te;:et &es:'it Ssre{at o.ffiffiffiffiffiffiffiffi abd b#&ffiffimwetrffi# sffimwwwffiffiffi#w

Volume10 Number8 April, 1967

COMMUNICATIONIN BEHAVIOR AND BEHAVIORALSCIENCE

Edited by JouN H. WrlxllNo, assisted bg JlNnr BEAVIN, M ental ResearchInstitute. Palo Alto, Calif ornia

Communicationand Behavior - An Introduction John H. Weakland Some Formal Aspects of Communication Paul Watzlawick and Janet Beavin . 4 On the Structuring of Communication Albert E. Scheflen,M.D. 8 Communication as an Approach to Politics . Robert C. North 12 Unintended Consequencesof InterpersonalExpectations Robert Rosenthal 24 The Ethnography of Experiment Roy Turner 26 Cybernetic Explanation 29

NEW STUDIES SECTION . ... 13 SOMEFORMAL ASPECTS OF COMMUNICATION PAULWATZLAWICK and JANETBEAVIN

The study of communication, like other fields, depends paper. For the present,we will primarily identify the subject greatly on the basic prernises involved. Too often this /eve/ matter of the problem. is taken for granted, but lt is explicitly examined here by Paul Watzlawick and lanet Beavin, Research Associates at PRAGMATICSAS RECIPROCALPROCESS the Mental Research lnstitute (Palo Alto, California). Their Morris (3) proposedthat semiotic (the generaltheory of paper is based on a study of human communication sup- signsand languages)could be divided into three main areas. ported by the Robert C. Wheeler Foundation,and presented We suggestthat thesecan, analogously,describe three levelsof in greater detail in Pragmaticsof Human Communication: analysis of human communication: syntactics as the study of A Study of InteractionalPatterns, Pathologies and Para- the formal relationsof signsto one another;semantics as the doxes, by Paul Watzlawick,Janet Beavin and Don D. Jackson study of the relationsof signsto the objectsto which the signs (New York: W. W. Norton, 1967). Parts of this paper were refer, i.e. the study of ; aod pragmatics as the study presented by Dr. Watzlawick at a recentconf erenceon Sys- of the relation of signsto their users.The last, the pragmatics tematic Researchon Family lnteraction at the EasternPenn- of human communication,encompasses our interest, which sylvania Psychiatric /nstitute, Philadelphia. is the behavioraleffects of human interaction.However. it is necessaryto make at least one important qualificationof the I This paper will describean approachto the study of human aboveframework, which may be saidto be not so much about interaction which is basedon the assumptionthat communi- communication,as about signs, senders,and receiversand, cation is synonymouswith what is observablein such interac- thus, still primarily concernedwith individuals in isolation. tion. That is, communicationis seennot asjust the vehicle,not We would prefer to use the term pragmaticsto refer not to any just as the manifestation,but as a better conceptionof what is sender-signor sign-receiverrelation but rather to an inter- often looselygathered under the rubric "interaction." personalrelation. We would not even say "sender-receiver" We have,of course,no completeor formal theory.We will, relation, if this could be avoidedin our language,in order to rather, presentwhat appear to be very simple and obvious be able to focus on a reciprocal processin which both (or all) premiseswhich, hoWever,lollowed through to their necessary personsact and react, "receive" and "send," in such detail conclusions, seem to yield a fundamentally new and quite and complexity that these terms lose their meaning as verbs productive outlook. of individual action. As Birdwhistellhas put it: Before this, however, some generalcomments about the An individualdoes not communicate;he engagesin or becomes "obviousness"of thesepoints should be made. First of all, it part of communication.He may move,or makenoises . . . but is often the intimately important, especiallyin our own be- he doesnot communicate.In a parallelfashion, he may see,he havior, which is overlookedor difficult to see,precisely because may hear,smell, taste, or feel- but he doesnot communicate. it is, like breathing,Iargely out of awarenessuntil drawn to our In other words, he does not originate communication;he attention. Second, and more important, while few would participatesin it. Communicationas a system,then, is not to be understood model flatly exclude in theory the ubiquity and importance of the on a simple of actionand reaction,however complexlystated. As system, social context, a it is to be comprehendedon the actual researchand applicationtoo frequently transactionallevel. (4, p. 104) stop at lip service, so that involvement with a particular (monadic) subjectof investigationresults, in practice,in the Herein our focus will be on dyadic, in-person communica- neglectof the interactionalperspective. That is, communica- tion, in which the cues exchangedemanate directly from the tional factors are often regarded as random, or potentially voice, the body, or the immediate context. Such communica- excludable,sources of variance.This problem arisesboth in tion is clearly amenableto analysis in terms of the principles the definitionof what is to be studiedand - sincethe behavioral to be outlined below. However, only further study will prove sciencesare ultimately self-reflexiveand all researchby humans whether theseprinciples might not be fruitfully applied to the on other humans is social - to the strategyand analysisof how study of, for instance,the massmedia, international communi- the data are studied; that is, to methodologyas well as to cation, the psychoanalyticconcept of communication with content.In either case,even when the interactionalcontext is one's introjects,or animal communication,though we would thus, in effect, ignored, it does not go away; there always expect that this is so. remains a valid communicational interpretation of the data. The latter is, unfortunately, often in conflict, or at leastdifficult IN THE PRESENCEOF ANOTHER, to integrate with the investigator'sintended, more monadic AtL BEHAVIORIS COMMUNICATIVE interpretation(e.g. 1, 2). Third, and perhapsmost important, Our casefor the generality of such phenomenaas will be these basic notions may be obvious, yet they still are neither describedrests primarily on an assumptionof the inevitability systematizedinto an adequatetheory of communication, nor of communication in social situations (which certainly include consistently utilized in research. more than thosedefined by mere physical presence- tbe above But such considerations take us beyond the aims of this premise states a minimum whose upper limit is yet to be

THE AtuenlceN Br,nevronAt- ScrENTrsr )

defined). It is first of all necessary to remember that the scope tionship level (discussed immediately below) ls conveyed to of "communication" is by no means limited to verbal produc- the receiver; this point has been especially well taken in regard tions. Communications are exchanged through many channels to schizophrenic behavior, e.g. Haley (7), which may be seen and combinations of these channels, and certainly also through positively, as occurring and meaningful, rather than as gib- the context in which an interaction takes place. Indeed, it can berish outside the pale of human communication. be summarily stated that allbehavior, not only the use of words, is communication (which is not the same as saying that be- THERE ARE MANY LEVETS OF INFORMATION IN havior is only communication), and since there is no such EVERY COMMUNICATION, AND ONE ALWAYS thing as non-behavior, it is impossible nor to communicate. PERTAINS TO THE RELATIONSHIPIN WHICH THE Recent animal studies show, for instance, that certaiD monkeys COMMUNICATION OCCURS. will seat themselvesduring their rest periods in a forest clearing A prisoner is held by two guards in a room with,lwo doors. so that no animal looks at any other while staring straight He knows that one door is locked, the other unlocked, but does ahead into the forest. This is not only in order to keep watch not know which. He also knows one of the guards always tells but also for the purpose of resting.They seem to find it neces- the truth, the other always lies, but again the prisoner does not know which. Finally, he has told that way to sary to avoid even the communication inherent in a glance, been the only regain his freedom is to identify the unlocked door by asking very much as a man in a waiting room may stare at the floor one questionof one of the guards.For a long time the prisoner if he wants to be left alone by other personspresent. But this pondersthis seeminglyunsolvable problem, but eventuallyasks behavior itself amounts to the message"Leave me alone" and the correct question: he points to one of the doors and asksone is normally understood by the others as such. of the guards (it does not matter which door or which guard), All behavior is communication: this is true but not trivial. "If I askedyour comradewhether this door is open, what would All behavior has an effect as communication, an often very he say?" If the answeris yes,then that door is locked, and, vice powerful effect which may be one of the most proven assertions versa,if no, then it is open. of social science - though not always deliberately proven. That The charm of this unlikely story lies not only in the fact is, it is quite common that experiments with a variety of inde- that a problem with two unknowns (the doors and the guards) pendent variables, including pharmacological agents, show is elegantly solved through the discovery of a simple decision significant changes in human behavior which are, however, not procedure, but in that a fundamental property of communica- replicable with that variable. However, the original effect did tion is involved in the solution. The prisoner has been given in fact occur, and is usually construed as a "placebo" effect. two quite distinct orders of information to work with. One has We concur with those who interpret each such case as a to do with impersonal objects (the doors), the other with demonstration of powerful though as yet unspecified com- human beings as senders of information, and both are indis- municationalelTects. pensable for the solution. If the prisoner could investigate the At the present state of knowledge there is no ultimate doors himself, he would not need to communicate with anyone evidence whether all behavior is really completely free of about them, he would merely have to rely on the information "noise" in the information-theoreticalsense. However, to date, supplied by his own senses.Since he cannot, he has to include explorations of the problem of noise (randomness) versus the information he has about the guards and their habitual redundancy (order, patterning) in behavior seenas communi- ways of relating to others, i.e. truthfully or mendaciously. cation have shown so consistently an almost unbelievable Therefore, what the prisoner does is to deduce correctly the degree of order and structure behind the protean manifestations objective state of the doors through the medium of the specific that it becomes more and more plausible that communication relationship between the guards and himself, aDd, thus, even- in the widesl sense is at least as rule-governed as natural tually arrives at a correct understanding of the situation by language is determined by its grammar and syntax (e.g. 5, 6). using information about objects (the doors and their state of And, just as in learning a natural language, the ability to com- being locked or unlocked) together with information about municate is based on the acquisition of a very abstract structure this information (the guards and their typical ways of relating or code which is never formalized and, in this sense,is never - specifically, conveying object information - to others). truly conscious. Not only in the abstraction of this story, but in real life as The suggestion that there is virtually no sender-originated well, these two orders of information are present in all com- noise in human communication almost immediately raises the munication. They are called the content and the relationship related issues of "conscious," "intentional," or "successful" aspects of message material; and while the one or the other communication. Many communicational events are routinely may have greater relative importance in a given piece of excluded from, or differently classified within, theory based on communication, communications composed of only the one a simpler "information" model because they were either not or the other are impossible, just as a computer needs data intended to be comnrunication or did not succeed in communi- (information) and a program (information about this infor- cating what was intended. To take an extreme position, one mation), and cannot function with only one of these inputs. which rests precariously on the edge of infinite regress, any The relationship as well as the content aspect is a basic, ever measureof intention is, ultimately, garnered in a communica- present property of communication. tional setting. Thus, I may say that I did not (consciously) To exemplify: if woman A points to womaD B's necklace intend to ignore you, or you may yourself label my behavior and asks, "Are those real pearls?", the content of her question in this manner. But this labelling procedure is no more and no is a request for information about an object. But at the same less than further information in our ongoing communication. time she also gives - indeed, cannot not give - her definition This. of course. has nothing to do with whether the uncon- of their relationship. How she asks (especially, in this case, the scious really exists or whether people really have intentions. tone and stress of voice, facial expression, and context) would Seen as labels of the participantu, such information is only indicate comfortable friendliness, competitiveness, formal busi- communication - valid as such but not complete. Dess relations, etc. B can accept, reject or redefine but cannot The issue of successfulor unsuccessful (and, therefore, im- under any circumstances - even by silence - not respond to plicitly meaningless) communication obviously rests in part A's message. A's definition may, for instance, be a catty, con- on the same sort of judgment of participant or observer. But, descending one; B, on the other hand, may react to it with again, to take an extreme counter-example, verbal or nonverbal aplomb or defensiveness. It should be noticed that this part nonsenseis communicative: information, especially at the rela- of their interaction..has nothing to do with the genuineness of

Apnrl. 1967 \

pearls,or with pearls at all, but with their respectivedefinitions ations, possessinga complex structure in their own right. One of the nature of their relationship, although they may continue man's experiment is another man's in sita social situation . . . to talk about pearls. (lr, pp.2-3) Whether or not communicational closure is reached on the content level will produce agreementor disagreementbe- A STREAM OF COMMUNICATIONALEVENTS tweenthe communicants;on the relationshiplevel, it will result CONSISTSIN A SERIESOF OVERTAPPING in understandingor misunderstandingbetween them - two STIMULUS-RESPONSE-REINFORCEMENT TRIADS phenomenathat are essentiallydifferent, even though the In pointing to the relationship aspect of communication ordinary termswe must usefor labelsdo not reflectthe distinc- and to the fact that all behavior must be consideredcommuoi- tion with completeprecision. (The many levelsand vicissitudes cation, some necessarycorrections of a purely verbal-informa- of communicationin this areahave recently been the subjectof tion model of communication may have been achieved,but the an excellentstudy by Laing et al. t9l). Thus, it is possiblefor result is still not too different from the standardsender-receiver two communicantsto disagreeabout an objective issue but model in its individual focus.In order to stepfrom the emana- understandeach other as human beings,to agreebut fail to tions of participantsto communicationalprocess, a model of understandeach other as human beings,or, to agreeand to this processis needed. understandeach other; by the same token, of course, two If A and B are interacting, that is, exchanging behaviors, communicantsmay fail at both levelsand, thus, both disagree these behaviors can be seen as a stream of alternating com- with and misunderstandone another. A particularly frequent municational events, and this stream may be representedas and clinically very important type of communicationoccurs ar b1 a? b, a, br . . . .bru d, b, a78 . an b, when the two levelsare confused.in the sensethat the com- Let us consider some of the ways the observer may punctuate municants attempt to resolve a relationshipproblem on the this stream of events,that is, what conception of what "really" content level- e.g. argue about a specificissue in order to occursis imposedon it. establishwho is the better. Another situation of particular One of the most common formats (indeed,nearly impos- clinical importancearises when a personis somehowcoerced sibleto avoid in our language)is that of individual actions,so into denyinghis own correctperceptions (on the objectlevel) that the eventsare seenas in order to maintain lhe statusquo on the relationshiplevel. aaa2a3...... d,i.i ars.,. 3n Asch's experimentson independenceand submissionto group b1 b2b3...bro br, .. .b, pressures(9) provide an excellent example and so do, of And, especially if one of the participants, say A, is the subject course,all doublebind(10) situations. of study (for example, a psychiatric patient or an experimental From the aboveit appearsthat it is easiestto illustratethe subject), then only items a' a2, a3, etc, are examined. The relationshipaspect of communicationwith pathologicalexam- behavioral events are divided and isolated on the basis of plesin which incongruencyof definitionoccurs. But, of course, whether they are ascribed to A to B. Laing et al.have focussed in those relationshipswhere there is understanding,or con- sharply on these "banal and unproductive errors": gruency of definitions,omnipresent though often very subtle The failure to see the behavior of one person as a function cues support and reinforce this state of affairs. Pathological of the behavior of the other has led to some extraordinary examplesmay be helpfuly clear, but they also mistakenlypro- perceptual and conceptual aberrations that are still with us. For instance,in a sequenceof movesin social interaction between mote a common notion about "breakdown"of communication a person A and person B, a, -) b, -; a, -+ b, a. b' which generality * ) missesthe of the underlying principles.This the sequencea1 -> a, -7 a, is extrapolated.Direct links are formal aspectof communication,then, is presentedhere not to made between at ) az ) a' and this artificially derived divide "good" and "bad" communication,but to define an sequenceis taken as the entity or processunder study. (8, p. 8) inevitableproperty of all communication. Often this sequence is further collapsed into an overall state- Thus, stated more strongly, i la Heisenberg,there are no ment about A. the actor. That is. the static sum of these ob.lectivefacts outside the relationshipcontext in which they behaviors areexperienced. Even - perhaps,especially - the interpersonal ar|_ar*ar....+an:A' situationsin which scientific data is gatheredabout human is seen as a characteristic of A - e.g, incomplete seDtences, behavior should be examined in this light. For example, a literality, hostility, etc., depending on the level of abstraction young medical student was assignedan exercisein epidemi- at which a statement is made about A. In such cases, any ology as a summerproject in which he was to make a survey "explanation" of A's behavior must be intrapersonal, attribu- of theincidence of suicidein a particulararea. Candid coroners table to some intrapsychic structure or process which then informed him that this was not strictly possible,as a verdict of becomes the primary object of aDalysis. A less extreme, quasi- suicideversus natural or accidentalcauses of deathwas based interactional approach is to include the behavior of the other well - on social as as medicalfacts delicacy,the testimonyof but in the same form as above. That is, b, * b, --l- b. . . . . + b" involved survivors,the context in which death occurred,and becomes B/ which is then somehow related to A'- for example, other considerationsroutinely entered into the final verdict. the concept of sadomasochistic symbiosis from a sequence of Indeed, Rosenthal'sstudies of "experimenterbias" reported behaviors between two persons. on in this issueexamine what might in our languagebe called In the above cases, either the individual is conceptually effectsstemming from the relationshiplevel of communication isolated from the behavior of those around him, or his be- betweenexperimenter and subject,while Turner considersthis haviors are seen as time- and order-free "properties" of him samekind of relationshipcontext more broadly: as an individual, or both. The alternative has been illustrated Humanbehavior is so organizedthat subjectscan anddo enter by Scheflen: experimentswith conceptionsas to whatbehavior is appropriate We notice three people standing on a corner facing each other to produce,what degree of complianceis desirable,and what one and talking. A number of abstractionsis possible.Each has on a owesto the experimenter.I suggest that theseremarks, which brown suit. Thus, browness is abstractable. So is humanness,or merely draw upon the commonplacesof the role conceptionof standing-ness,or two-leggedness,But it is also possibleto make social action, would be utterly trivial if made in referenceto another kind of abstraction. We can abstract relatedness,e.g. almost any other role relationship;and yet with regardto the lab proximity, kinship, cooperation, and so on. Once we abstract situation of subject and experimenterthey are likely to seem relatednesswe no longer have organismic wholenessor individu- merely nihilistic. . . . Like job interviews,courtroom proceedings, ality,'V,le have a concept. Thbre are only qualities or behavioral therapy sessionsand card games- experimentsare social situ- arrangementsin a concept; there are no people. (5, p. 58)

6 THB AuenrcAN BEHAVToRALScrENTrsr One schemewhich includesboth parties in a dyadic relation concern and anxiety he feels doubly guilty for causing them and which also takes note of the relation between events and emotional pain; on seeing his depressionthus increase they the order in which they occur is the stimulus-response-rein- try harder and at the same time feel more desperatefor being forcement format from psychology.Thus, a, is called the stimu- unable to help him; which in turn compounds his depression lus, b1 the response,and a, the reinforcement. However, prob- to the point of considering suicide for being so "bad,' to those ably as old as studiesof this kind is the joke about the labora- who love him. tory rat who boaststo another rat, "I have trained my experi- These examplesof vicious circles can be multiplied almost menter so that every time press I the lever he gives me a piece indefinitely.The point is that when the model of overlapping of cheese."What this rat does is simply to impose a different stimulus-response-reinforcementtriads is adopted, no one punctuationon the streamof events: what to the experimenter participant'sbehavior can be said to causethe other's: each is the rat's response,the rat considers its stimulus to the is both causeand effect of the behaviorof the other. In most experimenter;what he then does and calls a reinforcement,the ongoingrelationships, it becomesobvious on examinationthat rat seesas a response,and so forth. What we are proposing, the behavior of each participant is predicatedon that of the following Bateson(e.g. 12) is essentially to extend this con- other. The exact nature of these relationship links and the ception of things to encompassboth the experimenter'sand the more abstract rules which govern them emerge as a new, rat's positions.That is, any behavioralevent in a sequenceis a virtually unexploredarea of human behavior. stimulusfor the eventwhich follows it, and both responseand One final, very generalpoint derivesfrom such considera- reinforcementto the one which preceded it: tions. In the communicational perspective, the question ...'&rc bru &r; b' whether there is such a thing as an objectivereality of which SRRf somepeople are more clearly awarethan othersis of relatively SR Rf little importance compared to the significance of different s RRf viewsof reality due to differentpunctuations. However, aware- SRRf ness of how one punctuates is extremely difficult owing to SRRf another basic property of communication. Like all other SRRf complexconceptual systems which attemptto make assertions Just as all behaviorin interpersonalsequences is communica- about themselves(e.g. language,logic, mathematics) com- ,.simul- tion, so all such behavior has this triple aspect,being munication typically encountersthe paradoxesof self-reflex- taneously a stimulus, a response,and a reinforcement, accord- ivity when trying to apply itself to itself. What this amountsto ing to how we slideour identificationof the triad up and down is that the patterns of communication existing betweenoneself the series."(L2, p.a) We proposethis as the minimum com- and others cannot be fully understood,for it is simply impos- plexity of any inrerchange. sibleto be both involved in a relationship(which is indispens- Far from being spurious, however, the punctuation of the able in order to be related) and atthe sametime standoutside sequenceol events is a highly significant corollary to the it as a detached,uninvolved observer (which would be neces- premiseof an essentiallyunpunctuated stream. As philosophers sary in order to encompassand to be awareof the relationship of science,e.g. Popper (13), have pointed out, man is born in its entirety). This is essentiallysimilar to the impossibility with a propensity to look for regularitiesin the constantstream of obtainingfull visual awarenessof one'sown body, sincethe of events surrounding and involving him. The Berkeleyan eyes,as the perceivingorgans, are themselvespart of the body question whether these regularities exist in actual fact or are to be perceived. merely introducedby the observer,will be of no concernhere. As Russell,Gijdel and Tarski have shownonce and for all, (There is no reason why the day should be divided into 24 no systemcomplex enough to includearithmetic can achieveits hours,but this arbitrary divisionis a very usefulone).-The fact its own fully consistentformalization within its own framework remains that the sameprinciple is at work in human communi- and in its own language.Whether or not humancommunication cation: maD teDdsto pattern the stream of communicational is a comparablesystem and, therefore,beset by the sameprob- eventsinto an order which to him is familiar and predictable. lem of ultimate undecidability, is not yet clear at all. There is This, however,presupposes selection and the criteria applied much that speaksin support of this assumption,mainly the to this processof selectionare anything but simple and obvious. above mentioned problem of subjective awareness,or the fact In particular, thesecriteria neednot be obviousto, or shared that to communicate(or evento think) about communication by, the other communicants,Discrepancies in the punctuation itself is itself communication.In this senseboth one'ssubjective of jointly experiencedeveDts are in fact at the root of many experienceof communicativeprocesses with others as well as conflicts in most areas of human interaction, and the ever the study of communicationas such has to employ concepts presentblindness for the other'spunctuation, coupled with the whoserange includes themselves and, thus,lead into Russellian naive conviction that reality is the way 1 see (punctuate) these paradoxes of self-reflexiveness,into an infinite regress of events,almost inevitably leadsto the mutual chargesof badness assertionsabout assertions,and into the problem of undecida- or madness. bility in Gtidel'ssense ( 15). Bronowski,in a lucid studyof this Thus, for instance,nation A may arm to protect itself vexing problem (16), has shown precisely that on the one againsta real or imaginarythreat by nation B. Nation B con- hand "any description in our present formalisms must be siders lftl's a threat and increasesits armaments,justifying this incomplete,not becauseof the obduracyof nature,but because step as a purely defensivemeasure, made necessaryby nation of the limitation of languageas we useit," (p. 5) but that on the A's threatening attitude; nation A now has further "proof" of other hand it is obvious that the mind somehowsolves these nation B's aggressivedesigns, etc. etc., a mechanismextensively problemsin a highly typical way for which mathematicsor studiedby Richardson(14). A paranoid patient suspectsthe logic offer no analogies. motives of others; this prompts the others to prove to him It seemsto us that in the field of human communication the honesty and sincerity of their intentions, which not only the main evidence for the correctnessof this assumption is confirms but increaseshis suspicions,for, he argues, if they supplied by the phenomenonof growth and changein relation- were not out to hurt him, they would not be trying so hard to ships which are the equivalent of Baron Munchhausen'sfeat of make him believe that they meant well. A depressedpatient pulling himself from the quagmire by his own pigtail. The withdraws; his withdrawal worries those close to him; they paradox of changehas occupiedthe human mind since the days try to help him by increasingtheir attention; on perceiving their of the Presocratics,and it will not be solvedhere. But it remains

Apnn. 1967 a question of the greatest importance, especially in the light 7. Iay Haley, "An Interactional Description of Schizophrenia," of communicationalpatterns and the questionof their change- , 22 (1959) 321-332. ability which, after all, is the main factor in ,in 8. R. D. Laing, H. Phillipson, and A. R. l*e, Interpersonal Percep- conflict resolution oD a personal as well as on an international tion (New York : SpringerPublishing Company, 1965). scale,and ultimately in man's awarenessof reality. 9. S. E. Asch, "Studies of Independenceand Submission to Group Pressures,"Psychological Monograph,7O, No. 416 ( 1956). REFERENCES 10. Gregory Bateson,Don D. Jackson,Iay Haley, and John Weak- land, "Toward a Theory of Schizophrenia," 1. Don D. Jacksonand , "TransferenceRevisited," Sehavioral Science r (1956) 25r-264. tournal ol Nervousand MentalDisease, 137 (1963) 363-371. 11. Roy Turner, "Problems in the Study of (Paper 2. Robert Rosenthal,Experimenter Effects in BehavioralResearch Interaction," read at the Pacific Sociological (New York: Appleton-Century-Crofts,1966). Meetings, Van- couver,April, 1966). 3. Charles W. Morris, "Foundations of the Theory of Signs," in Otto Neurath,Rudolf Carnap,and CharlesW. Morris (eds.), 12. Gregory Bateson, "Exchange of Information about Patterns of Human International Encyclopediaof Unified Scienci (Chicago: Uni- Behavior," in William S. Fields and Walter Abbott (ed,s.), versity of ChicagoPress, 1938) I, No. 2, 77-137. Inlormation Storage and Neural Control (Springfield: 4. Ray L. Birdwhistell,"Contribution of Linguistic-KinesicStudies Thomas,1963). to the Understandingof Schizophrenia,"in Alfred Auerback 13. Karl Popper, Coniecturesand Relutatioas (New York: Basic (ed,.), Schizophrenia.An Integrated Approach (New York: Books,1962). Ronald Press,1959) pp. 99-123. 14. Lewis Fry Richardson, "Mathematics of War and Foreign 5. Albert E. Scheflen,Stream and Structure ol Communicational Politics," in James R. Newman \ed.), The lltorld ol Mathe- Behavior, Context Analysis ol a Psy-chotherap,-Session (Be- matics (New York: Simon & Schuster, 1956) II, 1240-1253. havioral StudiesMonograph no. 1. EasternPennsylvania psy- 15. Ernst Nagel, and James R. Newman, Gijdel's Prool (New chiatricInstitute Philadelphia, 1965). York: New York UniversityPress, 1958 ). 6. R. E. Pittenger,Charles F. Hockett,and J. J. Danehy,The First 16. J. Bronowski, The I-ogic of the Mind," American Scientist54 FiyeMinutes (Ithaca: Paul MarlineauPublisher. 1960). ( 1966)r-r4.

Tne AprcnrcAN BEHAvToRALScrENTIsr