in its paiings without anything

round them, an unseemly and disgracefbl sight; and so the borough council

had it taken away.

The architect. however, made off meanwhile to Afnca or America.

We perceive form, although form has no "comprehensible stimulus equivalent" (J.J.

Gibson, 1950, p. 19) reaching the perceptual organ. The probiem of the perception of form was well stated by Ehrenfels and Rubin. However, its answer, gestalt quality or contour, proved unsatisfactory for succeeding psychologists. One of the first experimentd psychologists in perception, Car1 Stumpf, would argue that we perceive elementary units like tones and, in addition, also their relation. That means we perceive the abstract, the relation or form, in the concrete, the tones. His famous students, Wertheimer, Kohler and

Kama, would point out that an element gets its meaning From belonging to a fom. They came to the conclusion that we have to perceive the form first and then its elements. There are some objections made ag& their idea, but we will treat those later. First we will look at their teacher, Car1 Stumpf CARL STUMPF ( 1848- 1936)

Ash ( 1982) wrote that Sturnpf s roie in the development of the Berlin gestalt school

"was central, though by no means simple" (p. MX). Ash (1982) touched on some of

Stumpf s work but he did not go into great detail. Therefore, his staternent is in need of a more solid foundation. Kun Lewin ( 1937), a gaduate student of Stumpf and an important gestalt psychologia in his own right told us that: "Many of his [Stumpf s] views have a close relation to fundarnemal ideas in gestalt psychology " (p. 189). Lewin's article is affect ionat ely written, but it also did not substantiate the daim about the relation between Stumpf and his students.

Stumpf was bom in 1848 to a middle-class, Roman Catholic family in Wiesenthied, a smaii German village. In 1865 he went to the University of Würzburg, where Franz

Brentano was teaching. That University had at that time a conservative catholic orientation.

Stumpf was impressed by Brentano and Brentano's influence, as Stumpf acknowledged in his autobiography ( l924), would remain strong throughout his work. Brentano, who was not high enough in the academic hierarchy to supe~sea dissertation, advised Stumpf to move to Gdttingen to Merhis studies under Rudolph Hermann Lotze. Lotze would also sponsor Stumpf s habilitation-Gottingenwas a leading university Ui the naturai sciences in

19th-century Gemany (see Men, 1904). Stumpf spent some tirne there in laboratones under Kohlrausch, obtaining hands-on experience in the research methods of the natural sciences. Mer his studies in Gottingen, he held positions at several universities before becoming professor of philosophy in Berlin, the most prestigious university. He camied out experiments on tones before he was appointed in Berlin. Those expenments were reported in the two volumes of the Tone Psyçholoq. In Berlin Stumpf lectured in psychology and the history of philosophy, and developed and directed the Berlin Psychological Institute. At his retirernent in 192 1, Stumpf ensured that Wolfgang Kohler would succeed him. With

Kohler's succession, Stumpf put the gestalt school in an excellent position to grow. Stumpf died in 1936, at the age of 88.

Four of Stumpfs books will be discussed nameiy On the Psvchological Orign of

Spêce Perception den psychologischen Ursprung der RaumvorstelIung] ( 1873); Tone

[Tonpsychologie] (vol 1: 1883, vol II: 1890); Speech Sound [Sprachlaute]

(1926); and Theo~of Knowleu ~rkemtnklehre](1939). The 1st book appeared three years after his death. As fw as I know. none of these books has been translated into English.

The focus will be on the fim two books, since only they could have influenced his students during their formative years. The last two books were published well afier his students' ideas had taken form. Stumpfs last book, however, provides an oveMew of his psychology. It also contains a critical account of the Beriin gestalt theory. Therefore that book will be used to clarify Stumpfs earlier work and to illustrate his criticism of the Berlin gestalt school.

. . of Space Pe

. . Stumpf s firsi book On the Psvchological On.gin of S~acePercept io~(1873) begins with a meticulous treatment of the several existing theories on space perception. His descriptions are fair and accurate, especially those of Immanuel Kant and the British empincists. He went on to criticize them in the wne thorough, penetrating marner.

Mer the lengthy historical introduction, Stumpf unfolded his own theory. Central in Stumpfs anaîysis is the 'presentation' [Vorstellung]. At the end of his Iife, in a discussion on George Berkeley's idealism, Stumpf would translate Berkeley's 'idea' with the German

Vorstellunq (Th-, p. 587). That translation gives a wrong impression of

Stumpf s thinking. Both scholars agreed in the reference to mental phenomena. Stumpf s

Vorstellung, however, indicates Brentano's notion of psychologicd phenomena in their unifieci aspect; it is a whole consis~gof rnany phenomena present at a certain time, whereas the British ernpiricist's 'idea' refen to single entities (mostly images of extemai objects). The two notions are clearly distinct and it is therefore better to express them with different words. That is why 1 will translate Stumpf s Vorstellung as 'presentation.'

A presentation is a unified phenornenon, and like many wholes it has certain parts

[Theilel. But those parts are not objects like a dog, a paw, and so fonh. Instead, Stumpf uses the word 'part' to indicate dimensions on which values cm Vary. Colour is an exarnple of such a dimension. The part is then the colour dimension, and on that axis the different colours occupy thei. values. Space is another part, or aspect, or dimension, and like colour it is a possibility for change of values. This dwsnot mean that we perceive possibilities. The dimensions as such are not given directiy in perception; only the values that a perceptual object realizes on them are perceived. Psychological phenomena are not possibilities. But they are the result of processes that lead to values on certain dimensions; they thus presuppose possibilities of realization.

The question now aises, what is the relation between those dimensions? Some dimensions cm be expenenced separately from one another. We can taste the sweetness of sugar with our eyes closed and we can see the whiteness of sugar without tasting it . Hence, the dimensions of taste and colour can be experienced independently of each other. Stumpf would name those contents of consciousness 'detachable' [trembaren]. Other dimensions, however. mua be perceived together. Those dimensions belong together like the two sides of a coin; when one is present, the other rnust be there too. Stumpf narned these dimensions

'undetachable' [untrembaren]. Undetachable contents dways must happen together in consciousness. They cannot be isolated fiom one another in real life. Stumpf thought that colour and space form an undetachable content. We cannot see a colour without seeing some extension; nor space without it haWig a culour. Therefore, according to Stumpf, space and colour are different aspects of the same experience. Each colour we see is extended and each extension is coloured. Shirnpf believed there is no colourless space, not even in one's imagination. In that case its colour is black. The separation of space and colour fiom each other cm only take place in abstract thought. Stumpf ( 1939) formulated that insight as follows:

Stimulateci by Brentano and under his decisive influence, the author showed . . in his work On the Psvcholp$ical Onpof SpaPerwtion (1873) through

an anatysis that colour and extension do not just habitudy accornpany each

other, as the empirical associationist psychologists taught, but that by t heir

nature they are indivisible and most deeply comected, that they are different

sides or ways of change of one and the same indivisible content of a

sensation or presentation.. . (-, p. 183). There is a parallel between Stumpfs On the Psychol@cal OriPin of Spacc

Perceotion and Meinong's Hume-snidies 1. Meinong's essay appeared in 1877, four yean

&er Stumpfs treatment of space perception. Meinong's study - his Habilitationschrift, which was sponsored by Brentano - had a more modest aim, namely to contribute to the history of British empiricism. Both books discussed opinions of the major British ernpiricists and both dealt with the problem of abstraction. Meinong defined abstraction "as a mental act through which one or more presentations are obtained from a larger presentation complex"

(p. 12). Furthemore, he noted that if such a presentation complex is not there, then one cannot have abstraction. One way of rnaking this more concrete is imagining a coloured ball.

According to Stumpf, the bal1 with al1 its aspects, its size, the colours, and so forth, is represented in us. From that representation the colour is 'taken,' as well as the size, and so on. There is a multi-dimensional presentation of the bail (size, colour, etc.) f?om which cenain dimensions are abstracted. Abstraction meant for Stumpf isolation of cenain aspects. The% abstraction is only possible when there is something to be abstracted, when there are more dimensions present, when there exists a multi-dimensional presentation. In that, Meinong and Stumpf agreed. Langfeld ( 1937) said about his tacher's course ihat:

Stumpf explaineci that an abstraction takes place when something is

separated in thought which is not separated in the phenomenon (p. 49).

Stumpf did not discuss abstraction as if I separate one pile of money into huo piles in thought. Instead, he noticed that we cannot see the size of the pile without its (silver or copper) colour. But we can talk about its size independently fiom its colour, and that can corne, according to Stump, only from further analysis. Meinong did not mention Stumpf in his thesis. He believed, however. dong with

Stumpf, that one cannot imagine colour without extension, although every space has the potentiai to be differently coloured. Meinong believed the same was tme for qudity and intensity. There is no quality, like colour or tone, without intensity. But, of course, every colour and every tone can have many intensities. Still, there is not one without the other. In that sense they are inseparable; they can only be isolated from each other in abstract t hought .

Now Stumpfs criticism of the British empiricists in Oof

Space Perception becornes understandable. In associationism, colour and space are united in consciousness through association., that is, through a process taking place in the observer.

StumpÇ to the contrary, held that colour and space essentidly belong together. that they are always present together. Not an active process in the observer, but an objective. factual, existing relation arnong stimuli is responsible for the unity of space and colour. A fused content, like a coloured surface was a unitary phenomenon and not bound together as the

British empiricists held. At the end of his life (Theow of Knowledgg, see, for instance. p.

1 1) Stumpf pohted out that the loudness of a sound and the sound's pitch are not connected with each other just habitualIy, meaning just through association Stumpf did not believe that originally we perceive the loudness of a sound without its pitch, or its pitch without the loudness, and that those properties of sound become co~ectedin the course of tirne. He believed that we perceive loudness and pitch together from the onset.

This inseparableness of the essaniai properties (attributes) is, of course, not

founded on gradua1 habitualisation, so that originally we perceived indeed volume of sound without pitch or sound pitch fk from intensity, and both

properties becarne comected with each other in our presentation just

habitually, instead, from the onset it [this inseparableness] lies in the nature

of impressions, how they are already innate in every person. (Theory of

Knowledge, p. 22/23)

This understanding was aiso a topic of Stumpfs course. Let us look again at what

Langfeld ( 193 7) recollected:

Stumpfin his positive arguments for a nativistic theory laid weight upon the

fact that the optic nerve gives us colors with the attribute of extension as a

unined impression. He thought that this fact aione makes an empirical theory

impossible. He added that it is not synthesis which is gradually leamed but

rat her analy sis (p. 46/47).

Stumpf s nativistic idea how coloun and space belong together was thus anathema to the ideas of the British empincists.

The consequenca of Stumpf s view for a theory of leaniing will be discussed later.

Let us first continue with his recognition that colour and extension are partial contents, that these, by themselves, are incornplete. This signifies, according to Mulligan and Smith

(1988), a break with atornistic psychology. And those scholars give as an argument:

It implies that the simplistic notion of atornicity, derived as it was from the

wrpusailar theories of t he Newtonian era, cannot serve within psychology

as an adequate basis even for the treatrnent of sirn~lesensations (p. 143). Stumpfs students would complete this break.

Stumpf thought the psychological presentation to be made up of dl the tones, colours, and so forth, present at a certain moment. The presentation as such is an immediate or un-mediate expenence, but it carmot be descnbed adequately. Only the dimensions can be named and the values objects have on them. Stumpf thus agreed with Brentano that introspection is in principle impossible. Brentano used as an argument that we cannot observe Our anger in the same way as we can observe a rose. Stumpf thought that one cannot be conscious of the presentation as such. Instead one is only aware of that which is abstracted fiom the presentation.

The presentation reflects, according to Stumpf, the continuous flux of the world. It lies before the psychological analysis; the psychological analysis is carried out on it. That analysis @vesus nameable results, such as colour or space. To appreciate this analysis, we have to tum to Aristotlets differentiation between perception and cognition. Perception brings us in contact with changeable qualities; in cognition we penetrate beyond the appearances to that which rernains unchanged. In Stumpf s analysis, a presentation is always changing. Although he did not say it expticitly, it seems that the psychological analysis gives a more stable outcorne. Perception is then a process of abstraction, a drawing out from the concrete presentation to the kinds or classes, like space, colour, tone pitch, loudness, and so fonh He accepted the Aristotelian outlook that we amive from changing appearances to stable concepts through psychological means. That, of course, does not mean that he accepted everything taught by Aristotle.

Stumpf was silent on whether his process of abstraction is carried out by the perceptive or the cognitive systern. The gist of his argument, however. points to the perceptive system. Ka red baii is shown to an infant, NO things cm happen. (1 ) The infant has leamed to discriminate for colour and wiil notice the red, or (2) the infant hasn't learned it and still ignores the colour axis.

We saw that Brentano mentioned two aspects of mental phenomena. They have ( 1 ) a mental inexistence and (2) an aboutness. Brentano distinguished between a presentation, for instance a mental representation of a tone, and mental acts, such as hearing a tone or being conscious of hearing a tone. Observe, however, that a note, or a sight, and also hearing, or seeing, are mental phenomena. Brentano stressed that mental events such as hearing and seeing are conscious. Stumpf, on the other hand, paid lip service to the aboutness; his focus lay on the mental inexistence of an object.

Someone who believes that objects have mental inexistence will amve easily at a belief in the existence of psychological unconscious States. At this moment we are not conscious of a flower we admired yesterday. Now, remembering that flower, it becomes

(again) a mental object. The flower must have been nored in us and since we were not conscious of it d the time, it mua have been kept unconsciously. Hence the assurnption of the existence of mental objects may lead to the assurnption of a psychologicai

'preconsciousness.' That idea was even more likely, suice the notion of unconscious was aiready widespread at that tirne. In passing, the prophet of the unconscious, Sigmund Freud, attendeci Brentano's lechires and visited Brentano once or twice at home (see Freud, 1990).

The conclusion that the storage of the flower is psychological, however, can certainly be challenged. The memory can very well be a physiological thing that happens in the brain. Stumpf s 'preconsciousness,' however, is not a repository of past experiences. and neither is it an agent that determines what can or cannot become conscious. Stumpfs preconsciousness is a step in the process of perception. It is the result of physiological processes and it is the foundation for the mental realm.

Stumpf was silent about the physical realrn in his On the Psycholo&cal Onein of

ace Perception. That book deds exclusively with mental phenomena, which shows the influence of Brentano. A presentation is a mental phenornenon, and there is no speculation on the existence of a counterpart of a presentation in the physical world. In his next major work, Tone Psychol~,Stumpf would say a linle about brain processes. Sensory stimuli lead to brain events and those evoke mental events, or presentations. A sensory stimulus is then represented in the brain and mentally. Why does Stumpf need presentations? We do not know his reasons, but perhaps the following is a plausible explmation. Stumpf had learned From Brentano that physiological phenomena - after dl, brains are physical- have no intentionahty. Perhaps Shimpf felt that a mental presentation is needed in order to explain the phenornenon of uitemionaiity. However, this is neither obvious nor necessary. Granted, even if intentionality is grounded on a presentation of the concemed object, it does not follow that the presentation is mental, and that is what Stumpf Md. A presentation on its own does not have intentionaiity and can therefore very well be a physical event. Perceiving, thinking, feehg and so forth have intentionality; but why are their underlying presentations not just physiological? Keep in mind that we need a physiological apparatus, including physiological representations, in order to experience. Smpfhad even less reason to assume the existence of a mental representation, since he recognized that the mental is fed by the physical and that the physical influences the mental: "The mental and the physical are in continuous exchange" (Theory of Knowledw; p. 822).

Another explanation could be that Stumpf did not believe that the brain is capable of integrating information into a complex whole. A presentation is such a complex. In it, many different kinds of information are integrated. That whole gets unravelled by a prychological analysis. But even then, the question still remains why the mental needs its own presentation and why it cannot use the one present in the brain. In any case, for whatever reasons, Stumpf developed Brentano's teachings of the mental inexistence into a dudistic view of presentation.

The parts are the result of an analysis of a whole and therefore they are secondary to that whole. The presentation, the psychological whole, must exist prior to its anaiysis and pior to the results of that analysis, hence pior to the parts. The unity of the stimulus. the whole thus, is grasped by the observer. The active part of perception is not the creation of a whole, siice that is already given. The active part is the abstraction of the elements from that whole. In Stumpf s view one abstracts, for instance, loudness and pitch from the tone.

Therefore, the whole cannot be a synthesis of the parts, instead it is that which grounds the parts. This idea has, of course, a gestalt ring to it. However, Stumpf did not equate presentation with gestalt.

One ddask ifthis priority of the whole applies across senses as weU as within. as when we see and hear somebody talking. And if this is so, cm we then dispense with

Aristotle's sense that is common or with his concept of sensoiy integration? 1 think that

Sturnpfs answer would be that sight and sound are detachable dimensions; the one cm occur without the other. Their integration is not laid upon the organisrn by the outside

world, and that seems to argue for an integration taking place within the observer.

Stumpf seemed to anticipate Wendel1 Garner's work. Garner (1974) noted that

stimuli have structure and that they are evaluated according to dimensions such as colour and size.

Structure is mily part ofthe physical world and cm be defined independently

of the perceiver of it. Any concepts related to it are ais0 part of the real

world. Thus for me to talk about a dimensional structure is to talk about a

property of the real world which does not depend on the processing of the

real world (p. 18 1).

Garner used experimental means to detect whether subjects could ignore ceriain dimensions while carrying out tasks in which other dimensions are involved. Stumpf and Garner seemed to agree. We abstract dimensions such as pitch and loudness fiom a tone, or colour and space fiom a sight.

A paradox of history is that later in life, as a result of reflecting on the theories of his pupiis, Stumpf wodd readdy criticize the Berlin gestalt school with an argument similar to the one he used agallist the British ernpiricias. Then he wodd argue that the elements and th& relations are presented together; that one is given with the other, not, as his students advocated, that one (the gestalt) is Qiven prior to the other (its elements). A gestalt, amrding to StumpS is the relation between the elements. A gestalt is then perceived with the elements grounded. However, let us now return to Stumpfs understanding that space and colour belong

intimately together. On the basis of this insight Stumpf would argue that Anstotle's concept of substance is unnecessary. A tone has a certain pitch and a certain loudness. You cannot

have a pitch without a loudness and vice versa, even though Stumpf noted that one may Vary

one aspect more or les independently from the other. Stumpf concluded from the fact that certain aspects always CO-occurthat qualities are not always grounded independently fiom one another in a substratum. ûtherwise, we could have one without the other. That, however, is not the case; both are always present together. Both form an inseparable pan of a whole. ûther qualities, such as white and sweet, can be perceived independently from each other. But in those cases there is neither a hidden bearer that carries white and sweet.

They are just parts of sugar.

Stumpf wrote in his autobiography that early in his career he gave up the notion of substance (p. 396). Stumpf could not agree with Aristotle's solution for the problem of identity, namely substance. In he wrote about the idea of substance:

The concept of the unitary bearer of the properties can really be shown in

perception: the whole is after al1 bearer of ali the properties, which we think

united therein... . Indeed, an "object" consists only of its propenies.. . but

these are not a bundle, but a whole (p. 28).

One page later we find: The sentence, "Every property has to be rooted in a substance".. . proves

itself self-evident or a priori. We express it ... as follows: "Every property

should belong to a whole of properties." (p. 29)

There is thus not a substance grounding the qualities. But the qualities are linked to each other They al1 form part of a whole. The example of a coin could help to clan@ this idea. Its sides are not grounded in a substance; they are simply part of that whole. There is then no ndto postdate a bearer of the sides. A linle fùnher in Theory of Knowiedqg we find:

[Substance] is for Aristotle not the whole of the thing, but rather its essence

as opposed to its attributes. This "essence" is, however, nothing else than the

whole of its essential properties, as they are indicated in the definition (p.

36).

The coin is not opposed to its sides; the coin is nothing other than its sides and the rest of its existentid properties, such as its metal. Nor is Freddie the dog opposed to his hair, colours, legs and so forth. Freddie is the whole of his essential properties. Stumpf wrote a page later:

One can consider these main properties, according to which we classi@ and

name the sensations, surely as the perceptual foundation of the concept of

substance, if one wants to understand by that the essence of a thing (p. 3 7). The qualities can be perceived: we see the sides and the metal of a coin the haïr and

size of Freddie. They form the perceptud basis of Our notion of a coin, or of Freddie.

The concept of a something outside the properties (conditions, activities)

that would be added to these and somehow formed the basis of al1 this

appears in al1 cases dispensable... (p. 40).

Stumpf thus thought that Anstotle's idea of substance is superfluous and that as a

consequence there was no need to postdate a bearer of a thing.

Stumpf would not, I think, deny that psychological processes lead to an

identification of essences. The essence of a coin is not its sides and not even the metal it's

made of A coin is money and we can buy goods with it. Remember that Stumpf wrote that

we see a glove, not the sense impressions leading to that expenence. To corne back to the

example of Freddie, his essence is not his hair or size; Freddie is a nice pet. In Aristotelean

psychology, the essence is not given in perception. One arrives at it through psychological

mechanisms working on perceptuaI qualities. Stumpf agreed with this Aristotelean position.

Psychological phenornena are caused by sense stimuli reaching the sense organs

which leads to a process in the brain from which somehow a psychological whole emerges.

What then in the stimuli grounds the holistic character of the experience? Ehrenfels' answer

was something which is extra to the elements: the gestalt qualities. Stumpf could not accept

that extra. He would argue that qualities are presented together and that we do not need to

suppose an underlying entity, like a substance; or an additionai element, iike a gestalt quality . Sturnpf s students would take over his criticism of non-physical notions like substance.

1 do not believe that Stumpfs analysis hoids up, and 1 think that he confused the mental and the physical domains. Anstotle held that the mental domain does not need substances. to which Ehrenfels added that gestalt qualities will do. But does that mean that the concept of substance is superfluous for the physical domain? Aristotle pointed out that the sensory experience of an object varies over time, but that the object keeps its identity.

To explain this phenornenon, one needs to resort, Aristotle argued, to something that underlies the change. If Stumpf did not like to consider that substance as not matenai, or opposed to the physical appearances, then why did he not see it as a physical bearer of propenies? Still, I think, he could not escape the notion of an underlying pnnciple of identity .

Stumpfs view on presentation also has a bearing on learning theory. Stumpf argued - not unreasonabiy - that we have to learn to anaiyze the presentation. Leamkg, in his eyes, is a graduai process whereby one differentiates more and more sharply the aspects of the presentation. He thus postulateci a development fiom a weak to a better discrimination of the parts. Perhaps the following epigram would not be unfair to Stumpf s position: We lem not by association but by dissociation! And that dong pre- existing dimensions of variation.

Stumpf did iiot elaborate on this view. His premises, however, seern to indicate a view of learning which runs opposire to mainstrearn psychology. This may help to explain why he is now a forgotten psychologist (Spning et al, 1986). Stumpf left it at that. His insight, however. was worked out by one of his students, namely Kurt Koffka. It is instructive to see how Kama developed Stumpfs thought and how Koffka, in his tuni, influenced an

Amencan psychologist. Therefore let us now tum Our attention for a moment to these successors of Stumpf

Koffka ( 1924) believed that infants learn colours gradually. First they leam to differentiate between colour and non-colour. Then they are ready to notice the differences between colours, fim between warm and cold colours, and thereafier within the two groups.

Only after having reached that stage can the child learn the narnes of colours.

Koflka offered spatial position as another example of the graduai process of leaming.

The orientation of figures is irrelevant to children. It does not matter to them if sornething is upside-down, or turned on its side. That means that they have to learn to attend to the features of position; "to a cMd a form is much more independent of its absolute spatial positio~than it is to adults" (p. 293). The position of features becomes important only der one has realized their importance. Thus,

for adults the absolute orientation of the figure is a very powerfùl factor.

Right and lefi, above and below, become characteristic properties of the

Merent members of the configuration, and consequentiy of the total-fom.

(p. 293).

Eleanor Giison's undentandhg of learning was in many respects similar to Stumpf s intuition. She attended Smith College fist as an undergraduate, and then to obtain her master's degree. During her student years Kun Kama taught a course on gestalt psychology, which she took. Koffka also ran a seminar at Smith and Gibson attended that as well. Gibson ( 199 1) remembered that seminar as follows:

When Iwas a graduate student, I regularly attended Koffka's seminar as

moçt of the Smith psychology faculty did. It was the local forum for debate,

despite Koffka's rather authontarian style. 1 was never much attracted to

gestalt psychology, as 1 have ever been uneasy with a phenomenologicd

approach (p. 6).

She also wrote that in Smith College:

The real excitement came in my senior year when 1 had a year-long course

in advanced experimental psychology, with James Gibson, a young assistant

professor fiom Princeton (p. 6).

We will see how James Gibson was influenced by Koflka later on, now we will retum to

Eleanor Gibson. Perhaps she was never attracted to the gestalt school. Nonetheless, Gibson

( 1969, p. 26) recognised sirriilantes between her ideas on perceptual development and those of Koffka.

In 1955 Eleanor and James Gibson published together an article on perceptual learning. 1 wiii corne back to their authorship afier having diswsed the article itself Therein they asked: '7s perception a creative process or is it a discriminative process? 1s leaming a matter of enrichhg previously meagre sensations or is it a matter of differentiating previously vague impressions?' 34). And their merwas: 'The observer sees and hears more. or infers more, but this may not be because he images more, or infers more. or assumes more, but because he discriminates more. He is more sensitive to the variables of the stimulus array" (p. 40). They aiso noted that "it is not novel, of course, to suggest that perceptuai development is a matter of dxerentiation. As phenomenal description this was asserted by some of the genalt psychologists, notabiy Kama and Lewin" (p. 34). It is interesting to note that James and not Eleanor was the first author of the 1955 article, but it was Eleanor who later would defend their view.

In 199 1 she would write that:

The uitimate goai of perception... is differentiation, the reducing of

uncertainty; or to put in another way, the extraction of distinguishing

features and invariants from a stimulus flux where information is very great

(p. 361).

Perception is in this view a "search for critical features and invariants" (p. 360). And what is lemed "is the distinctive features of objects and the invariance features of events over time" (p. 360).

It is not clear to a baby which features, which invariants are relevant. Growing up, the child learns to abstract and filter the nght features and invariants, to attend, so to speak, to the right dimensions.

Percephial leaming goes on because in the beginning ody gross imprecise

information is obtained. Development proceeds Ma differentiation of

information that specifies things and events in the world, by discovery of

invariants as changes in stimulation are produced by movements of things and of the observer, and by encounters with novel and broader environments

( 199 1, p. 289).

Or, as she formulated it in 1969:

Children (and animals) lem to attend to distinctive features of things, to

invariants that lead to perceptual constancy and permanence, and to higher

order structures and niles (p. 445).

Children have to lem to perceive those features that teii them about the structure of the environment.

She agreed with the gestalt psychologists that the "features are not constnicted by the mind but are discovered by the perceiver" ( 1969, p. 82). There is a Line from Stumpf to

Kottka to Gibson. Al1 three believed that we learn gradually to attend to the right features and dissociate ourselves fiorn the irrelevant ones.

Tone Psvcholoqy

Mental life, in StumpCs view, starts with a presentation. The function of the psychological analysis is to distinguish, to abstract the parts present in the complex presentation. That is an active process taking place in the observer. Stumpf used the word . . 'part' in a technical sense in Psvchological Onan of Space Perceptiori, namely as a dimension on which objects realize values. He would continue to do that in his next book,

-. -. There, it seems, he was reluctant to use the word 'part' to describe two or more notes within the "part" sound. It is only in his later work that part would acquire its regular meaning, that is an element belonging to a whole.

Stumpfs experimental work was devoted to musical and speech sounds. His main experimental work and perhaps his magnum opus is Tone Psychofoa. The first volume appeared in 1883; the second in 1890, the year of Ehrenfels' publication. fhe last one is. like

Stumpf s first book, dedicated to Brentano.

The book is titled Tone "because it intends to describe the mental fictions, which are evoked by tones" (Tone Psycholoq. vol 1, p. V). Tone Ps~chologis thus Sturnpf s attempt to establish the scientific study of mental phenornena, especially the ones caused by physical tones.

Tone Psvchology is a vastly different work from PsychoIonical Origin of Space

Perception. The change of scale, from perceptual space to single tones, is striking. That, however, is not the only change. Stumpf moved, as he wrote in his autobiography ( 1924)

"frorn theory to observation, from meditation to facts, fiom my writing desk to the laboratory" (p. 397).

From what has been said, the modern reader would expect to find descriptions of experiments as well as conclusions based on them. The first volume, however, does not mention a single one, and ody far into the second volume do we find reports of experiments. That is not as surprishg as it might seem at first. The mental realm is open for ber perception - to be precise, Stumpfs inner perception. But this raises a question, would

Sturnpfs inner perception match those of other persons? The only way to find that out is by empirical means, to ask (other) subjects to report their auditory perceptions. Now. StumpPs modest experience in scientific laboratones came to fniit. "ln order to control my own judgement, I have taken yet another route.. ." (Tone

Psycholog. vol II, p. 142), narnely the route of experiments. Of course, no one has direct access to the inner perceptions of othen. Those perceptions cannot be observed externally.

The best one can do is ask subjects about their experiences. The experimenter can increase the reliability and usefunes of those repons by ( 1 ) controlling the environment in which the perceptions are evoked, and (2) standardizing the reports of those perceptions. That is exactly what Stumpf did. The experiments were thus canied out to get an idea of the imer perception of musically naive subjects. We saw that Stumpf wrote in his autobiography t hat he started to occupy himself with facts and laboratory. Facts are the contents of imer perceptions. And laboratory is not restricted to some speciaily-equipped room in a university; it is any setting to elicit reports of i~erperceptions.

Shrrn~fs scientific Method

Before we continue with the findings published in -q, let us look more closely at Stumpf s scientific method. Stumpf used that method for the experiments described in Tone Psycholpgy, but he explained it in his next book, The Speech Sound.

There, he defines a subject as an "observer, whose task is the description of a perceived appearance" (p. 49).

Stumpf learned the methods of the natural sciences dunng his stay in Gottingen. He adopted those techniques for the use of psychologists. Then he used his method in his own experiments and ais0 taught it to his students (see Ash, 1982). -4s a student of Brentano. Stumpf was well aware of the difference between the mental and the physical sciences, between a biological experiment and a psychological one

(see Macnmara, 1993). Brentano advocated a scientific study of psychological processes that respecteci the intemal appreciation of reality. The intemal perception, obviously, is not grasped by the methods of the other sciences. That forces psychological experiments to be different nom them. In the experiments done by other scientists, intuition of the object studied does not play a role. Intuition, however, is essentiai in psychological experiments.

Stumpf explained the difference between the two types of sciences by contrasting his subjects with Frogs used in physiological experiments. The contrat is, of course, that the task of Stumpfs subjects was to descnbe their experience, while the biologist is not interested at al1 in the frog's experience. Remember that Stumpf defined his subjects as observers!

Brentano reaiized that the methods of the sciences that study physical objects were not suited to the sîudy of mental phenomena. But he did not spell out how the study of mental phenomena could be carrîed out. Stumpf accomplished that task and he deserves credit for developing a research method that is both sound and gives intuition its rightfbl place. Brentano was one of the German scholars who started the transition of psychology from philosophy to an independent science and Stumpf was in the Brentano circle responsible for the transition penod. So Stumpf, dong with G.E. Müller, Wilhelm Wundt and perhaps a few more researchers was one of the founders of scientific psychology.

Stumpf combined a philosophical education with a scientific training and, under his authority, so would his students. We have to appreciate this. A training in laboratory techniques takes time and can therefore be detrimental to a wider development. In his research for Tone Psvchology, Stumpf would carry out microscopic investigations and let his students take over his experimental method, narnely the use of self-observation by a few well-trained subjects. Wertheimer ( 19 18) acknowledged this indebtness in a speech to

Sturnpf on the occasion of Stumpf s 70th birthday:

For a long time now the center from which mind streams fonh has been

science; and not the general words that are taught, not what someone

preaches is essential here, but how one to work, how, in what spirit the

sciemist works. in grasping particuiar problems as well, has iduence in al1

directions. We, too, have been led into the garden from which this (spirit)

streams forth in the midst of the living; and it is not a matter of indifference

in what spirit apprentices are initiated. The style is passed on (quote cited in

Ash, 1982; p. 62).

Hence, Shimpfs siudents were technicdy well-prepared. Like Stumpf, they studied the irnmediate given, the psychologicai reality of tone or space. Stumpfs students also showed their broad education in treating a wide range of subjects. Wertheimer also acknowledged t hat in the aforementioned birthday spsh:

As much as you love and support work in specialized science, you have

nonetheless taught us to keep Our gaze directed to larger questions of

pnnciple, to work toward the fhitful cooperation of psychology and the

theory ofknowledge, with the highest problem of philosophy in view. None gf us wished to be locked up in the workroom of specialized

science...(Wertheimer, 19 1 8; quote cited in Ash, 1 982; p. 62)

In the naturd sciences, facts are unproblematic, in the sense t hat t hey are undisputed since they can be observed by everyone, at least in principle. This is not the case in Stumpf s psychology. Two listeners can hear the metone, but they have no access to each othef s hearing expenence. That hearing expenence, not the tone, is the psychologicai fact.

Therefore psychological faas must be proven as facts; t hey are not objective by t hernselves.

Stumpfs experimental method aimed to establish psychological phenomena as facts, not to prove or falsie a theory When one deals with mental phenomena, one has to be sure that one talks about reai phenomena.

We find an echo of this fact-finding mission of psychology in Koffka's (1935)

Principles of Gestalt P

Nobody can reproach psychology with having discovered too few facts. A

psychologist who knew dl the facts that have been brought to light by

expenmentd methods would indeed know much, very much. And such

knowledge is today regarded as an aim in its own right. "Find facts, facts,

and again facts; when you are sure of your facts try to build theories. But

your facts are more important." This slogan expresses the creed of a

philosophy which is widely accepted today (p. 4).

A few sentences fùrther Kama wrote: With the advent of experiment more and more facts were discovered which

played havoc with the old theories. Only when psychology determined to

becorne a fact-finding science did it begin to becorne a real science (p. 5).

Koflka then continued by saying that:

Science is not comparable to a catalogue in which al1 the facts are listed

according to an arbitrary principle, ke the books in a library in the

alphabetical order of their authon; science is ratiod; the facts and the order

are one and the sarne; facts without order do not exist... (p. 6).

KofTka, afier ail one of the leading gestalt psychologists, pointed out that facts get their meaning only by king embedded in a general hework, in a theory. In Koffka's eyes, facts are meaningfùl if they are part of a theory; the proper place of elements is within a whole.

I think Koffka was right here; a collection of facts is no science, science also requires an understanding of facts.

Stumpf gave psychologists a task to establish Eicts through the experirnental method.

Having done that, the researcher can think of a possible explanation. Kotfka's (1935)

Pnnciples of Psvcholqgy descnbes many experiments proving certain psychological facts.

Facts, which, according to KofEka, can ody be explained with the use of the gestalt principles.

Wertheimer published in 19 12 the results of experiments on stroboscopie vision or apparent movement. When two visual objects in close proximity are very bnefly shown, one immediately afler the other, one does not perceive two objects, but instead perceives movement. Thus, two stationary light sources flashing at given intervals give the impression of movement between locations. In this famous publication Wertheimer du, aimed at establishing facts. His paper describes experiments and their resuits. but the reader hardly finds anything about Wertheimer's theory. The phenomenon of apparent movement was known before Wertheimer's research. His publication however, re-established it experirnentally as an undoübtable fact of mental life. The emphasis of the establishment of tàcts cm explain why a theory is so noticable lacking in his paper. More likely, however, is that in 19 12 Wertheimer and his fiends had not fully grasped their gestalt notion, but that would happen soon. The first paper containing an indication of the thoughts of that school was by Gabriele von Wartensleben (1 9 14) (see Harper et al, 1985). Koffka's ( 19 15) paper gave a more authorative account of that school (see Seaman, 1984).

It is iiiustrative to look at another quote nom Wertheimer's speech on Stumpf s 70th birthday:

There are researchers who attack nature as though it were an enemy, set

traps for ha, take the off've, tq to bring her down like an opponent, like

cooi technicians or sports people who want to feel their own strength,

achieve success. ûthers are like travellers who write arnusing feuilletons;

others have their card boxes, write thick books in which everything swims

confusedly together; still others work busily on a certain small piece, have

real C.N.V.F. (concentric narrowing of the visual field exists in fields other

than the optical), are blind to the right and to the lefi. How different you are! For you the facts are not objects of attack, nothing

that should bring flashing results. For you the facts are as though they were

in a fathex's hand. In Africa there is a custom in one tribe: when one wishes

to show trust to a guest, the mother lays her nursling in his arms and says,

hold the child. So do you hold the facts in your hands, and so have you

taught us: devotion to the real.. . (Wertheimer ( 19 18), quoted by Ash ( 1982)

p. 62).

Fritz Heider ( 1970) conunented on the above quote:

One can understand these words in different ways, but it is quite likely that

Wenheimer was taiking here about the free unbiased description of

experience, the khd of phenomenological description which we meet in his

paper (p. 13 5).

The passage, however, becomes meaningful when one keeps in rnind that Stumpf taught his student s to establish psychologicai facts t hrough the experimentai rnethod, no t, as Heider wrote, through phenornenological description. His reverence for facts was taken over by Stumpfs students. The cornparison Wertheimer made is colourfùl, but aiso to the point. The psychological expenence is unique to the listener; others have no direct access to it. It is in a sense like a nursling that belongs to its mother and has a unique relation with her. In this Afncan tribe the mother gave her nursling to the guest to hold. Stumpfs experirnental method was devoted to letting his experirnental subjects give their psychological experience to him to hold. and so was the work of his students.

Tone Fusion

Volume 11 of Tone Psycholog

deals with the questions: How does Our consciousness [Urteile] react

towards several simultaneous notes, to these related to intervals, and, finally,

to short tone-sensations (p. VI).

Sturnpf was interested in the perception of notes that exist at the sarne time, thus in the mental result of the simultaneous presentation of two or more notes. In this experimental set-up one can ask questions such as how much will the presence of one note influence the perception of anothd How important is the relation between the notes for their individual as weii as conjoint perception? Stumpf thus explored in an experimental way the effects of context on perception.

In order to explain the subject's answer - whatever it may be - Stumpf had to know what a note is. In this he drew heavily on the work of that great 19th-century Geman scientist, Hermann von Heimholtz. Helmholtz ( 1862) gave a plausible account of hearing tones in Die Lehre von den Tonempfinduneen Gr die Theorie der Musik. Helmholtz based his own research on the works of the French mathematician Jean Fourier ( 1768- 1830) and the German physicist Georg Ohm (1789-1854). Ohm reasoned that noise must be an irregular, non periodic movernent of air, whereas notes should be transmitted by a periodic, or regular movement. Ohm based his conclusion on the undentanding that it is impossible to analyze sometfiing irregular. Regular patterns, on the other hand, can be decomposed and hence simpler patterns retneved fiom them. From this, Ohm concluded that noise must be irregular and therefore unanalyzable, whereas it is possible to obtain meaningful units of information From tones, because tones are believed to be periodic air patterns. Helmholtz saw that Ohm's insight was supported by the findings of Fourier. Fourier proved that ( 1) any penodic wave is a sumrnation of simple sine waves, and (2) there eBsts only one solution, or one specific summation, for each periodic wave. That made Ohm's physical theory mathematically possible. Helrnhoitz then had to show that sounds are indeed airwaves. He dernonstrated this by making sounds by bringing air into motion. This led to the invention of the siren, an instrument that artificially produces airwaves. Helmholtz had now proven that one can make sounds through air movement. His next step was to demonstrate that one can detect a simple wave within a cornplex one. Helmholtz pointed to

Nning forks and piano strings. They are selectively responsive to specific airwaves and thus capable of analyzing ahaves. Finally, Helmholtz had to indicate at which spot in the body the decornposition of the air wave into sine waves takes place, and he pointed correctly to the cocldea.

Helmholtz thought hearllig to be a process which analyzes the incoming energy into simple, elementary units. Stumpf would accept that idea, but he was well aware that we experience a unitary tone. Therefore, the decomposition of the airwave in the ear must be followed by an integration at a more central level.

One of Stumpf s main experimental findings was tone fusion. When one presents a subject with two notes - even when each one (made by a naturai instniment) is played to a different ear - the subject hem in certain cases just one note. Snimpf named this 'fusion,' because the notes are fùsed into one. He aiso found that there are aeps in the degree to which fusion takes place. or to which the subject gets the impression that he hem one note instead of two. The fusion is strongest when the notes differ by an octave, and weakest when an discordant combination is presented. Stumpf defined fusion as follows:

We cal1 fusion that relation between two contents, especially contents of

sensations through which it does not fom a mere sum, but a whole (Tone

psvchology. vol 11, p. 128).

Stumpf located the cause of the fusion in the brain. Each note on its own has a certain effect in the brain. Fusion occurs because those effects interact in such a way that a fusion is experienced in the psychological realrn. Stumpfs experimental work showed that the fusion has to take place centrdly. A peripheral explmation must be excluded because fusion also occurs when each note is presented to a difFerent ear. Stumpf argued that two auditoiy stimuli cause different nerve fomtions to be active, and these two activities corne together in the brain. Fusion, however, is a psychological phenornenon and therefore must be distinguished from the physical ones. Remember that, according to Brentano, psychological phenomena have characteristics that physical ones lack (namely they have intentionality, directness toward an object). That implies that we cannot identify the psychologicai phenornenon of fusion with a certain brain state. In Stumpf s view, t hat brain state is the basis from which that particular psychological phenornenon arises. Tied to a particular brain state is a sensation, in this case that of a fùsed tone.

Later. Stumpfs students would use the term 'isomorphic relation' to express their

idea of the role of the brain for the perception of gestalten. They would argue that the

perception of a gestalt is the result of an underlying gestalt-like pattern in the brain cells.

Their idea thus clearly had roots in the psychophysiological notions of Stumpf, namely the coming together, or patterning, of nervous activity in the brain, which results in the experience of a single tone.

Kohler (1947) explained that idea of isomorphic relations as follows: First he rejected the view as entirely wrong that

"Gestalten," i. e. segregated entities, exist outside the organism and simply

extend or project themselves into the nervous system (p. 1 60).

Then he explained his view of the modality of vision. We receive light waves from objects. in that information, however, "no trace is l& of the units which actually exist in the physical world" (p. 16 1). Obviously this is an exaggeration, but it makes it very clear what he meant, namely that one has to reconstmct the physical world fiom the visual information reaching the retins- That can be done thanks to the way the serwry system is organized. or the visual neurons are comected; "this organization tends to have results which agree with the entities of the physical world" (p. 163).

Kohler rejected the view that segregated entities exist outside and sirnply extend or project themselves inside. This passage, however, should not be understood to imply that there are no segregated entities outside, nor that we do not have a good image of the outside world. His objection is aimed at the notion that the projection is done simplr- Seeing is then not simply a projection from the outside to the inside. Instead it is a reconstruction carried out by the neurological organization of the visual system. Something is lost during the transmission in the vidrays and this is restored in the organism. John Sullivan ( 1968) pointed out that "Kohler's notion of isomorphism is fùndarnentaily a physiological account of inexistence" (p. 260). Indeed, Kohler taught that the outside gestalt exists in, has its image - to use Aristotle's term - in the observer.

If we go back to Stumpc the neurological organization of the auditory system leads in cenain cases p the hearing of one füsed tone, instead of the two presented ones. Stumpf was aiready aware that the organization in the brain leads to the perception of structure. His midents worked that insight out in an elegant theory. Stumpf s ideas were speculative, not supported by neuro-psychological research, and neither is the theory of his students. It should therefore not be a surprise that Murray (1994) noted:

Recent developments in sensory neuropsychology have made the

neurological theorizing by Kohler and the other gestalt psychologists

obsolete (p. 5).

Stumpfs work on fusion shows a rernarkable resemblance to the experimental work of the Berlin gestalt schwl. The presence of one item influences the experience of another.

We have one note, say 4 and we present it together with a series of different notes, in such a way, however, that at each presentation of A will only be accompanied by just one other note. The note A. according to Snunpfs acperimentai wok will be perceived very distinctly

£kom the other one if they form a discordant combination. The notes will be perceived less diainctly. or more tùsed. the more their relation approaches that of an octave. .And Stumpf

found that A will be perceived as fused with the other one when they are an octave apan.

Wertheimer ( 1925, p. 12), defending the gestalt phenornenon, used an argument

strikingly similar to Stumpf S. He said that one can prove esperimentally that how one

perceives two colours. namely as one or two. depends on the ensemble in which the colours

are embedded. One can evoke ftom the sarne patched stimuli something homogeneous or two different colours with just a small alteration. Observe that Wertheimer's account of colour is also a good description of Stump's work on fusion if we replace, of course,

'colour' with 'note.' Sturnpf s students would thus continue his way of experimenting, but they wouid propose a different theory to account for the results, and Stumpf could not have agreed with their expianation.

The unitary psychologicai phenornena be it çome separate notes or a tùsed note. has cenain characteristics. We can ask ourselves questions such as, what is the intensity of a sound cornplex, or its pitch?

Simultaneous notes, when they are recognized as several appear to us not

as a mere sup but as a totality of notes. From this it directly follows that

there is also an inclination to ascribe to this totality, per se, a tone-pitch,

although stnctly speaking one can only attribute a pitch to every individual

note. (Tone Ps~chol~gy.vol II, p. 383) We cm play several notes at the same moment. Those notes will form a unity. As a unity those notes have certain charactenstics. We can Say that this unity, these notes is played too loudly, althwgh no one will deny that certain individual notes are played softly.

Nonetheless, the totality of the tones, this whole set of notes, has certain characteristics. in

Our example loudness. Thus, according to Stumpf, one can talk about the loudness of these tones as a totality, although they are not fused and in that sense do not form a unit.

We already find in Tone Psvcholoq an awareness of the problem of the relation between the totality, a whole, and its elements. But we have to wait for Theory of

Knowledgg to find a thorough analysis of and a solution to the problem. Of course, the analysis in this last book is a reaction of the master to the work of his own students.

Rubin's Criticism

We saw that Stumpf claimed that colour and space. loudness of a sound and its pitch, the pitch of a sound and its intensity are inseparable and that this co~ectionis imate in every person. Stumpf held that conviction when he wrote his book on space perception and he reiterated it in his last one. Stumpf had no experimental evidence to support his opinion. This is not surprising for his first book - after dl it was written prior to his experirnentai work. But even later in We, when Stumpf had carried out expenments to prove the existence of psychological facts. he did not deem necessary, or perhaps he overlooked to camy out experiments to prove the existence of this fact. Rubin concluded from his experimental work that we can perceive lines without width. Obviously, by their nature, it is impossible for these lines to exist. 'lonetheless, the experience can be evoked. Rubin believed that one has to separate the experience fiom physical reality. Rubin doubted the ciaim (se. p. 176)considered clearly evident by its proponents, that colour and extension beiong together. Rubin's point was that what is necessq physically or frorn the common sense point of view rnay not be necessa- psychologically. Therefore Rubin did not want to exciude the possibility that we can experience one without the other, although such a separation is physically impossible. That is an empirical matter and it has to be settled by empirical means.

Rubin did not mention Stumpf when he doubted Stumpfs claim. In a footnote. however, he referred to Edmund Husserl, another farnous student of Stumpf There is even a good chance that Husserl got his idea fiom Stumpf Anyway, Rubin's book was published in German in 192 1, when Stumpf had developed his expenmental method. Stumpt's own framework - that one should establish psychological facts through expenmental rneans - lads to the question, aiggested by Rubin, about the experimental support for the supposed fact that we motperceive colour and extension separately fiom each other. It seems that

Stumpf was not justified in leaving his opinion without experimental support. Having said this, let us now look at expenments Stumpf camied out dunng the World War 1.

Speech Sound

Sturnpfs Speech Souna published in 1926, followed Tone Psychology. Its aim is to the analysis of the different elements of the speech sounds themselves and

the conditions leading to their perception (p. 3).

Again. Stumpf was interested in the components that make up the experience of a sound.

But now. he was not interested in notes but in speech sounds. We have seen that Stumpt's earlier work on sounds drew on Helmholtz's research, which, in tuni, was based on Ohm and Fourier. In Speech Sound Stumpf would also recognize their importance to his work.

Helmholtz taught that the experience of hearing a note is caused by a process in the cochlea that decomposes air waves into simple sine waves. Those sine waves cannot be decomposeci further and are, therefore, the basic elements of acoustic sensations. Stumpf s student Kohler, however, did not agree with Helmholtz's thesis and Stumpf mentioned

Kohler's criticism (Speech Sound, p. 3 7 113 72).

Helmholtz thought that the ear decomposes the incoming energy. That would imply that nigher up' in the braiq those basic elements must be combined to form the bais of the unified experience of one sound. StumpPs treatment of Kohler's work seemed to indicate. however, that Kohler believed that such a sumation in the brain could not account for important phenornena of our hearing experience. Facts such as the interval berween tones, the specific sensation of a chord, the quality of a speech sound cmot be comprehended, according to Kohler, £tom a theory that assumes summation of basic elernents. Stumpf believed that Kohler was completely nght in his assertion.

In Speech Sound (p. 2771278) Stumpf offered an analysis of the whole. When

Stumpf wrote his book the word 'gestalt' was already used by his students. Stumpf started by saying that a complex exists where a whole of 'absolute' contents is present. What did Stumpf mean by that? He indicated with 'absolute' that somethîng can stand on its own. that it is not in need of a wpporting framework. A tone and colour are obvious examples. They are the elementary units, the basic elements, from which forms - to use another word for whole - are made. One cmot have a visual form without cotour, a melody without notes.

These elementq units are presented as having a certain relation toward one another. They form a propomonal whole. A melody is, of course, our prime exarnple of such a whole. This proportion or relation can be accidental, in which case we are unable to install a change without changing the experience. Or the relation between the elements can be fixed. Then one can transfonn the individual elements without changing the experience, as in the case of a melody. As long as the relations between the elernents remain constant, the gestalt will remain the sarne. The fact that a melody is dependent on the relations between the notes and not on their absolute values makes it possible to play it in different musical keys.

Transformations are possible, because relations are not tied to the absolute values of the elements. Ehrenfels' influence in this description is, of course, obvious.

in some gestalten we perceive the individual elements, as in a melody. In others we are not conscious of the elements. Ehrenfels gave as an example a chord, and Stumpf mentioned fused notes.

Before continuhg with Stumpc 1 would keto compare hirn with James Gibson. We already heard - firom Eleanor Gibson - that his fkst academic position was at Smith College, where he was a junior colleague of Ko&. Gibson ( 1971) acknowledged in an article titled

The mcy of KoWs PMicipk his admiration for and debt to Koffka. Gibson mentioned

Stumpf in none of his works and it is therefore uniikely that he had read Stumpf. Nonetheless, Gibson's ( 1979) theory was squarely in line with Stumpf s work.

One of Gibson's central concept was &ordance. "The affordances of the environment are what it ~ffers,what it provides or hrnishes for good or ill" (p. 127).

Gibson gave as an example an apple; an apple is eatable to us. it provides food and we see t hat directly.

We find the first indication of Gibmn's notion in an article he published wit h Eleanor

Gibson in 1955. (1 rnentFned that article when discussing Eleanor's work on perceptual leanùng). They ( 195 5) wrote: "Let us assume tentativeiy t hat the stimulus input contains within us everything that the percept has. What if'the flux of stimulus at receptors does yield al1 the information anyone needs about the environment" (p. 34).

Gibson ( 1979) also said that

Afliordances are properties of things faken with reference to an observer but

not properties of the pxperiences of the observer (p. 137).

.And a few pages further we find:

The affordances of things for an observer are specified in stimulus

information. They to be perceived directly because they a perceived

directly (p. 140).

Gibson would agree with Brentano that perception is characterized by intentionality, that there is an aboutness in perception. Brentano and Stumpf, however, argued that intentionality implies a mental representation of the perceived object. Gibson, on the other hanci, would dispense with the representation; one simply perceives directly the edibility of the apple. The article of 1955 contained a remark that is here of interest: "Repetition or practice is necessary for the development of the improved percept, but there is no proof that it incorporates mernories" (p. 40). Thus the Gibsons formulated in 1955 the view that an apple's edibility is not based on earlier expenences with apples, which lett, for instance. memory traces that would add knowledge of the edibility to the sight of an apple. Instead. the Gibsons believe the one sees edibility directly. that means without assistance of earlier expenences. E'cperience, in Gibsons' ide% would only help to focus better on the relevant feature of the stimulus array.

Gibson ( 1979) said about affordance

"It" is an invariant combination of variables, and one might guess that it is

easier to perceive such an invariant than it is to perceive al1 the variables

separately (p. l Ml3 5).

An affordance is invariant and that means "the afTordance of something does not chan."

(p. 139)' which implies "the afZordance, king invariant, is always there to be perceived" (p.

139). In other words, the affordance presupposes "veridical perception" (1955. p. 40).

Let us compare this with Stumpf s theory. Stumpf held that since we detect the relation between tones, we are capable of extracting invariants - to use Gibson's tem (p.

14 1). Giison worked in vision. His elements "are those of surfaces and edges" ( 19 50, p. 8).

But we are able to perceive "the order or pattern of the retinal image" (1950, p. 9).That order "can be considerd a stimulus" ( 1950, p. 9). Stumpf would also hoid that we perceive directly the relation between notes, and he beiieved as wel1 that this relation is a basis for hrther analysis. We saw that Kohler's concept of isomorphism was a development of Stumpf's notion of psychological analysis. Note that Stumpf and KoMer's idea contrasts with

Gibson's view of direct perception.

Sturn~fsDualism Versus his Students' Monism

Brentano differentiated the mental tiom the physical. Stumpf accepted that division and came to the conclusion that there are two systems working closely together, two separate parts of one reality (see Theory of knowleds, p. 822). Stumpf "did not believe in psycho-physical pardlelism, but in interaction" (Lewin7 1937. p. 192). Kohier came to believe there is an isornorphic relation between the mental and the physical (see laeger,

1992); thus he seemed to believe also in the distinction between the mental and the physical.

This is, however, not tme. Heider (1970) wrote that the Berliners were fundamentally monist (p. 136). We have corne here to a difference between Stumpf and his students that deserves careful analysis. KofRa's ( 1935) Principles of Gestalt Psvchology is especially revealing .

Koma started his discussion on monism on page 28, where he distinguished the geographical from the behavioural environment, or, in other words, the physical from the mental world. Here Ko& seexneci, like Stumpc to accept a dualistic point of view. On page

56 he remarked that Wertheimer brought the two realms "as closely together as possible"

Whatever the basis of bis remark about Wertheimer, that still holds the two worlds apart.

Koffka's next step was an appeal to Kohler's theory of isomorphisrn "a term irnplying equdity of form" (p. 61). Henle (1987) described isomorphisrn in a review of Koffka's book as

The hypothesis of a sameness of structure between rnolar physiological

processes in the brain and stmctured phenomenal facts (p. 16).

We see gestalten [stmctured phenomenal facts] because gestalten are forrned in the brain

[molar physiological processes]. Now Koffka had amved at a point where he could hold that the two worlds "are essentiaiiy alike" (p. 6 1 ). If two entities are essentially alike and we know one of the- then we know the other as well.

Koflka must have had something like this in rnind, because he believed that

"observation of the behavioural world reveals to us properties of p hysiological processes"

(p. 6 1 ). If by knowing one entity, one knows the other as well, then one of the entities ceases to be important. One is enough and Koffka chose the physical world. He stated that

"our theory . .. is a purely physiological theory, even though mental facts, facts of direct experience, are used in its construction" (p. 64).

Koffka's last step was to combine the two worlds into one, for which he uses the term 'psychophysical field.' Behaviour - we see in the use of this word the influence of

Amencan psychology at the time the book was written - is then seen "as an event in the psychophysical field" (p. 67). We should, however, not equate Koffka's psychophysical field with Stumpf's two separate parts of one reality. The word 'psychophysical' eady leads to that interpretation, but that goes against KoBa's remark that his was a purely physiological theory. Henle (1987) gave the following definition of the psychophysical field: "the brain field corresponding to psychological events" (p. 17). This definition is in line with Koffka's train of thought, where he reduced the phenomenal events ro molar brain processes. Then the term 'psychophysical' field is misleading, since Kofia meant only physical field. hyway. whatever we tMof KoEka's arguments and definitions, he started with Stumpf s duaiistic position and then tried to convince us to abandon this notion for a monistic position, in fact a materialistic Mew of mind. The Berlin school would thus not take over Stumpf's dualistic and interactionistic position, but would argue for an intesrated field, for a monistic position.

We saw that Brentano concluded that since the physical and the mental are different. the scientific methods to mdy them must be different too. In Kohler's ( 1947) treatment on the psychologicai method, we also see a move away fiom the position of 19th-century

Gerrnan psychology to a monistic stance. Kohler started wÏth a recognition that methods can differ arnong sciences:

A method is good if it is adapted to the subject matter; and it is bad if it

lacks regard for this material, or if it misdirects research. Hence, what has

been shown to be an excellent procedure in one science. or for some

problems, may be aitogether useless or even a hindrance in another science

or for other problems (p. 37).

He continueci the^ but not by recognizing that the memal sciences differ fundamentally from the physical ones and that therefore different research methods are warranted. Instead, he said that psychology is a much younger science than the natural ones and that, in its present state, psychology was not ready for the quantitative, indirect methods of the natural sciences. In this view psychology is not fundarnentally different fiorn the natural sciences and one day psychology should be able to use the methods of today's natural sciences.

Ifwe wish to unitate the physicai sciences, we mua not imitate them in their hi-ghiy developed contemporary fom. Rather. we must irnitate them in their

histoncal youth. when their state of developrnent was comparable to our

own at the present time (p. 42).

There is. according to Kohler, nothing wrong with psychology irnitating the natural sciences. as Ions as the present situation of the mental sciences is taken into account.

However. Kohier recognized in the same book that "Aristotle wrote his te.xtbook of psycholog+' (p. 1 38). In other words, he acknowledged that psychology is an old science.

That admission nihilates. of course, Kohlef s argument, unless one is willing to interpret the word 'young' in terms of stages of development rather than chronological age. The latest stage is then that psychology is a monistic and matenalistic science like the others. But even then the question remains, is a science of more than two thousand years really at an early stage of its development?

Stumpf and his Students

Brentano's speculations could easily have resulted in a negation of extemal reality.

Stumpfdid not fdinto this trap. Brentano's influence, however, is felt in the interest in the intuitively given, the psychological reality of tones and space. We cm contrat this with behaviounsm, which threw psychological intuition deliberately out of the field of psychology .

We saw that Stumpf rejected associationism because it was blind to structural regularities in appearances. Stumpfs students ais0 rejected 19th-century associationism. They would become world renowned for formulating the gestalt laws. laws dealing with regularities in appearances.

The Berlui gestalt school applied Stumpf s experimental method, the use of intuition by a few weli-trained subjects in an experimental sett ing. Wundt is senerally considered the founding father of experimental psyc hology . The Berlin SC hool. however. leamed the trade from Stumpf

Both Stumpf and his students were aware that brain States must be related to psychological phenornena. All showed an openness towards physiological theorking.

The above is clearly not enough to explain the rise of the Berlin gestalt school.

S~mpfsstudents had to deal with a weakness of StumpE Kun Lewin ( 1937) provided an insight when he ended his article on Stumpf as follows:

One can find in Stumpf s works al1 the outstanding characteristics of the

pioneering epoch of experimental psychology in Germany : the nearness to

the philosophicai ongin as background; the will to establish a variety of facts

as nch and as well proven as possible (this was the center which kept things

together); a marked predominance of perception psychology; a more or less

open disappointment in this iimitation; and a longing for the deeper problems

ofpsychology, the experimentai approach to which seemed so utterly out of

reach. To my knowledge Stumpf has never shared the view of deep

pessimism for the outlook of psychology which was noticeable among some

of the younger psychologists at t hat time. He was rather optimistic for the

final outcorne, perhaps because he himself contributed so much to its progress. (p. 194)

Kama ( 1935) pointed to the unease that Stumpfs students had with psychology during their student years:

.4 colleague of mine with whom 1 was going home asked me the question:

"Have you any idea where the psychology we are leaming is leadin%us?" 1

had no answer to that question, and my colleague, after taking his doctor's

degree, gave up psychology as a profession and is today a well known

author. But I was less honest and less capable, and so I stuck to my job.

But...

his question never ceased to trouble me ... (p. 53; dso quoted in Ash. p. 56).

Ash beiieved that the unnameci author is Robert Musil. Lewin's and Koffka's quotes indicate a dissatisfaction with Sturnpfs psychology arnong his students. Koffka's rernark also indicates that it troubled him. The dissatisfaction thus drove him to look for something more inspiring.

In the works examined to this point, Stumpf was indeed not a successful 'salesman.' as he failed to communicate his main thoughts clearly. His basic ideas are hidden in those books; they are oniy implicitly present. This has the obvious disadvantage that the reader has no easy access to them. Not being clearly fomulated, his theory is not explained and defended, much less venfied. Furthmore, an implicit idea cannot easily arouse enthusiasm on the part of the reader. That could explain why Sturnpf s work seerns detached and lacking in passion. Passion and enthusiasm, by the way, are qualities of which there are plenty in Brentano's writings.

Stumpfs students would not oppose him. The Berlin gestalt school cannot be understood in terms of revolt by a younger generation against their eiders (nor for that matter against another school). On the contrary, Kohler and his colleagues would show respect for Stumpf. But the students rnissed something exciting and inspinng in Stumpf s works. Stumpf would not give his students an ideai to [ive for. Hence, it was up to them to develop an intelligible, coherent and inspiring program; and it is one slogan, one word that captures ail: Gestalt.

We saw that in The S~eechSound the elderly professor mentioned and criticises some of the ideas of his pupils. This book thus gives an indication of his views on the Berlin yestalt school. The treatment was sympathetic. but rather short and perhaps in need of hrther clarification. Clearly. Stumpf did not exercise his analytical talents on the theones of his -dents. However, he did so in his last book the Theory of Knowledgg ( 193911940).

There he started his discussion on the Berlin gestalt school with the statement that he saw "a gestalt as a whole of relations" (p. 229) and that he considered "transposability as its main property" (p. 229).

A gestalt... can be preserved when, during the changes of ail absolute partial

contents, proportions (relations) of a certain type between them are

preserved (transference, transposition). If it is offered in this changed material. it can be recognized as the same. (p. 230)

In other words, Stumpf accepted Ehrenfels' criterion of transposability as defining a ystalt. One cmdeduce tom the quotes that Stumpf did not accept Ehrenfels' solution of a gestalt quality. Mead, Sturnpf assumed that we perceive the relations between elernents directly. This idea built on his derwork. We saw that fusion takes place when two notes stand in a specific relation to each other. Hence the relation between the elements is detected.

The complex way Stumpf thought that incorning energy is analyzed is striking. He acknowledges two processes. In one, elernents and in his later work aiso their relations are detected. The other analysis is orthogonal to the first one. Values or measurements taken dong axes are abmacted from those elements and figures. There is thus not only a process of obtaining single elements and figures, but aiso an analysis in which those elements and figures are being appreciated. We detect a gestalt and the gestalt gets a value on the loudness dimension: this piece of music is too loud. Or we hear a tone and that tone is - on the pitch dimension - too high.

Being a student of Brentano, Stumpf was well aware that at the psychological leve!

Our experiences are holistic. He rejected Arinotle's notion of substance and also Ehrenfels's gestalt quality. Stumpf was a more modem scholar. He saw brain states influencing psychological states. Hence the holistic phenornenon of the psychological level must be connected somehow to atornic events at the physiological level. There were thus opposing forces puhg at Stumpf, the psychological one calling for unity, and the physiological one for disseting. Snimpfs way out was a compromise. Both are perceived. The whole as well

as the elements are perceived and further analyzed. Stumpf concluded that the perceptual

systems perceive not only the elements, but aiso their proponions or relations. In other

word, Stumpf believed that while hearing a series of tones, we also perceive the

mathematical regularity t hat exists between the tones. Stumpf believed t hat those relations

have to be the sarne each tirne in order to expenence that gestalt. We thus perceive the rule

that unites the elements into a whole. The information conceming elements and conceming

their relations is then brought up to the psychological level. The presentatio~then contains

information regarding elements and their relations. In this view. the psychological analysis

is canied out on the elements and on their relations. For instance, tone pitch and loudness are abstracted from single tones, but also from a melody. Stumpf had found a solution to

the problem posed in Tone Psvcholo . How is it that we can speak of a pitch of a single

tone, but also of a set of tones?

-4ccording to Stumpf, the experience of a gestalt stems from a cornplex. made up of elements that stand in a certain relation to each other. These elements are simple tones,

coloun, and so fonh; the basic units of which a melody, painting, or whatever is made up.

Stumpf noticed that there cannot be a gestalt without basic elements. Even a highly

repetitious melody like Ravel's Bdero contains many tones. It is however, not necessary

that to be able to expenence every individual element as such. Remember, Stumpf found that in a fused tone one cannot hear the two individual tones which are presented.

In Theory of Knowledgg, Stumpf formulated his cnticism of the Berlin school as

follows: [The newer gestalt psychology].. . has fallen from one biased view into

another in the works of the gestalt theonas it appears oflen as if there were

no colours and tones anymore, nor any absolute in Our sense perceptions,

only rnere gestalten. These themselves appear as the original and only given,

from which can be extracted the tones, colours and everything that we

usualiy name sensations, only through abstraction (p. 242).

The same point is repeated a few pages later as follows:

Stria gestalt theorists stress that with gestalten we are not dealing with mere

relations or even relation wholes. Gestalten must be something prirnarily

existing that do not need a foundation, as is the case with relations.

Relations - one says - are something abstract, gestalten something vivid.. . .

indeed the single so-called sensations are just products of division from the

solely originally given gestalten.

We cannot endorse this view of gestalten. Whenevn there is a gestalt in the

sense of a content in consciousness, distinguishable parts are there, then a

structured sensation complex is present, then relations between the

distinguishing parts, which are perceived together, exist as well to a certain

degree. Gestalten are never perceived, per se, as their own foundation, like

colour or tones, but are grasped only with a specific material. Iust as one

cannot perceive a tone gestalt without the contnbuting tones, one cannot

perceive the pitch of a single tone without the tone itself But just as the pitch of the tones is secondary, the product of an abstraction, so too is the

tone gestalt. It is merely perceived together with [the Foundation] (p. 246).

Stumpf s cnticisrn is strong and Our example of a melody can make them better understandable. The Berlin school claimed that we first hear the melody and through it the individual notes. Stumpf. arguing against this view. pointed out that we don? need to hear a melody in order to hear individuai notes. We don't need a gestalt to perceive its elements.

Wendel1 Gamer (1962) dealt with the same issue, although from a slightly different perspective. He did not refer to Stumpf s criticism and it is indeed unlikely that he knew of it. It is therefore noteworthy that both scholars came up with the same criticism against the

Berlin gestaltias. Garner would, however, propose a different resolution from the problem than Stumpf S. Let us look what Garner wrote:

I am disagreeing with classicai gestalt theory, and I feel that the major fading

of such theory has been its assumption that it can deal adequately with the

single stimulus pattern (p. 342).

Garner ( 1 974) wrote about his own experiments:

Structure is what I manipulate in my experiments, and 1 arrange al1 these

things long before bringing an experimental subject into the process. When

the subject is brought in, it is for the express purpose of finding out what

consequences my manipulations of structure have on the subject's

performance (p. 18 1 ). This is nor a bad description of Stumpfs expenments either. The influence of Sturnpf s students on Gamer is apparent where Garner remarks that it is the set of ail stimuli that determines which dimensions are relevant for the experimental task at hand. The whole gives value to its parts.

But how is it then possible that ive evaluate one singe stimulus? Mer dl. the major failing of the Berlin school was its incapacity to deal with the perception ofjust one element!

Garner ( 1974) argued

That any single stimulus leads to an inferreci subset and total set of stimuli,

and that the single stimulus then has structure by being a member of these

different sets and subsets (p. 97).

If necessary, the subjects create From the individual stimulus a set of stimuli. On the basis of some perceived characteristics of the stimulus, the subjects try to figure out to what set this stimulus belongs. The whole is then implicitly present in or can be inferred from its members. Whatever the tmth of this intuition, at least it provides an answer to Stumpfs pertinent questions. I do not think it is a good answer. There is too much sophistry to my liking, and an inferred whole stands too fw away from what 1 think a whole is. Nonetheless, f do not want to disregard Garner's intuition altogether. In order to understand the sentence

'John kissed Mary,' one has to apply a whole background of knowledge to the stimulus. One will not understand ihis sentence if one does not bear upon it knowledge about noun phrases, in order to determine who is doing what. The background of knowledp must be present to experience one element. Sathat background cannot be equated with a total set of stimuli. In this case, part of the background must be innate, because it can never forrn part of a set of stimuli*

It is odd that the Berlin school created such a problem of the one- eiement stimulus. k~easy

way out would have been to consider a single dimdus as a gestalt. One tone played b y itself

has qualities such as offset, onset, duration, and so forth. It is a coherent entity of experience

and is that not the criterion for calling something a gestalt?

Let us now retum to Sturnpf We saw that he assumed that when one perceives

severai elements, one notices their relation too. Hence one perceives the gestalt at the same

rime as one perceives the individual elements. The elements and their relation are given jointly in perception; it is not one first and then the other, but both together at the same

ttme.

Stumpfs aîtack of the Berlin school went further. He said that a gestalt is abstract.

That notion follows from (Ehrenfels') criterion of transposability. In order to detect that

something is transposable. we have to process the immediately given. We have to abstract

from the immediately given that which can be transposed, namely the relation between the

elements.

A rnelody is not at ail the totality of these successive momentary heard

tones, but that which is preserved when we transpose them. And it is really

sornething abstract. To arrive at tones, and colours through abstractions

from tone and colour gestalten, means just to define something concrete

through abstraction from an abstraction (p. 247).

This, of course, is foolish indeed from a cornmon sense point of view. Remember, however, Rubin's waming that expenence does not always follow the dictates of physics or common sense.

A gestalt is thus an abstraction and it is of course impossible to derive something concrete out of an abstraction. As a consequence, a gestalt cannot be that From whch the elements are obtauied. Stumpf thought it absurd that a relation - something abstract - could be the prirnary given, and that &om this relation the elements - concrete things - are derived.

Smith ( 1988) formulated Stumpf s understanding as follows. According to Stumpf

There is something cornitive in our awareness of that specific structure

which is a gestalt. For to grasp a gestalt is to grasp not merely an individual

as such but also that abstract net of transferabte relations which is its essence

(p. 24).

Stumpf also formulated his criticism of the Berlin gestalt school in question format:

Why are gestalten transferable? By which basic property cm this be

explained? Further, why must each gestalt form the basis of a structured

cornplex, why doesn't it make sense to describe contents of consciousness

as gestalten which are in our view simple (even though the corresponding

physical or physiological state of flairs migh be put together)? In other

words, how and in which way does the unity of a gestalt distinguish itself

from the unity of an impression, in which we cm distinguish absolutely no

(independent) parts (eg. a single tone), if not through the diversity of its

elements, which become or are combined in the gestalt? Finally, why in -cenerd do gestalten require a foundation, why are there no gestalten without

several elements? .And why are the gestalt laws so different for the different

senses. for the sense of colour as opposed to the sense of tone. if. after dl,

gestalten are the prirnary and concrete, but tones and colours the secondary

and abaract? On the bais of the newer gestalt view al1 these questions seem

to us insoluble (p. 253).

To the best of our knowledge, the Berlin gestalt school never responded to these questions. The questions were hidden in a book of several hundred pages. a book which seems to have had no impact. Nonetheless, the questions are well fomulated and they go to the very hart of the Beriin gestait t heory . It is no sign of strength that Stumpf s students ignored them.

We have now anived at the explicit point of disagreement between Stumpf and his students: the relation between e1ements and the whole. In the 1 890s the gestalt appeared as an independent entity. dongside the elements. We saw that Stumpf proposed that elements and figures are processed in parallel. Elements and the whole enjoy an equal status. The

Berlin gestalt school had already taken a different stance. In the eyes of the Berlin gestalt psychologists, there exkt fht and forernost wholes that determine their parts. In that view, the elements have ceased to have an independent existence alongside the whole, but are subjugated to the whole. Wertheirnefs fmous experiment with two altemating lights proved to him that one sees one movernent under certain conditions, not the two elements that give rise to it. The mental representation of the altemating lights was thus not atornic. Stumpfs students also held the tiew that something has to exist prior to its parts. What we are conscious of the parts, are grounded in something that is analyzed. For Stumpf, that ground is the presentatio~containing information about the elements as well as their mathematical, or calculable relation. His students had a monistic stance. They rejected the psychological phenornenon of presentatioq and it seems that they repiaced it with the gestalt. In their eyes. the gestalt grounds the elements and the elements are analyzed from it. They Ieamed that loudness and pitch, colour and space Ire abstracted frorn a presentation and rhey would teach that elements are singled out fiom a gestalt. Therefore the elements have the same relation to the gestalt as loudness and pitch have to tone, one of subordination. Stumpf described this idea as absurd.

In Theo- of Knowled~,Stumpf also seemed also to indicate the preconditions of a gestalt expenence.

The view of gestalt perception that is defended here presupposes: (1) the

possibiiity of perception of proponions, (2) the possibility of imperceptible

contents of consciousness, (3) the differentiation between appearances and

mental t'unctions as heterogeneous data of consciousness on which is based

the possibility of imperceptible contents of consciousness (p. 254).

This is a rather cryptic statement. A possible explanation is the foilowing. Number 1 means that with the presence of individual elements, such as colours or notes, the relation between t hose elements is given as well.

Number 2 could be aimed at a Mew expressed by Kohler in 19 13. Kohler's article helps us to understand Stumpfs point. In the beginning of his article, Kohier wrote: where a mass of stimuli of the sarne modality occurs. and it is not possible

to find everywhere in direct experience those sensations that would be

expected fiom the previously established relation between single stimulus

and sindg sensation (p. 14).

Kohler could have in mind the experience of a concert. One does not hear al1 the tones; some remain umoticed, even though their stimuli reach the ex. It seerns that Stumpf argued t hat. nonet heless, the tone stimuli evoke sensations in the concert-goer. and t hat t hose sensations are caused by those umoticed tones, which "have psychological reality in the same way. except that they are unnoticed" (Kohier, 1913, p. 1411 5). At stake here is the relation between stimulus and psychological reality. Stumpf believed that there is a simple, direct relation between them, and in order to rave that theory, Stumpf assumed the existence of imperceptible contents of conscious. KohIer disagreed, aibeit his view was still nascent in 19 13. In his later works, he clearly expressed that the experience evoked by a stimulus depends on the configuration in which the stimulus is embedded. There is no one-to-one relation between stimulus and experience according to Kohler. Instead, the psychological reahty of the stimulus depends on the gestalt. There is another possible criticism of Stumpf s position. What is a content of consciousness if it is imperceptible? This seemingly cuntradictory statement strengthens my conviction that Stumpf would have been better off to put his presentation at the physiologicai level.

And number 3 in Stumpfs statement could mean that there exists a difference between the presentation on the one hand and the apparatus that analyzes it on the other.

That is, a difference between the appearances that are analyzed and the psychic functions that do the analyzing. Again. Kohler ( 19 13) can enhance our understanding. Kohler explained Sturnpfs view - with which Kohier disagreed - as follows:

Judgements which follow a sensation or perception are in theory to be

distinguished sharply from these. But they ofien deceive us so badly about

the sensations actuaily present that we believe we have experienced different

ones (p. 15).

Sturnpf, it seems, held that our apparatus of analysis can err in its examination of the presentation, or, if one finds the word err too harsh, does not appraise everything of the presentation. Certain data remain unanalyzed and therefore umoticed.

StumpPs last book, Theory of Knowledp, published derhis death, gives the best account of his theory. It clarifies his thinking and therefore indicates what the gestalt psychologists were taught. That book, however, also contains an intelligent criticism of the

Berlin gestalt schwl. It is unforninate that Wertheimer, Kohler and KofEka did not react to

Stumpfs criticisrn that, fier & is based on a life devoted to thinking and experimenting on the relation between pan and whole. The mdy and incorporation of that criticism into their own work would have strengthened the gestalt theory. Even though the Berlin gestaltists saw psycholog in a materialistic way, they applied Stumpf s experimental method-The gestalt psychologists asked their subjects to report their inner perception and their method was initiated by Brentano and developed by

Stumpf The gestaltists are interested in the immediately given. the expenence, the mental reality of perception. The repons of Wertheimer's experirnent indicate perception of movement. Hence. the question had to be answered: How is it that we do not see the elements before we are aware of a quality such as movement? How is it that in the stroboscopic experiment. we have the illusion of movement? Wertheimer's answer seemed to be forced upon him by the facts: One sees first a whole and then, when the conditions are right, one derives the parts from that whole. In the stroboscopic experiment one does not çee the parts - two lights at two locations. Lnstead one sees movernent and that is what one reports.

In a speech before the Kant society, Wertheimer ( 1925) said

A melody does not arise as a secondary process fiom its constituents; on the

contrary, the perceptuai properties of each constituent depend upon the

whole (p. 1 1).

Here Lies the principal difference between Ehrenfels and Stumpf, against the Berlin school. As we saw, Stumpf was aware of the disagreement. However, it can be further spelled out. Before doing that, remember that Stumpf argued that the Berlin school failed to appreciate the role of the element.

E hredels explained the structure of our experience through eestalt qualities accompanying the sensory data. Stumpf taught that one perceives the relation with the elements. Ehrenfels and Stumpf, each in his own way, believed that one sees the parts and the whole together. although Ehrenfels ( 1890) acknowledged that the whole can ovenvhelm the element S.

The gestalt quality may sometimes force itseif into the foreground. Le. may

make dernands on Our attention to such an extent that it is difficult to resolve

its foundations into elements (p. 95).

In contrast, the later Berlin school held that a collection of sensory data does not have a gestalt; it is a gestalt, a whole, whose parts are themselves determined. A whole is then not a unification of its parts. but has its own laws and charactenstics and these cannot be understood as a functicn of its elements. Lnstead, the elements are just functions of the perceived wholes. The parts are themselves determined only as parts of the whole. Or, in other words, a collection of sensory data does not exist prior to the gestalt. On the contrary, it is the gestalt that determines the interpretation of its parts. Remember that Weriheimer stated that a melody does not arise from the tones, but that the psychological reality of tones, on the contrary, depends on the melody. The Berlin school put al1 the emphasis on the gestalt as causal &or and rejects any interesting existence for the parts as distinct from the role assigned to them by the gestalt.

According to the Berlin school, parts and wholes do not mutually determine each other. This seems odd. Let us take for example a sentence. .A word such as hand can mean several things depending on the sentence. It denotes something different in "May I shake your hand" than in "Give me a hand." In the first sentence hand means, of course, a specific body part, while in the second it stands for help. But what is a sentence afier dl? It is a structured combination of words. whose very structure is a function of the words that constitute it. So, the whole determines or constrains the constituents. And the constituents determine and constrain the whole. Neither is reducible to the other.

Eleanor Gibson ( 1969) had a sirnilar criticisrn on the gestalt school. She ( 1969) even considered that "the very notion of parts and wholes in perception is mistaken" (p. 447) She felt that there are two processes involved in perception. One is that of differentiating the relevant stmctures. or the parts. Perception of the whole is then the result of an integration of those parts. That is clearly a bottom up process. On the other hand, a top down process exins in perception as well. That follows from her theory of mental developrnent. Children perceive an object and have as a task to learn which features provide relevant information.

If two processes play a role, then one cm ask which is the most important. The answer depends, according to her, on the task at hand and to the stimulus information provided to the observer. Then she continued, %th [processes] may even occur at once" (p. 448). Then it is not just bottom up, nor ody top down; instead it is sometimes bottom up, sometimes top down, and sometimes both at the sarne time. The picture we get fiom her is this: The input of the perceptual system (or systerns) is cornbined into a unity. This unity is constrained by its inputs. But the unity also constrains the perceptual system(s).

Another cnticism of the Berlin school is that science is impossible within a strict holistic way of thinking. How do we study the structure of the whole? We study it by analyzing the parts. If we want to study the bloodstrearn, we have to study the hem. the lungs, and so fonh. However, the hnctioning of the heart is intelli@Ie ody within the whole of the bloodstrearn. But if you want to know about the blood system, you need to study its parts. In order to understand the whole. one has to study the pans. And in order to understand the parts. one has to know the whole. It is not one or the other. but both.

Kohler reacted to this criticism:

Some cntics maintain that gestalt psychologists repeat the word " whole"

continually. that they neglect the existence of pans and wholes, and that they

show no respect for the most useful tool of scientific procedure, which is

analysis. No statement could be more misleading (p. 168).

In our example, anaiysis could stand for the study of the hean and lungs to know more about the bloodstream. Kohler continued after saying that his critics were misleading. The

main tasks of gestalt psychology is that of indicating the genuine rather than

any fictitious parts of the whole. Al1 visual things are such genuine pans of

the Iields in which they occur, and mon things have again subordinate parts.

The very principles of organization refer to the segregation of such parts as

much as to their unitary character (p. l68/ 169).

Kohler did not address his cntics straightfonvardly. First he claimed that gestalt psychologists must differentiate genuine from fictitious parts. This sounds impressive, but it goes without saying. To appreciate Kohlef s statement, let us remember that scientific progress is only made by recogniting genuine parts. Of course, one needs the right pieces of the puzzle. Imaginary or fictitious ones exist only in the mind of the scientist. and there as a false notion. Then. Kohler rnoved away from scientific progress to perception, a very strange reply indeed to his opponents. But he referred there to the principle of segregation.

This remark is not very clear. The most likely interpretation. however. is that the whole takes care of the segregation of its parts. It seems that he reiterated his view concerning parts-whole. For, what is a genuine part? Just that, a part, not something that has meaning in its own nght and therefore worth studying. Kohler used the word 'misleading', but who is misleading whom?

The Berlin school ion sight of the dialectic relation be~eenparts and w hole. 1 agree with Gibson that both play a role. Then both have to be 'actors' and that implies that both have to be individuated. that one has an independent role to play vis a vis the other. Thus the individual tones individuate the melody, but the melody individuates the tones. The individual words individuate the sentence, but the sentence also individuates the words. The parts of the bloodsveam individuate the bloodstream but the bloodstream individuates the heart, lungs, and so forth. THE GRAZ SCHOOL

MEiNONG

Ehrenfels' On 'Gestalt Oualities' appeared in 1890. One year later rUexius Meinong discussed it in Zeitschrift fùr Psvcholoeie und Phvsioiogie des Sinnesor-. Meinong was a student of Brentano and Ehrenfels studied under him, as well as under Brentano. So.

Meinong discussed one of his students' works in a journal edited by another scholar belonging to the Brentano circle, namely Stumpf G.E. Müller was also an editor of this journal, but, since Ehrenfels' article concemed music, Stumpf was most likely its editor.

Meinong came from the Austrian lower aristmcy, who served mostly as state officiais and priests. David Lindenfeld (1 980) told how the bureaucratic delay of Meinong's promotion around 1880 led to an dienation from Brentano's circle and ultimately to an animosity towards Jewish midents and scientists. .i\lthough he was not politically active, Meinong was sympathetic to the moderate fàctions of the German nationalist movement. This could have caused the delay and bureaucratic secrecy surrounding his promotion. Whatever the case, this hurdle in Meinong's career resulted in pain, suspicion of conspiracy, and an unjustifiable anti-Sernitism.

Ehrenfels always used the word 'gestalt' together with 'quality,' as in "gestalt quality"

(disregardhg one or two exceptions). Meinong, on the other hand, used 'gestalt' on its own, without adding the word 'quality.' We will see why Meinong did not use that word.

Furthemore, Meinong made a shift from audition to vision. He restricted the word 'gestalt' to the visual modality. A gestalt is then a visual figure. something occupying space, a unie made up of spatial elements. It is the visual counterpan of a melody. a unity of tonal elements.

,Meinong wote: "1 know of a reader of the essay under review who is suspicious t hat outer mental realities lie behind the "gestalt qualities" (p. 153). It is difficult to know what

Meinong's reader meant by outer mental realities. Since that u~amedreader. however, cannot accept it, he or she disagreed with Ehrenfels' t heoq. Meinong could understand the reader's suspicion - he thought it wrong and a misinterpretation of Ehrenfels' term 'gestalt quality.' Meinong, however, did not give any reason why the suspicion is wrong. So. he does not say that Ehrenfels agreed wit h him. Neither did he provide other arguments, based, for instance, on Ehrenfels' text, to substantiate his position. Thus, Meinong gave no proper rehtation of a reading of the text contraq to his own interpretation. Note that Ehrenfels' article is weU wrinen, and the tem 'gestalt quality' could very well have been aptly chosen.

If the reader meant that form is a causal factor. then that interpretation is correct. Note that if one beiïeves that only tones have real existence, then gestalt quality can be considered as an outer mental reality .

Ehrenfels argued that we perceive a gestalt quality. The important word is perceive.

How one wants to interpret gestalt quality - an outer mental reality like Meinong's reader. or not, like Meinong - it is undeniably Ehrenfels' position that we perceive it. Meinong considered that position to be counter intuitive. He held that we produce mental presentations, although he did not deny that in some cases we produce the complex presentation more or less automatically from its foundation, such as a melody from tones. In order to appreciate Meinong better. it is wonhwhile to look at a distinction already made by Thomas Aquinas. Aquinas (ca. 1252), in his On Being and Essence, differentiated being from essence. Being, he said "has two meanings" (p. 30). It denotes

"something positive in reality" (p. 30), or "signifies the truth of a proposition" (p. 30). Note that in the latter case, the proposition does not need to reflect an extemal reality That

"affimiation is opposed to negation" (p. 30) is true, but we cannot point to this tmth as we can point to tables and chairs. Some things do exist in extemai reality. while other things exist, but not out there. not in the real world.

The other terrn in Aquinas' essay, essence, denotes the quiddity, the (Aristotelian) fonthe nature. "The essence is what is signified through the definition of a thing" (p. 3 I ).

If something does not need to exist in outer reality, then its essence surely needs not to have reai existence. Aquinas went even further writing:

Matter, then, cannot exist without form, but there cm be fom without

matter: Fom as such does not depend on matter. There is no necessity, then.

that the essence or quiddities of these substances be anyt hing else t han form

(p. 53).

And a few pages funher:

Every essence or quiddity cm be understood without knowing anything

about its king. 1 cm know, for instance, what a man or a phoenix is and still

be ignorant whether it has being in reality. From this it is clear that being is

other than essence or quiddity (p. 5 5).

In other words, essence has to be separated 6orn (material) existence. Meinong agreed wi t h this view. In a later work. On the Theory of Obiects, Meinong ( 1904) noted that one can talk about Santa Claus. the Golden Mountain, a round square, or the riot which did not take place. We know that Santa Claus Lives in the North Pole and brings presents to children; the mountain is made of gold; that square is round (and therefore cannot exist); and that the rebellion against the authorities did not take place. Like Aquinas and Brentano. Meinong held that objects of knowledge do not need to exist. He named them ideal objects. and he supporteci his notion funher by pointhg to the figures dealt with in geometry. These figures

- lines without width. perfect circles, ifinitely small points, and so on - do not exist.

Nonetheless. their properties can be known, which is a position Augustine had already taken

(see, for instance, Augustine's Confessions, ca. 395). Meinong added as his own contribution t hat "difference" between objects, or t heir " similarity" is non-existent as well.

DifTerence, or sirnilarity as such is not something concrete, out there. Smith (1988) said it as follows:

We cannot. Meinong argued, the 'difference' or the 'similarity' between

two colours in the sarne sense in which we can see the colours themselves.

Similarly we cannot higher order objects such as geometrical shapes,

velocities, distances. For such objects are && iike numbers and concepts,

that is they are outside space and time, and what is outside space and time

is not capable of being grasped in acts of sensation (p. 22).

Meinong's thwry is best explaineci in his (1904) On the Theory of O-. Meinong considerd it self-evident that one cannot have a mental presentation, a judgement, an ide* and so fonh on their own. There has to be something presenta judged. or thought. In other words. there has to be reference to an object. Meinong got that notion. of course. from

Brentano. Meinong tried to understand Brentano's characterization of the mental as having reference to an object. Especially, he wanted to Say more about that mental object.

Meinong's research led him to the object theory He formulated it in a book explicitly aimed at celebrating the 10th birthday of the psychological laboratory in Graz. Meinong was professor in Graz and founder of that laboratory. Theory, however, is not the right word.

Meinong proposed a new science. It was the science of objects, of objects as such; and its subject matter was that one can potentidly have an intentional relation towards any object.

It is the science of all possible mental representations, of dl potential or possible objects of knowledge. It is an "a-priory science which concerns everything which is given" (p. 106).

Quite rernarkable! Meinong launched a new science at an occasion intended to celebrate the progress of the psychological science.

Thus, we know and can talk about essences that cannot exist and therefore can also not be perceived through the senses. What is then the foundation of these thoughts9 How do we know them? Meinong's answer is that we produce them ourselves. They are not 'out there,' ready to be perceived by the senses. This leaves as the most obvious answer that we create them ourselves.

Let us now renim to Meinong's treatment of Ehrenfels' article. Meinong argued that a tone, or its counterpart in the Msual domain (a part in space) will result in a content in a person's consciousness. Meinong used the word 'cornplex' to refer to the mental presentation producd by the person. Certain contents together will more or less automatically create a cornplex in consciousness. That is the case when tones form a rnelody and (spatial) pans a figure. .Meinong noted that those complexes are often one of observation [Beachtung] or naming [Benennung]. Complexes that are not created so easily, and where more effon on the part of the person is needed, can oflen be expressed through a relation tem. .4n example of such a complex is a cornparison between two entities. There is. in Meinong's eyes. a gliding scale of gestalt complexes, frorn the ones that are very easily evoked to those that need more effort. Note that there is no room for Ehrenfels' gestalt quality in Meinong's theory. That notion is replaced by an effon within the person that works on the contents provided by outside stimuli and that creates from those contents complexes such as rnelodies. tigures, or comparisons. Meinong did not explain how complexes are created. It is. however. clear that it is through an unconscious process.

Meinong indicated that he was well aware that his view went against that of

Ehrenfels. He acknowledged that, according to Ehrenfels. the gestalt quality is given with the foundation. That cm only mean that the gestalt quaiity appears in consciousness without any effort. Meinong, in other words, said that Ehrdels did not propose a production theory of gestalt, since the gestait quality is not produced. This, of course, places his opposition to the view that Ehrenfels' gestalt qualities are outside the rnind on weak grounds.

Ehrenfels said that we perceive a gestalt quality. That means that we also perceive directly what belongs or does not belong to the whole. The gestalt quality belongs to the tones of the melody, not those coming from the background. Hence, we perceive which tones belong to the melody and which sound is background noise. Meinong rnust have disagreed. Since he believed that the complex is produced by the observer, he must also have held that the observer is actively involved in forming the boundaries of that complex.

Let us look at what the implications of "perceive" versus "produce" are for an explmation of the phenornena of similarity and dserence. E hrenfels explained t hem through the agency of a gestalt qudity. Similarity or diflerence is something that we perceive.

Remember. one recognizes two men as brothers because their gestalt qualities are sirnilar.

Meinong disagreed. He held that by comparing A with El, one produces a new entity (cdled an objective) that includes A and B. In that new presentation (the objective) A and B have a certain relation to each other, narnely of sameness or difference. Judgernent, according to

Meinong, is based on inner production and not through the perception of gestalt qualities.

That new. produced presentation has as ils foundation the presentations of the entities being compareci. Thus, by comparing NO entities a cornplex mental entity [Thatbestand] is created through which more basic entities stand in a certain relation to each other, and, of course, also to the complex.

Relations are one-sided dependent, namely on that which is related. A relation cannot exist without the entities that are related. One cannot have equality, or difference, without tbgs being qua1 or different. One cannot have a family without farnily members.

The other way around, naturally enough, is not tme. An entity cm exist without it having a padcular relation to othen. Lindenberg (1980) wrote: "Relations are dependent elements; fundarnents are independent" (p. 96). Meinong's notion contains a strong atomic element, which he shares with Brentano and Ehrenfels. Note that how the elements are related falIs under the realm of logic and not of causality. Production Theorv

We saw that Meinong used the word objective to single out a special type of ideal object. ".An objective is thereby treated as a complex of some kind, . . . the object belonging to it as a kind of cornponent" (p. 85). For instance. if blue is an object. then not blue is an objective. Not not blue is a funher objective and so on. When we have a mental object .4, the negation of not A, is as good a mental object as A. We can judge A to be true or false, the sarne for not A. We can become angry or happy about A and also about not A.

(The negation theory of La Palme Reyes et al ( 1994) is based on a sirnilar insight.) ..Another example is a unicom. If we start with unicorn, the next step could be to assen the being of a unicorn, a statement we judge to be fdse, while the not being of a unicom is true.

There is, theoretically, no end to the range of objectives, but there is a start: the object. The objective has no real existence. Still, it is an object of knowledge. We can. for instance, judge it to have being or to fail to have being. Indeed, as Meinong remarked. ".A mus be 'given' to me in some way or other if I am to grasp its non-being... " (p. 85). So, we can judge the sentence, 'After the election, there was no not,' to be true. Some objects cannot even exist, such as a square circle. We have seen that sameness and difference are also ideal objects. We can judge things to be the same or different; we can even measure their difference or perform other mathematical operations on them. That, still, does not make sameness or difference red objects.

"We now lcnow that those objects which exist, and even those which have being, nin far short of the sum-total of objects of knowledge" (p. 87). Meinong logically remarked that the potential objects of knowledee far exceed the real existent ones.

At this moment. one may wonder what this has to do with the human rnind, or with psychology That, 1 hope, will become clear when we discuss the works of Meinong's students Benussi and Witasek.

Since Meinong published his proposa1 to celebrate the 10th year of his own psychological laborato., he had to indicate what the task of psychology should be. One thing was clear. psychology could not deal with ail the potential objects of knowledge, since that was given to his new science. Meinong restncted psychology to occupy itself "only with real psychological events" (p. 9 1 ). In other words, the objects which are actually p-nt.

It seems that Meinong thought the science of objects would give the general framework within which the empirical science of psychology would carry out its research. It is, however, unWtely that the distinction holds between a science occupying itself with potential objects and one restricting itself to ernpirical facts. A modem viewpoint of this division is formulateci by the philosopher and psychologist John Macnarnara (in press). He noted that if it is an empirical fact that humans have access to al1 potential objects. than those objects fa11 under psychology's realm.

Besides, it is doubtfiil that this division would lead to progress in human knowledge. hgthe last forty years linguists and psychologists have made rernarkable progress - that during a tirne when their areas are blurred. Psychologists take into account the fact that hurnans can formulate an infinite number of utterances, while linguists look also at factual utterances. BENI !SSI

~Meinong'stheory is higNy abstract and fomulated in very general terms. Fritz

Heider. a student of Meinong wrote of his teacher ( 1970):

Meinong in his passion for cleariy articulated theories, developed the idea of

his friend and former mident [namely Ehrenfels] and said that there is a two-

step process. The first step leads fiom stimuli to the sensations. and the

second from sensations to the gestalt. The first step is mainly determined by

extemal sources - that is, by the stimuli - while the second goes on by virtue

of an intemal factor, an act of the subject, which he called the act of

production (p. 133).

It is not imrnediately obvious from On the Theory of Owthat Meinong's ideas may be summed up as Heider did. Notice also that the ones who cailed themselves restalt psychologists (the Berlin school) did not accept this view. Heider's quote is. however. cleariy valid for Meinong's collaborators, Witasek and Benussi. They translated Meinong's thoüghts into a psycholo~caitheory. Witasek was for years the de facto head of the psychological laboratory in Graz. He rnay be considered Meinong's first assistant. Benussi experimented very oRen in the laboratory at Graz and can be counted as Meinong's second assistant. We get a good idea of Benussi's psychology 6om his contribution to the collection celebrating 10 years of psychological experimenting in Graz. Another good source for

Benussi's position is an article Wntten by Koffka.

Ko& (1 9 15) abstracted fiom Benussi's (experimental) work the main theoretical notions of the Graz school. Kofka's description is accurate. Thereafier he Save a detaiied criticism of Benussi's notions and introduced the views of the Berlin school, or in his words,

the Wenheirnerschen Gestalttheorie (the gestalt theoty of Wertheimer]. Note that the Berlin

rhool started to pubticize its viewpoints about 10 years after the Graz school. and that the

Berlin school identified itself partly in opposition to the Graz school.

Benussi's contribution in his laboratory's collection is titled On the Psvcholo~of

Gestdt Perce~tion.The word 'gestalt' is used here in its modern meaning: a whole made up

of parts, irrespective of whether those parts are colours, tones, space elements, or others.

Meinong used 'gestalt' only for the visual domain, but that is not the case with Benussi. That

difTerence noteci, one mua wnclude that Benussi followed the train of thought of his teacher

and boss obediently.

Benussi started his article with the claim that a melody and a figure by their nature

have no extenial reality, that they are ideal. Note that Benussi made the same distinction as

Meinong between tond and spatial gestalten. Also implicit in Benussi's statement is

Meinong's idea that there are basic, fundamental presentations and ones that are the result

of funher imer processing of those presentations. The resulting presentations, or the

outcomes of the processing stage, do not exist in the world. Benussi held that tones and

colours really exist. Melodies and figures, on the other hand, are constructed fiom the

sensations, which, in tum, are caused by stimuli that reach the senses. Melodies and figures

are not really out there; they are ideal objects, unable to affect our senses.

Meinong pointed out in On the Theorv of Qbiects that objects that cannot be

perceived by our senses must be constructed in ourselves. Benussi appiied the sarne logic to Our perception of gestalten. it cannot be denied. of course. that one experiences melodies and figures. But they camot stimuiate Our senses and therefore one has to obtain them through different means. Since one does not receive them From the outside, there remains only one option: One has to make them inside. In Benussi's psychology. a gestalt takes the sarne position as an objective in Meinong's theory A gestalt is the result of further processing of (more) basic presentations. Certain energies have an impact on the senses. resulting in presentations of colours, spaces, tones, and so on. These presentations are comparable to the objects in Meinong's system. The next step in Meinong's system is the production of objectives. In Benussi's psychology, the next step is the production of a new presentation, such as difference (for instance between two colours). But the next one can also be the production of a gestalt, like a melody or figure. Meinong's system of basic objects, objectives and more elaborate objectives becomes in Benussi's hands essentially a system with two phases. The first one is that of the senses. and the second one of production.

Kama (19 15) concluded that according to Benussi, a melody and a note are descriptively the same, but functionally different (see p. 15). It means that in imer perception one cannot differentiate between the ongin of a note and of a melody. We hear then they are both mental objects. Functionally, however, they differ. A note is directly evoked by the auditory system in consciousness, while a melody cannot becorne conscious directly, since it is the result of a production process. Koffka (see p. 16) also made it clear that, according to Benussi, the production process is a mental process and should not be confused with a physical one. The production process works on notes, colours, and so fonh. Thus its building blocks are mental objects. It is not a funher processing of physical effects of stimuli reaching the senses that lead to gestalt perception in Benussi's theory. Here lies an important dserence wit h the Berlin gestalt school, which related the gestalt experience not to funher processing of mental presentations but with brain States, or physiological processes taking place higher up.

Benussi noted that erron can occur in the sense phase as well as in the production phase. A wrong presentation of the outer reality may take place in both cases. The errors which can arise in the first phase, that of the senses, are stimulus bound. They are the result of the interaction between the stimulus and its sense. Subjective factors inside the observer play no role in the appearance of these erron. They are related to the particular sense organ.

Enon happening in the production phase are more independent of the energy reaching the senses. Benussi offered the Müller-Lyer illusion as an example of an error of the second type. He explained that illusion by assuming that in the production of a gestalt, the parts coming from the senses affect each other, causing some of them to be slightly altered. creating the illusion. These errors are more independent of the sense organs and subjective factors play a role in their formation. As a consequence, they cm be infiuenced by the condition of the observer (such as how tired he or she is), by the attitude of the observer, former expenences, and so forth. These errors are caused by a change in the elementary, sensory presentations. Note that Benussi could not account for Stumpf s experimental finding of tone fusion. Benussi holds that each tone gives a sense impression in consciousness and those sense impressions are further processed. That must mean that one should tint hear two notes which then fuse into one, and that is not the case. Koffka, on the other hand. provided an explanation similar to Stumpf s: The two notes lead to an overall brain actiçity resulting in the conscious experience of one single tone.

Kottka ( 19 15) airnrned it up again acnirately (see p. 18). The sense impressions are

( 1 ) a prerequisite of the gestalt, and (2) they are the components of the gestalt. The form of a part as component, however, can differ slightly from the fom ir had as sense impression. The production process can change the form of the parts when putting them together in a gestalt.

WITASEK

Witasek discussed gestalt perception in his ( 1908) textbook Foundations of

Psychology [Gmndlinien der Psychologie] and in his (1910) Psvcholow of Visual Space

Perception [Psychologie der Raumwahmehrnung des Auges]. Bot h works implernent

Meinong's ideas in the area of psychology. The striking similarity with Benussi's work is therefore not accidental. From Benussi's and Witasek's work one can only conclude that

Meinong exercised a profound influence on the thoughts of his coiiaborators. There is an irony in this; when Meinong was a pupil of - the no doubt overpowenng - Brentano, he preserved his independence even at the cost of straining his relationship with his teacher.

Brentano's influence is felt imrnediately at the beginning of Witasek's textbook, where Witasek dserentiated the mental from the physical, based on the criterion that the mental has reference. Psychology, according to W~tasek,should be the study of mental facts.

We find in that passage also an influence of Meinong, who was of the opinion that psychology should only be concerned with facts, and Witasek seemed to agee.

In Psvcholoev of Visual Space Perception, Witasek ( 19 10) stated that the impressions of the senses are grouped together into cornplex unitary objects and it is the latter objects that we experience (see p. 292). The grouping together of sense impressions however, is not a mere ordering of them. Rather, that grouping is a distinguishable mental process resulting in a new and unique mental formation. Witasek's notion of perception becornes clear from the names he gives to parts of the perception process. The grouping gets the name of 'production of presentation,' and its results 'produced presentation.' lnner activity, or production thus lads to a new presentatioq the produced presentation The last presentation can become conscious in its own right.

W~tasekbelieved that the sense impressions are fùrther worked on, resulting in more cornplex presentations of gestalt, movement, and so forth. Both the latter and the former one can becorne conscious. Perception, in this view. is the result of an imer process. Then. the whole is more than the sum of its parts (received ffom the senses). There is something extra added to the parts, namely the production process. He denied that gestalten affect the sense organs. There are, according to him, no outer stimuli that cause the sense organs to perceive a gestalt. Neither does the impact of a stimulus on the sense's organ evoke the presentation of similarity or difference. [mer action is necessary for these presentations to arise. The question then arises, how does the mind know which gestalt to produce? Witasek did not deal with this remark, but I think that his answer would be: It depends on the task at hand. If one is interested in the difference between two stimuli, then, naturally enou~h~ one will look ai that. One has a certain control over the production mechanisms. Still. more production processes can be active at the same time. Therefore. even when one attends to one system others can still be active

Wenheimer ( 1925) did not mention the Graz school by name, but surely his following explanation of the gestalt phenornenon applies to them:

StiIl "some" "higher processes" amount to what is actually given - partiai

sum - which attaches to the sum of what is given, and makes it happen (p.

1 1).

Wertheimer rejected this explanation, which will be clarified with the example of a rnelody.

A melody is not directly perceived, according to the Graz school; only the individual notes are. Those notes are, in Wenheimer's words, what is actually given, the partial sum. That implies that something else must be responsible for our capability to heu a melody. That sornething else has to be attached to the tones and makes it happen. Makes what happen?

The melody. And that somethlig is, of course, the production process. Koflka ( 19 15) noted too that in the Graz view. strictly speaking, the presentation of a gestalt has no outside stimulus; a gestalt is indirectly evoked by outside stimuli. In his textbook Foundations of

Psvchology , Witasek (1908), explaining the hearing of a melody, wrote that the ear produces the experience of single notes. These must be grouped together to form the experience of a melody. That grouping is not a mere ordering, or a puaing together side by side of those notes. Mead, the grouping is a process that uses as input the individual notes and produces as output a gestalt. This process can be smooth, easy, and unrecognized, or more difficult. But its result is always a new presentation. That presentation is thus

çornethùig new, extra to the sense elements, something which the subject has added to the sense elements and which stands side by side with those elements.

Witasek drew an interesting conclusion fiom his schema. He thought that there are different production processes working at the same time. One can, for instance, notice the difference between two colours - the result of one production process. Or, one sees moument - the result of another production process. One can also perceive a gestalt - the result of another production process. He even acknowledged the existence of several production processes leading t O different gestalten. Since there are t hese different production processes, each with its own outcome, one plurality of sense impressions can have various outcornes.

Witasek held that the production processes take place unconsciously and that one is only aware of their results. Nonetheless one has the abiiity to choose among the production processes and select the one or ones needed for the task at hand.

Witasek argued further that the sense impressions as well as the presentations resulting from the production process can becorne conscious. It is the element and the gestalt that are both conscious. Witasek explained why we can perceive individuai sense impressions as well as gestalt impressions. Witasek's framework makes it possible for sense impressions to arise without the experience of a gestalt, a dserence, and so fonh. The other way around, however, is not possible. We camot detect a difference between elements without at least two elements there, nor a gestalt without some parts.

Wtasek explained movernent as follows: The individuai sense impressions result in individual space presentations. Those presentations are compared with one another, resulting in the presentation of a movement. That process of cornparison takes place unconsciously: the experience of movement is considered the result of a cornparison of the various space presentations following each other. Witasek believed that the mechanism of cornparison can explain the presentation of a melody too, or that of a note increasing in loudness. Individual auditory impressions are compared, leading to the gestalt expenence.

Witasek also accounted for the phenornenon of stroboscopic vision. Perception of movement is possible, even when the production process receives incomplete information from the senses. because the production process will still compare the sense material.

Witasek went on to remark that the production process can even distort basic materiai. And he gave as exarnples stroboscopic vision and the Müller-Lyer illusion.

Witasek also mentioned differences between the process taking place in the sense organs and in the production process. The first process accepts only sense-specific energy, whereas the latter accepts input from several or al1 senses. Errors taking place in one process are different From those happening in the other. We saw already that the Müller-

Lyer illusion was explained as an error of the production process.

KOFFKA'S CRITICISM

Before discussing KoWs ( 19 1 5) criticism of the Graz school, let us first compare that school with Stumpfs notions. Mer all, Sturnpf was Koffka's teacher. Witasek believed that there is a relation between the functioning of the brain and consciousness. In that, he agreed with Sturnpf Like StumpÇ he taught that psychological experirnents are intended to evoke mental events under such conditions that they can be observed by the subject and reported to the psychologist.

Benussi's and Witasek's theory held that one is aware of individual sense impressions and the gestalt presentation. Stumpf proposed dso that the elements and the whole are present separately in consciousness We saw that Stumpf, at the end of his life - thus 20 yean afler the publications of the Graz school - was mnvinced that the Berlin gestalt school could not account for the fact that we expenence parts as well as the whole. The whole is in Stumpfs view, however, not the result of imer processing, but of regularities in the stimuli. Here, Benussi and Witasek disagreed with Stumpf; they argued that the content of the whole is functionally different From the content of the parts. Stumpf said that we perceive the elements as weil as their relation. The elements and their relation thus have the same status in consciousness. Benussi and Witasek, on the other hand. held that the elements and the whole are different from each other, but not independent, since there is a causal process from the parts to the whole. We saw that Kama ( 191 5) understood this

Graz notion as follows: The parts are components of the whole and the form of the parts in the whole does not need to be completely equal to their forms when evoked as sensory impressions (see p. 17).

Stumpf began his fùst book Un the Psychological Ongin of Space Perception, with a rejection of associationism. Those theonsts believe that presentations are put toget her f?om more elementary ideas. That process is lawfbl, because association of ideas is related to the regularities of the process of bringing them together. Lawful, however, does not mean that there is a direct relation between the outer world and the way the ideas are associated with each other. Stumpf opposed associationism. He held tht there are regularities witkn . - the stimuli and that the task of the observer is to detect or to abstract those regularities.

Ehrenfels and Rubin also believed that perceived regularities of the stimuli are not put together into a meaninfil whole in the observer. They held as well that those wholes exist outside the observer and that the parts of the whole belong together for reasons not related to the observer. Like Stumpf, they believed that the observer perceives those meaningfùl wholes. The Berlin school continued in this tradition. Gestalten are out there and they are reconstructed in the brain and then presented to consciousness. There exists an isornorphic relation between a gestalt and its perception. The Berlin school would agree that this demands a most impressive physiological apparatus, but not a processing unit that works on elemental units, transfomiing them and in some cases even distorting them in order to create the more complex unit of a gestalt.

Meinong and especially his collaborators Benussi and Witasek proposed a theory which is in many respects opposite to the ones of Ehrenfels, Rubin, Stumpf and the

Berliners. They believed that perception involves processing units that have sensory input and provides a different output. By accepting an imer processing mechanism t hat works on atomistic units and brings them together, it seerns that the Graz thinkers, although they were no associationists, could have been criticized by the Berlin geaaltists for making the same errors as the associationists. KofEka (1915) took on the challenge of the Graz school. He had studied the Graz school very well, as proven by his fair and detailed account of his opponent. The controversy was: Are gestalten real or are only elements real and the gestalten pulled together by the observer.

Koffka noticed that the pillars of the Graz school were: (1) There are two kinds of illusions (one caused by the senses. the other by the production process); (2) there are multiple outcomes from a set of stimuli to gestalt presentations; (3) there are no gestalt stimuli. He would dispute al1 three.

Benussi held that a stimulus can lead to only one sense impression, whereas diRerent production processes can lead to different outcomes. Koflka argued that this strict classification between two separate and distinct phenornena could not be retained. He noticed that the same sense impression can lead to more outcomes. When one concentrates on a grey field, the experiences of this field Vary, sorne appear more yellow and othen more blue. On the gestalt side, there are presentations that never Vary, where there is always the same outcome. Therefore, Benussi's differentiation cannot pass scmtiny. The feature of one outcome is not only reserved for sense impressions, and neither is the feature for multiple outcomes for gestalten. In short, Benussi's classification of illusions in two totally separate entities rnust be rejected.

Kama pointed out that the classification between two distinct illusions can also not be rnaintained. The perception of contrast between two colours, for instance, can Vary even though contrast perception is a sensory, or peripheral phenornenon. Kof£ka believed. in contm to Benussi, that objective and subjective factors are involved in every perception.

He did not believe that one illusion is or@ caused by objective factors and the other one just by subjective factors.

Koffka opposed the view that there are no gestalt stimuli. He believed that there are real gestalten. So, he ( 19 15) asked rhetorically:

And what about the series of events in the central nervous system that we assume are related to the gestalt perception (p. 36).

These events are, according to Koffka. not a rnere addition of the individual excitations.

Mead. they form an overall gestalt process [gestdtete Gesarntprozess] (see p. 36). Kama argued here that a set of individual stimuli has an overall effect in the brain. That overall effect is not an addition of the individual effects caused by each stimulus. The overall effect is different fiom the sum total of the individual excitations. That overall effect evokes the presentation of a gestalt in consciousness. In this view, a gestalt stimulus is the outer condition Ieading to brain activity that precedes a gestalt experience.

The Berlin thesis thar the whole is grasped before the parts now becomes more comprehensible. The overall brain state leads to a gestalt presentation. That brain state is caused by individual stimuli, but their identity is somehow hidden in the overail effect. 1t is thus the overall effi in the brain that evokes the gestalt expenence. That gestalt expenence is primaiy, the elements becorne conscious after the gestalt, according to the Berlin theory.

Note the sirnilarity with Stumpf s explanation of fusion. That too depends on the overall brain activity.

In the Berlin view, the gestait presentation is not secondary to the sense impressions.

Also, the gestait is fiinctiondy more ori@ than the presentations of its elements. A gestalt cannot be comprehended as an assemblage of the individual parts.

KofTka referred to an article that Kohler wrote two years earlier for the same

magazine. Kohler's article was translated into English as On Unnoticed Sensations and

Enors of Judgement. Therein Kohler rejected the constancy-hypothesis, that is the notion that a one-to-one relation exists between an external stimulus and the sensation it evokes. He considered false the view "that the sensation depends only. or alrnost only, on the stimulus and its peripheral reception" (p. 20). Kama ( 1922) explained that rejected notion as foilows:

Given a certain stimulus and a normal sense-organ, we know what sensation

the subject mua have, or rather, we know its intensity and quality, while its

"clearness" or its "degree of consciousness" is dependent upon still another

factor, namely, attention (p. 534).

Langfeld (1937) recalled that in his lectures Sturnpf advocated this theoretical position - the one his students would oppose. Stumpf, according to Langeld, told his audience:

The description of our sensations is always a description of content.

Sensations are the entire content in al1 its parts whether we are aware of

these parts or not. Perception has to do with those parts of which we are

aware. At any given moment there is only a small section of sensory content

of which we are aware. As long as one does not hear the scratching of the

bow in violin playing, for example, one has a pure tone sensation. The

scratching is present in the mind as sensation but below the level of

awareness or perception. When one becomes aware of the scratching, the

perceived content resolves itself into two parts, tone and noise. In such an

example there is no change in the matenal but only in the function of

awareness (p. 3 7). Benussi's view is. of course. similar to the one that Langfeld ascribed to Stumpf

Kohler ( 19 13), on the other hand, believed that:

Apart 60m the stimuli and the well-known peripheral conditions, a number

of other factors. primarily centrai ones, also play an essential role in the

detemination of sense data (p. 38).

Kohler was not very explicit about the other factors. But he said that:

The momentary direction of attention set, and other central factors exen a

powerful influence on the sensory processes themselves (p. 20).

At the end of his article Kohler wrote:

... as soon as the remaining factors, apan fiom peripheral conditions, also

exert their influence. This applies particularly to the psychological correlates

of stimulus complexity, and speclfically to the everyday perception of 1-

(p. 39).

It seems that Kohler had in rnind that centrai factors. nich as attention play an important role in the perception of daily objects. Kahler's view undermined not only Benussi's theory, but any atornistic position, since he denied that there are pnmary elements. Instead, what psychologists considered elements are the result of an interplay between peripheral and centrai forces and as a consequence these so-cded elements are not basic at dl.

Then, however, even a question wch as 'How do two sensations compare?' becornes rneaningless, since, as Koffka ( 1922) noted, "The two sensations themselves do not exist"

(p. 542). That laves the question open: What does exist? The answer of the Berlin gestalt school is: an "undivided. aniculated whole" (Kotlka 1923. p. 542). It is difficult to understand what the Berliners meant with an articulated whole. The best guess, however. seems to be that a whole is not an element (of course not), but that it is nonetheless an identifiable unit with parts or divisions. But now we are ahead of ourselves. In 19 13. their annvec was that the experience of the sensations, the elements, is caused by an interplay of - for lack of a better word - forces. That interplay determines the qudity of the sensations evoked.

Kohler's article is thus a step in the direction of the holistic theory of gestalt, although Kohler did not propose gestalt as a causal factor. In 1 9 13 he was still looking at haon such as attention and fatigue to understand the different experiences of sensations.

The experimental proof of that insight, of the importance of attention on the quality of perceptioh was provided, as we know, by Rubin. Rernernber that in Rubin's experiments the figure, or the field to which the subjects pay attention, was experienced quaiitatively different from the unattended ground.

Kohler's ( 19 13) rejection of atomistic theories aiso becomes evident from how he then fomulated his concept of science.

For me too the task, the goai of scientific investigation, seerns to be the

finding of constants, the diçcovery of constant properties and t heir consistent

ordering in a theoreticai structure, or the discovery of unchanging laws

governing the variation of characteristics and of their relations. Nobody

doubts this... (p. 17).

From the "for me too." and from "nobody doubts this," 1 can only conclude that Kohler had no idea how far his view had developed away from Brentano. Ehrenfels. or Stumpf Those scholars believed that science aims to find basic units and their laws. Note that Kohier did not use the word elements. Instead. he mentioned "constants", "constant properties" and

"variation ofcharacteristics." Science, then, is not about basic units anymore. It is not clear what Kohler meant by properties and characteristics. But whatever they may be. they cannot be compared with, for instance. an atom in chemistry. The property 'red' and the characteristic 'loudness' - and now I assume that this is a correct interpretation - do not stand on the mefooting as a chernicd element.

KoîRa's article, to be distinguished from Kohier's, clearly stated whorn Koflka opposed (namely Benussi) and what he thought was wrong with his opponent's position.

Koffka's article, however, also demarcated KoMsown position. In 19 15, thus. the nascent notion of the Berlin gestalt school had rnahired to a weU-expressed opinion. That means that somewhere between 19 13 and 19 15. the Berlin gestalt idea dawned on its principal players. Brentano and the gestalt notion

My thesis provides a somewhat different analytical framework than Ash's ( 1982,

1995) or Smith and CO-workers(1988). My thesis is not written because 1 disagree with them but to draw attention to the line leading from Brentano to the later gestalt theorists.

The importance of that lineage is touched on, but not emphasized in their studies. This thesis. on the other han& shows the importance of Brentano for the gestalt school. Despite the imponance of other great rninds to the gestalt ide& like Herbart, Goethe and G.E.

Müller. 1 focused on specific schoiars belonging to the Brentano circle. In the discussion of the relevant books and articles 1 did not follow a strictly chronological order, but discussed every SC holar separately .

Brentano was forernostly an erudite philosopher. speciaiizing in Aristotle. He had the greatest respect for Aristotle and therefore it is not surprising that one finds many

Aristotelian concepts in Brentano's analysis of consciousness. Anstotle's notion of 'image' became in Brentano's hands 'presentation.' Object s, according to Brentano, have a presentation in consciousness. Brentano thought that there are complex presentations made up of more elementary ones. He also held that the presentations in consciousness form a unity.

His students would reflect further on how presentations are integrated. Stumpfs criticism of associationism was based on his view that the formation of presentations into a unity is an automatic process. Ehrenfels' gestalt quality is an original solution: it States how a (new) mental elernent groups presentations and groups of presentations into units.

Meinong differed from Stumpf and Ehrenfels in his teaching that the observer actively puts tozether presentations to form a unit. Rubin's contribution is unique; he did not deal with the problem of how presentations relate to one another. Instead, he focused on how one presentation is noticed and the rest not. Like Ehrenfels. he explained that phenomenon through a special mental element, narnely the contour, that encircles a field and thereby making it a figure that stands out from fiom the background.

Al1 the authors assumed that presentations do not aiter when they become part of a more complex presentation. Kohler's ( 19 13) first attempt at formuiating his own notion is a criticism of that supposition. He disputed the hypothesis that there exists a constant relation between a single stimulus and its experience. With that challenge he laid the groundwork for the later gestalt school. Kohler's second rnove was away From dualism: A gestalt is rnirrored in brain processes and, as a consequence, into consciousness. The reason why a gestalt is important now becornes clear. To formulate it in the terms of presentations:

An individual presentation does not (help to) determine the complex presentation; it is the complex that determines how one experiences the individual presentation.

Since presentations are the result of brain processes that depend on outside stimuli, it is the gestalt that determines how one is aware of an individual stimuhs and not the individual stimulus that contributes to the gestalt perception. The Berlin school failed to notice Rubin's explanation. I think that is because Rubin's findings, though not his explanation of them, fit so well into their own theory. How an element is experienced, as Emre+ or as -~ound. depends on the gestalt. according to their theory. There was no need to assume t hat Rubin's explanation would be different frorn theirs, since they had such a plausible expianation themselves.

Brentano's role as one of the founding fathers of psychology is oflen rnentioned.

Especidy his notion of intentionality is recognized. His concept of psychology as a science. however, is less well understood. Brentano appeaied to psychologists to use an empincal approach, to look at the fadsituation of mental phenomena and discover their laws. That is. the mental phenomena must be studied scientificaily. He also wamed against the use of introspection. The description of mental phenornena has many more pitfalls than that of concrete objects. Therefore, Brentano thought psychology was a much more difficult science than the natural sciences.

Ehrenfels. Meinong and Stumpf accepted his appeal to understand the mind on its own terms. Ehrenfels' way of studying the mental was by way of thou@ experiment. he reached his conclusions through reasoning. That was also true for Meinong, although he took case that his &had the use of a psycholo@cailaboratory. Perhaps Stumpf responded best to Brentano's challenge to formulate a science of the mind. His lasting influence was the development of an expenmental method to study mental phenomena. That method differs from the modem method and a study to spell the modern one out in more detail is warranted. Sturnpfs students would apply his method. They would not, however, accept his underlying motivation, that of a hdarnentai dflerence between the mental and the physical.

With Benussi, Rubin, and the Berlin gestalt school we enter the area of modem psychology. Rubin found inspiration in expenmental results. It was experimenting that brought him to the fiyre-gound phenornenon. Benussi's work and that of the Berlin school was also experimentally driven. Experiments give results which are sometimes surprising forcing scientists to consider new theories. The experirnental method, therefore, cm become a source of inspiration on its own. If one cames out expenments oneself. it is only natural t hat one looks at other expenmental sciences to see what cm be leamed fiom them, especially if those sciences show impressive result S. Wit h the evolution of t he Berlin gestalt school neither Aristotle and Brentano were any longer a source of inspiration to psych~logias.Psychology had left its Scholastic period and would look up to the physical sciences.

There lurks a danger in psychoiogy copying the methods of the physical sciences: it can lead to a discardimg of mental phenomena altogether. The so-called 'scientific' method. furthemore. can bring one to the erroneous conviction that the study of the mental is a subdiscipline of biology. The gestalt psychologists, though they saw themselves like other scientists. would not fa11 into either of these traps. They were interested in the study of the irnmediately given and were wiliing to use experience as data; Stumpfs method met their needs. Hence, they stand with mental phenomena and that made it impossible for them to reduce their work to biology. Neverthelesq I think that the greatest error of the Berlin school was their attempt to formulate a rnonistic view, setting aside the notion that the mental and the physical are different. They denied that the mental cm work according to fûndarnentally different laws, a different logic, different mathematics than the physical.

A theory, however, is not only important for how close it is to the truth. Even when wrong, (1) it can lead to important research and thereby improve our understanding, and (2) it can show some weak sides of other views. Ash ( 1995) believed that such is the case for the Berlin school. He wrote in his conclusion that

Gestalt theory's most important historical roles have been that of a

sometimes unacknowledged stimulant to funher thinking, or t hat of aent

provocateur, a constant irritant reminding overly enthusiastic or do-matic

scientists of the limitations in their current approaches (p. 4 10).

Murray seemed to agree that a valuable contribution of gestalt psychology was to point to psychological aspects negiected by other psychologists. He concluded in his ( 1 995) Gestalt

Psycholo~vand the Comitive Revolution t hat :

with respect to memory, we Murray] examined five sub-areas iduenced by

gestalt contributions and found that oflen it was not so much the

expetimental evidence for gestalt theory that was convincing, it was more

that certain issues were clearly brought out which had been overlooked in

the associationist or behaviounst literature (p. 165).

It is, of course, no credit to the gestalt school that the other schools in psychology have their flaws. But. 1 think, the gestalt theorists were aware of them exactly because they grew up in a certain tradition. They learned that conscious expenence is the object of psychological study. When one matures in such an environment, the mistake of neglecting expenence is immediately obvious. Their concept of naniral science included experience and in that concept the voice of Brentano can still be heard.

Brentano had rnany outstanding midents; he was a source of inspiration for a whole generation. His personality and cornmunicating skills can explain his appeal. Students are important, because they guarantee the continuation of a way of thinking. and the growth of the gestalt idea can partly be explained by Brentano's wide following. His student Meinong taught Benussi and Witasek. Stumpf took care of Kohler's nomination at the prestigious

University of Berlin.

The generation afier Sturnpf and Meinong, however. was not in the fonuitous position to train students and help them in their careers. Witasek died early, and Benussi. although highly respected among his Italian colleagues. cornmitted suicide. To escape Nazi

Gemany Wertheimer, Kohler and Koffka had to imigrate. .4nd in the U.S..4., although received with great honours. they were not allowed to train students at a doctorate level.

The ernigration of the Berlin school to the U.S.A. made the gestalt school respected and weil known in the leading country afler the war. Still, the conditions for its fùnher growth were harnpered. Wertheimer and Koffka died early. Kohler had some students in Germany. but they did not grow old either (see Ash 1995, p. 406). The favourable conditions for the gestalt movement at the beginning of the century were, unfortunately, not repeated.

Nevertheless, gestalt research continued to be carried out.

Murray (1995) noted that "the gestalt psychologists mticipated late twentieth century" (p. 185) cognitive psychology, but not "that they idluencd it very rnuch" (p. 285).

That, according to Murray, cm be explained because their "contributions to cognitive psychology have ken underknown (in comparison with their contributions to perception)"

(p. 185). Perhaps the following picture is acnirate for the generation growing up in the '40s.

Psychology students read texts on perception by the gestalt theorists during their training.

Since those theonas had no graduate students, their other ideas were not handed over and, \ as a consequence. cognitive psychologists had to work through their intellectual problems unaided by the work of the gestaltists.

Some ot her observations

Ash ( 1995) also wrote in his conclusion:

Gestalt theory has been a worthy participant in two central intellectual trends

of our the: the revolt against dualism in twentieth-century thought; and the

ongoing struggie between science as technological manipulation and control.

and science as an attempt to undentand and appreciate order in nature (p.

31 1).

[ have already commented on monism and I will leave it at that. As far as the other point brought up by kh, I do not believe in a struggle between application and knowledge per se: in Fact I do not see any conflict whatsoever between the two. There exists. in my view. a logical relation between them (to be contrasteci with a causal link). There is no application without knowledge; first one has to pursue academic interests and then one can see what is of use. It is thus one (application) after the other (knowledge), but both are quite distinctive fields. And the pursuit of progress in each field goes according to its own laws.

And those laws are not necessarily equal to the ones of the other area. 1 think that 19th- century German psychology shows that the development of knowledge and understanding is as activity independent of its practid application. Note, by the way, that knowledge does not guarantee a useful application. b Not one of the scholars discussed in this thesis defended what they did: not one of them pointed to the relevance for society of his endeavour. The application of their findings seemed to leave them cold. Science was for them pure. fundamental, only aimed at comprehending. Mat seemed to rnotivate them was a search for knowledge. for basic principles that pide the mental. .;Ul of them were. however, in the cornfortable position to be able to ban commercial influences from their lives. at least during the period discussed here. With not one exception, they came from reasonably well-to-do or even rich families.

The scholars that 1 studied had a basic financial secunty and were able to accept the low wages paid at the bottom ranks of the German universities. ûthers supplemented their incomes. mostly by teaching at the elite high schools.The state supported university research and there was thus no oeed to look for extemal iünding. There was also no need to defend the relevance of theu work as part of an appeal for more money. Financially. they were ûee and independent: in that sense they could to be as creative as they wanted. Note that there was no financial pressure on them to publish. Still, their output was considerable. gauged according to quantity as well as quality. There were, however, other pressures. For instance, after their dissertation they needed to write another scholarly book for their habilitation.

These scholars were highly motivated and the limitations they faced lay in the organization of the German universities. Upcorning men ni11 had to please their superiors who controlled the scientific joumals. But at the beguuung of this century, great minds couid pursue fundamental research at Geman and Austnan universities with awesome results. 1 beiieve that this finding should be taken into account when adapting our universities to the r challenges ahead. Brentano, Ehrenfels. Meinong. and the Berlin school tell us something about the fiamework in which creative scientists flourish.

It is also interesting to observe that al1 of the discussed scholars were broadly educated, pursued wide interests and were not afiaid to enter new fields. Brentano was trained as a pnest and left the priesthood. Ehrenfels, Stumpf and Meinong had geat love and talent for music. Ehrenfels was taught by the composer Anton Bnickner (see Heider,

1970. p. 13 1) and Stumpf played several instruments (see Kendler. 1987, p. 193).

Furthemore, Ehrenfels wrote on a broad range of ahtic, psychologïcal and societal issues, while Stumpf took an apprenticeship in scientific laboratones. Wertheimer had very wide intellectual interests, and Kohler a considerable knowledge of the natural sciences. The

Berlin school also adopted well to living in a new country; they were not restricted to a certain mode of living in their personal lives or a narrow specialization as professionals. On the contrary. This points, I believe, to the benefits of a broad education.

Divergent opinions

The gestalt notion as such does not exist. Brentano had a holistic view of mental phenornena. Ehrenfels' view can best be described as diaiectic; the gestalt quality and the elements exist together. It is not the elements or the gestalt, but both at the same time.

Rubin's view was that a gestalt jumps out from its background, but the gestalt itself is made up of a field and a contour. The Graz school started with the elements and anived at a gestait. That theory can explain non-existing objects. The Berlin school, on the other hand, * began with the gestalt and amved at the elements. revealing why the same stimulus can be differently experienced. Thus. al1 these schoiars thought deeply about gestalten. Their theories are quite different and sometimes irreconcilable.

Tasks for the Future

.bh situated the rise of the gestalt notion in its contemporary context, while I look at its antecedents. The definitive study of the gestalt schools should encompass both approaches. But that is not dl; not one of the works on the history of the gestalt school - mine included - provides a description of the penonalities of the major players. That, I thidc, has to be corrected, too. Part of Brentano's influence lay in his capacity to keep his entourage spellbound, whereas Stumpf had great difficulty getting his ideas across. Real. living people gave us their ideas and it was their personalities that shaped history. That means that they. in a sense, must be brought back through an artistic act of recreation. In short, the definitive study of the gestalt schools must embrace contextual, lineage-historie, and artistic approaches. And it must build out of al1 those elements the true and fascinating story, the whole that is more than the sum of its paris. hquinas T. (ca 12521 1949). On being and essence. (a. Maurer. Trans.). Toronto:

Pontificid Institute of Medieval Studies

Aristotle ( 194 1 ). The basic works of -4ristotle (R. Mckeon, ed).

New York: Random House.

Ash, M.G. ( 1980). Acadernic Politics in the History of Science. Experimental

Psychology in Germany 1879- 194 1. Central European History. 13, 255-285.

Ash, M.G. (1982). The emereence of gestalt theories: Experimental psvcholonie in

Gemanv. 1 890- 1920. Ann Ahor, MI: University Microfilms International.

Ash. M.G. ( 1995). Gestalt ~svcholoevin Geman culture' 1890 - 1967. Holism and thc

ectivity. Cambridge: Cambridge University Press.

Augustine (ca 397/ 196 1). Confessions. (R.S. Pine-Coffin, Trans.). London: Cox &

Wyman.

Benussi. V. ( 1904). Zur Psychologe des Gestalterfassens (Die Müller-Lyersche Figur. )

In A. Meinong (Ed. ) ,

302-448. Leipzig: Johann Arnbrosius Barth.

Boring, E.G.(1929f1957). A history of expenmentd osychology (2nd ed.). New York:

Appleton-Century-Crofts.

Brentano, F. ( 1866/ 1967). Die Psychologie des Aristoteles. Darmstadt: WissenschafUiche

Buchgesellschaft.

Brentano, F. ( 1874, l928/ l924-28). &chologie vom em~irischenStandpukt (Band 1-3). e Hamburg: Felix Meiner.

Brentano. F ( 1874, 192811973). Psvcholog hman amperical standpoint (AC.

Rancurello. D.B.Terrell & L.L. Mc Allister, Trans.). London: Routledge &

Kegan Paul.

Brentano. F. ( 1 882/ 1 980). Aristoteles Lehre vom t irsprun~des menschlichen Geistes.

Harnburg: Felix Meiner.

Brentano. F. ( l9OV 1 979). Untersuchun~en zur Simesgsychoio~e. Hamburg: Felix

Meiner.

Brentano, F. ( 1933). Katecroriedehre. Harnburg: Felix Meiner.

Brentano, F. (193311987). The theones of cate~ories.(M. Chisholm & N. Gutermq

Trans. ). The Hague: Martinus Nijhoff.

Chisholm, R.M. ( 1982). Brentano and Meinon studies. Amsterdam: Rodopi .

Ehrenfels. C . von ( 1 890). Ueber "Gestaltqualitaten" Vierteljahrsschrift fùr

wissensçhafiliche. Philosophie. 14, 249-292.

Ehrenfels, C. von (189011988). On "Gestalt qualities". In B. Smith (Ed. and Trans.)

Foundations of gestalt theo-, 82- 1 17. Wien: Philosopha Verlag.

Ellis, W.D. (1938/1969). A source book of gestalt psvcholoqy. London: Routledge &

Kegan Paul.

Everson, S. (1995). Psychology. In J. Bumes (Ed.), The Cambrid e com~aniont~

Aristotle, pp 168- 194. Cambridge: Cambridge University Press.

Fabian, R. ( 1986). Christian von Ehrenfels, ein Beitrag zur intellektuellen Biographie. In . . R. Fabian (Ed.) evonEhrenfels.1-64. Amsterdam: Rodopi.

Findlay. J. ( 1 963 ). Meinong's t heoy of obiects and valua. Odord: Clarendon.

Freud, S. ( 1990). The Ietters of Simund Freud to Eduard Silberstein. 187 1- 188 1. (W.

Boehlich, Ed. & A.J. Pornerans, Tram.). Cambridge MA.: The Belknap Press.

Garner. W.R. ( 1962). Un cert ain ty an d stnicture as ~sycholoeicalconce~ts. New York:

John Wiley .

Garner, W.R. ( 1974). The processine of information and structure. Potomac, Maryland:

Lmrence Erlbaum.

Gibson E.J. (1969). Principles of perceprual leamin and development. New York:

Appleton-Century-Crofis.

Gibson, E.J. (1991 ). An odysse~in leamine and ~erception.Cambridge MA: The MIT

Press.

Gibson, J.J. ( 1950). The perception of the visual world. Cambridge MA: the Riverside

Press.

Gibson J.J. (197.1)- The legacies of KofFka's Pnnciples. Journal of the History of the

Behavioral Sciences. vol 7. no 1, 3-9.

Gibson, J. J. ( 1979). The ecological ap roach to visual perception. Boston: Houghton

Mi&.

Gibson, J.J. & Gibson, E.J. ( 1955). Percepnial learning: Differentiation or enrichement?

Psvchol~calReview. vol 62,32-4 1.

Gill, M.L. ( 1989). Aristotle on substance: The paradox of unie. Princeton NI: Princeton

University Press. Gilson, L. ( 1955). La -holoeie descriptive selon Franz Brentan~.Pans: Vnn.

Grossman. R. ( 1974). Meinong. London: Routledge & Kegan Paul.

Heider, F. ( 1970). Gestalt theory: Early hiaory and reminiscences. Journal of the History

gf the Behaviorai Sciences, vol 6- no 2, 13 1- 139.

Helmholz, H. von ( 1857). On the hearing of physiological causes of harmony in music.

In R.M.Warren & R.P. Warren (1968) lgy

and development, 27-58. New York: John Wiley.

Helmholz, H. von (186?/1913). D-e

Gnindlaee- tùr die Theorie der Musik. Braunschweig, Gemany: Fnedr. Vieweg

& Sohn.

Jaeger, S. ( 1992). Wolfgang Kohler in Berlin. In (L. Sprung & W. Schdnpfiug, Eds.)

Beitrage mr Geschichte der Psvcholog& Frankfun am Main: Peter Lang.

Kendler, H. ( 1987). Historicd foundations of modem psvchology. Phiiadelphia: Temple

University.

Koffna, K. (1915). Beitrage zur Psychologie der Gestalt- und Bewegungssserlebnisse.

III. Zur Gmndlegung der Wahrnehmungspsychologie. Eine Auseinandersetzung

mit V. Benussi. Ze'tschrifi1 fir Psvchol~gieund Physiolo e der Sinnesore

-373 11-90.

KofFka, K. ( 1922). Referat . E. Rubin: Synsoplevende Figurer. Studier i psychologisk

cholo&che Forschung& 188- 190.

Koflkq K. ( 1922). Perception: An introduction to the gestalt-theorie. The Psychoiou

Bulletin, 19, 531-585. Koffka. K. ( 1924). The gowth of the mind. an introduction to child psvcholoey New

York: Harcourt, Brace.

KofRa, K. (1935). Pnnciples of Gestalt Psvcholoey. New York: Harcourt, Brace &

World.

Kohler. W. ( 19 13). c%er unbemerkte Ernpfindungen und Urteilstauschungen. Zeitschrift

Gr Psvcholoeie und Phvsiologie der Simesorgme. 66, 5 1-80.

Kohler, W.( 19 1311 97 1). On umoticed sensations and errors of judgement. (H.E. Adler.

Tram.). In M. Henle (Ed.), The selected wers of Wolf~aneKahler. New York:

Liveright .

Kohler. W.( 1947). Gestalt psycholow. An introduction to new concepts in modem

psvcholog. New York: Liveright.

Kohler, W. ( 1969). The task of epstalt pp~holo~.Princeton N'Y: Princeton University

Press.

Kraus. 0. (1919). Franz Br en tano mr Ke~tnisseines Lebens undLeh_remi_.t

BeitrQnvne o Car I Stum~fund Edmund Husserl. München: C.H.Beck'sche

Verlagsbuchhandlung.

Kraus, 0. ( 192 1). Zur Debatte über die Gestaltpsychologie. Einige kritische

Darlegungen. Lotos (Prwe?.69,23 3-242.

Langfeld, H. S. ( 1937). Stumpf s "Introduction to Psychology" . nf

Psychol~.50,33-56.

La Palme Reyes M., Macnamara, J., Reyes, GE. & Zolfaghari, H. (1994). The non-

Boolean logic of natural language negation. Philosop-a, 45-68. Lewin. K. ( 1937). Car1 Sturnpf The Psvchological Review. 443, 189- 194.

Lindenfeld, D.F. ( 1980). The Transformation of Positivisrn. Aiexius me in on^ and the

European Thou t. 1880 - 1920. Berkely: University of Calnifornia Press.

Mach. E. (1886/1897). Contributions to the analpis of the sensations. (CM.Williams.

Trans. ). Chicago: the open Coun.

Macnamara J. ( 1993). Cognitive Psychology and the rejection of Brentano. Journal of

fhe Theory of Social Psycholw. 233, 1 17- 137.

Macnamara, J. (in press). Historical reflections on -.

Macnarnara, J. & Boudewijnse, G.J.A. ( 1995). Brentano's influence on Ehrenfels's theory

of perceptual gestalts. Journalfor, 40 1-418.

Meinong, A. ( 189 1 ) Zur Psychologie der Komplexionen und Relationen. Zeitschrifi Gr

Psvchologie und Physiolqgie der Si~esoreane.2, 245-265.

Meinong, A. (1904/1960). "The theone of objects" (1. Levi. DyB. Terre11 & R.M.

Chisholm, Trans.). In R.M. Chisholm (Ed.) Realism and the background of

p henomenology. Glencoe, ILL: Free Press.

Metziger, W. ( 1963). Zur Geschichte der Gestalttheorie in Deutschland. Psychologia 6,

11-21.

Morgenstern, C. ( 1963). The fence [der Latteruaun]. In L. Forster (Ed. & Transl.)

-in book of Germa. verse, 387. Penguin Books.

Mulligan, K. & Smith, B. (1988). Mach and Ehrenfels: the foundations of gestalt theory.

In B. Smith (Ed.), Foundations of gestalt theory, 124- 157. Wien: Philosophia

Verlag . Murray, D J (1988). A histon, of western psvchoiog. Englewood Cliffs. NJ: Prentice

Hall.

Murray, D.J.(1995). Gestalt Psvchology and the coenitive Revolution. New York:

Harvester Wheatsheaf.

Nuyens S.J., F.J.C.J. ( 1939). Ontwikkelin~momentenin de zielkunde van Anstoteles.

Een historisch - ~hilosophischestudie. Nijmegen -Utrecht: Dekker & van de

Vegt .

Pol, W.H. van de ( 1966). Het einde van het conventionele christendom. Roermond: J.J.

Romen.

Rubin, E. ( 1 9 1 5/ 1 92 1 ). Visuel1 wahrgenomrnene Figren. (P. Collett, Trans. ).

Kopenhagen: Gyldenalske Boghandel.

Rubin, E. ( 1930). Kritisches und Expenmentelles mr "Empfindungszeit" Frohlins.

Psycholoeische Forschung 13, 10 1- 1 12.

Sajama, S. & Kamppinen, M. ( 1987). A historical introduction to phenomenolo~.New

York: Croom Helm.

Smith, B. (Ed.) (1988). Foundations of ggstalt theory. Wien: Philosophia Verlag.

Smith, B. (1988). Gestalt theory: an essay in philosophy. Ln B. Smith (Ed.), Foundations

stalt theory, 1 1-8 1. Wien: Philosophia Verlag.

Spmng, L., Sprung, H. & Kemchen, S. ( 1986). Erinnerungen an einen fast vergessenen

Psychologen? Car1 Stumpf ( 1848- 1936) mm 50. Todestag. Zeitschrift Gr

Psvchologie~194, 509-5 16.

Sprung, L. & Schonpflug, W. (Eds.) (1992). Z LJ r Gb.hi h Frankfurt am Main: Peter Lang.

Stumpf. C. (1873). Uber den psvcholo schen Ursprun~der Raumvorstellunq. Leipzig:

S. Hirzel.

Stumpf, C.( 1883-1890/1965). Ton sychologie. erster Band. meiter Band.

Hilversum/Amsterdarn: A.M. KnuiE.J. Bonset.

Stumpf, C.( 1895). Hermann Von Helmholtz and the new psychology. The Ps~choloeical

Review. 2:1, 1-12.

Siumpf, C.( l924/193O). Car1 Sturnpf (T.Hodge & S. Langer, Trans. ). In C. Wurchison

(Ed.): History of psycholow in autobi~phie~Vol 1, 389-441. Worchesrer.

Mass. : Clark University Press.

Stumpf, C. ( 1926). Die Sorachlaute. Exoerirnentelle - Phonetische Untersuchunaen.

Berlin: Juiius Springer.

S tumpf, C. ( 1939- 1940). Erke~tnislehre.Band 1 & II. Leipzig: Iohann Arnbrosius

Barth.

Sullivan, 3.J. ( 1968). Franz Brentano and the problems of intentionality. In B.B. Wolman

(Ed. ), Histoncal roots of conternporay psvchologr, 248-274. New York: Harper

& Row.

Wertheimer, M.(1912). ExperVnenteile Studien über das Sehen von Bewegung.

Zeitsc Te-h " r 61, 161-265

Wertheimer, M. ( 1925). Uber Gestalttheorie. Vortrw5

-amer 1924. Erlangen): Philosophische Akademie.

Witasek, S. ( 1908). Grundlinien der PsychoIo&. Leipzig: Dürr'xhen Buchhandlung. Witasek. S. ( 19 10). Psvchologie der Raurnwahrnehmung des .4um. Heidelberg: Car1

Wintets Univarsitatsbuchhandlung. IMAGE EVALUATION TEST TARGET (QA-3)

APPLIEC3 - IMAGE. lnc -= 1653 East Main Street , -. Rochester, NY 14609 USA --= --= -- Phone: 71 6/482-0300 ------Fa: 716/288-5989

O 1993. Appiied Image. Inc.. Ali Rqhts Re~e~ed