CANADIAN PSYCHOLOGICAL REVIEWPSYCHOLOGIE CANADIENNEVol. 16, No. 2, April, 1975

BEHAVIOURISM? COGNITIVE THEORY? HUMANISTIC PSYCHOLOGY? —TO HULL WITH THEM ALL!1

D. E. BERLYNEUniversity of Toronto

ABSTRACT

Some curious features of the current psychological scene are briefly examined.They include the notion that psychology is undergoing a paradigm shift and thatthe paradigm due for suppression can be identified with "behaviourism". It issuggested that, although a return to Hull's behaviour theory con scarcely beadvocated, some characteristics of Hull's approach which have now been widelyabondoned, could with advantage be revived. These are (a) the objective ofintegration, (b) attention to motivational problems and (c) the recognition of theimportant contrasts, as well as continuities, between behaviour controlled by symbolicprocesses and behaviour not so controlled. These might enable us to avoidopposite shortcomings of contemporary Skinnerian and neo-cognitivist positions.

RESUME

Examen rapide de quelques caracteristiques de la situation actuelle en psychologie.On y retrouve notamment I'idee que la psychologie est en voie de changer demodele et que le modele appele a disparaitre peut s'identifier au "behaviorisme".La suggestion est faite que, mane si un retour a la theorie behavioriste de Hullpeut a peine se defendre, il n'en reste pas moins que certaines de ses caracteristiqueslargement abandonnees aujourd'hui pourraient avantageusement renattre: a) lesouci d'integration; b) I'attention portee aux problemes de motivation; c) lareconnaissance des continuites ainsi que des contrastes importants entre le com-portement qui est controle par des processus symboliques et le comportement quine Test pas. On devrait pouvoir ainsi parer aux deficiences inconciliables despositions skinneriennes et neo-cognitivistes contemporaines.

PARADIGMS to circles of like-minded colleagues. Theyr̂ . . , , . . .. are all confident of riding the wave of theDuring the last few years, the progress . . . . . . . .. s , . . . .

. „ „ „ , , , , , ' . *, & . future, oblivious of the maelstroms that areof psychology has been marked by alarming a b o u t ' t o o p e n b e n e a t h t h e m . But apocalyp-jolts and grinding noases. And if it is a por ten tfare a p p a r e n t a t t h e a b r a s i v e in-T S ?T • P60**1* s a y l n g f Pea t" terfaces - where psychologists complainedly that there is a crisis, then psychology a b o u t t h e i r ^ * , s h o r t c o m i n g s t o t h e

is clearly in a state of crisis. Many com- m ! ^mentatorshavmg read Kuhn's book (1962) J [ ^ f r e e l a n c e i n t e l l e c * u a l s a n

Pd r e .

The structure of scientific revolutions, feel s e n t a t i v e s o f o t h e r d i s c i p l i n e s v o i c e theirthat they can recognise what is happening u n f l a t t e r i n g impressions of contemporary(e.g., Boneau, 1974; Deese, 1972; Dember, n < ! V P n n l n c r v1974; Gardiner, 1973; Giorgi, 1970; Palermo, P**"1010*^-1971). Psychology is undergoing a para- r h e Paradigm that is in the process ofdigm shift. being abandoned is variously labelled "be-

haviourism," "S-R psychology," "mechan-A high proportion of psychologists seem i s n l ) . . a n d «reductionism." At least, these

curiously unaffected by all this, as they a r e s o m e o f t h e pouter things that it isdoggedly pursue their work in their ac- calied. And the preferable paradigm that iscustomed manner, confining their contacts supplanting it is variously identified with

"cognitive theory" or "humanistic psy-iThis article is a slightly expanded version of the chology." These t w o expressions are , ofPresidential Address to Division 1 at the meeting of & J ,the American Psychological Association, New Orleans, COUrse, b y no means SynonymOUS, and eachSeptember 2, 1974. Its preparation was facilitated by nf {+.„„, sc nep/l Kv rtiffurvnt mithnrc tnresearch grant A-73 from the National Research Council. O I " iem IS USea Dy ail ierent autnors to

69

70 D. E. BERLYNE

denote very different positions and kindsof activity.

In the midst of the current turmoil, it isalmost impossible for us to gain a balancedview of what is going on. For that, weneed the detached perspective of posterity.The best we can do is to Imagine somefuture time, say 100 years from now, whenthe history of psychology will be as welldeveloped a discipline as the histories of artand philosophy have been for some time andas the history of the physical sciences isjust becoming. We must try to see thingsthrough the eyes of a young trainee in thisdiscipline — let us call him John Q. Gra-duate Student — who has been able to fulfilthe rigorous entrance requirements, includ-ing above-average credulity and a hyper-trophied sense of humour, and has gonethrough the requisite apprenticeship.

He will look back on the early 1970's,known as the "period of the great paradigmshift," with particular warmth, but, as withall good graduate students, a few questionswill occur to him and he will want to lookinto them.

First, did psychology at that time have aparadigm to repudiate? A paradigm is ini-tially defined in Kuhn's book (1962, p. 10)as an "achievement that is sufficiently un-precedented to attract an enduring group ofadherents away from competing modes ofscientific activity and sufficiently open-ended to leave all sorts of problems for there-defined group of practitioners to resolve."Kuhn (p. 13) recognises pre-paradigm stagesin science. Physical optics before Newtonin a case in point, and he writes, "being ableto take no common body of belief forgranted, each writer on physical optics feltforced to build his field anew from itsfoundations. In so doing, his choice of sup-porting observation and experiment was re-latively free, for there was no standard setof methods or phenomena that every opticalwriter felt forced to employ and explain.Under these circumstances, the dialogue ofthe resulting books was often directed asmuch to the members of other schools as itwas to nature." That description soundsuncomfortably familiar and certainly fitsmuch of psychology at present. Neverthe-less, particular areas of research in psy-chology have certainly had paradigms thatconform to Kuhn's definition. But this is

different from the existence of a paradigmfor psychology as a whole.

In later passages of his book, Kuhn as-sociates paradigms with principles on whichscientists are for a while agreed. But psy-chologists have never been noted for con-sensus or for approval of one another'swork. John Q. Graduate Student will there-fore be moved to ask himself the nextquestion. Was there more paradigm shift-ing in psychology in the early 1970's thanat previous times? Was psychology evernot in a state of crisis? In the third quarterof the 20th century, fruitful new paradigmswere appearing in such fundamental areasas learning, perception, motivation andthinking. But it could be argued that newparadigms in these and other areas hadappeared every 10 years or so since theirinception. As long ago as 1937, Spearman(p. 5) wrote, "But what really may startleand even disturb us is that those very psy-chologists who profess — and are every-where admitted — to bring such a vastaccretion of knowledge and power, arenevertheless in extreme discord and conflictwith one another! . . . . What would theworld say if presented with ten rival phy-sics or botanies or chemistries? Curiouslytoo this clash of schools is no special char-acteristic of modern times. So far back ashistory goes, there seem to have been con-flicts of a similar kind, if on a smal'erscale." There is no lack of comparablequotations from earlier and later authors.That psychology has had a history of unin-terrupted upheaval making the San Andreasfault look like permafrost can scarcely bequestioned.

BEHAVIOURISM

John will then be brought to his third setof questions. What was behaviourism? Wasit a paradigm, and if so, were its short-comings severe enough to call for its super-session? He will find no lack of writerswho were willing to characterise behaviour-ism for him. One British psychologist,Rowan (1973), claims to have shown thatbehaviourism is "wrong theoretically, wrongtechnically, wrong morally and wrong poli-tically." Von Bertalanffy (quoted in Koest-ler, 1967, p. 352) described it as a "sterileand pompous scholasticism which, withblinkers of preconceived notions or super-

BEHAVIOURISM? COGNITIVE THEORY? 71

stitions, doesn't see the obvious; whichcovers the triviality of its results and ideaswith a preposterous language bearing noresemblance to normal English or soundtheory, and which provides modern societywith the techniques for the progressivestultification of mankind." The stultifica-tion of mankind, as even a cursory readingof history will show, has been doing prettywell for thousands of years. There is littlesign that its practitioners have ever neededto take lessons from behaviourists.

Behaviourism has been represented as atissue of elementary errors. It has been de-nounced with ad hominem arguments, ethi-cal and political arguments, philosophico-linguistic arguments, metaphysico-theologi-cal arguments, and arguments of Chom-skyan inspiration. It has been charged withaping physics to an extent that is inap-propriate to the subject matter of psy-chology. It has been taken to task for notkeeping abreast of the revolutions that haveoccurred in 20th-century physics (which pre-sumably means that it has not been apingphysics enough!). It has been taxed withcategorising behaviour in ways that are notcorrect or meaningful. All these objectionscall for careful examination, and the ques-tion arises, as with many patrons of taverns,of whether they will survive large doses ofwhat they call for.

Time does not permit us to go into thisquestion here and now. But one might sup-pose that behaviourism, as a target for somuch opprobrium, would go undergroundand that its partisans would be at pains tohide their connection with it. Surprisinglyenough, the opposite is alleged to be thecase. Locke (1971) has complained, withspecial reference to Wolpe's behaviour ther-apy, that some psychologists falsely repre-sent themselves as behaviourists when theyare not entitled to this stigma. He points outthe danger that, if somebody uses a termlike "behaviourism" in a sense that is notits normal one, "behaviourism (defined inthe traditional way) will be given credit forsomething that is not deserved." But whatis the traditional meaning of behaviourism?As John Q. Graduate Student will readilydiscover, this question is far from easy toanswer.

The earliest definition appears in the firsttwo sentences of Watson's 1913 article in

the Psychological Review: "Psychology asthe behaviorist views it is a purely objectiveexperimental branch of natural science. Itstheoretical goal is the prediction and controlof behavior." As the earliest pronounce-ment on the subject of the man who coinedthe term, this must carry a certain author-ity. Watson went on to write books withtitles like Behaviorism (1925) and Psych-ology from the standpoint of a behaviorist(1919). Readers can be forgiven for ident-ifying behaviourism with the views present-ed under these titles. But John will notethat most of those views had ceased to haveany adherents long before the 1970's. Infact, it is doubtful whether anybody otherthan Watson ever subscribed to them.

Then, John will find that, from about1930 onwards, a number of theorists ap-peared who shared Watson's concentrationon behaviour as a subject matter for scien-tific inquiry but otherwise had little incommon with him or for that matter withone another, apart from a predilection forcarrying out animal experiments and forbuilding theories of learning. In fact, theyspent most of their time disagreeing withone another. They were commonly clas-sified, and were apparently not averse frombeing classified, as behaviourists in a broadsense. But, in order to distinguish theirpositions from the early behaviourism ofWatson, they are frequently and properlycalled "neo-behaviourists."

In the 1950's and 1960's there were manyoffshoots of the early neo-behaviourisms inevidence. But there were also writers likeBroadbent (1973) or like Miller, Galanter,and Pribram (1960), who accepted classi-fication as behaviourists, even though theirways of thinking departed radically fromthose of earlier bearers of this designation.

Meanwhile, philosophers made their con-tributions to the further complication of analready tangled situation. Psychological be-haviourisms are often confused with theposition known as "philosophical behaviour-ism," which equates the occurrence of parti-cular mental events or processes with theperformance of particular kinds of be-haviour (cf. Fodor, 1968). Ryle (1949, p. 327)and some of the philosophers associatedwith the logical empiricist school haveleaned towards it. Traces of it are even tobe found in works ot such dissimilar philo-

72 D. E. BERLYNE

sophers as Wittgenstein and Sartre. Philo-sophical behaviourism can derive some en-couragement from occasional remarks byWatson, such as that "what psychologistshave hitherto called thought is in shortnothing but talking to ourselves" (1925, Ch.X). But there is little reason to believe thatthese philosophical issues were a major pre-occupation of Watson's, and neobehaviouristshave, on the whole, studiously kept awayfrom them. On the other hand, one philo-sopher, Quine (1969, p. 97), has contendedthat behaviourism is "just the insistenceupon couching all criteria in observationterms . . . whose application in each parti-cular case can therefore be checked inter-subjectively." He thinks of behaviourismas "covering all reasonable men," This is,of course, going a little too far. The vastmajority of reasonable men are not be-haviourists, for exactly the same reason asthe vast majority of reasonable men are notbotanists. But can any reasonable man bean antibotanist?

John Q. Graduate Student will surely bynow be forced to the conclusion that, whilethere can be and have been behaviouristparadigms, behaviourism itself is not a para-digm. This is the conclusion reached bythe British philosopher, Briskman (1972),after examining the question carefully. Heasserts that behaviourism is a "methodo-logical-cum-metaphysical research p r o -gramme." Apart from their acceptance ofbehaviour as their subject matter, behav-iourists share a number of working assump-tions, some of which are common to all ex-perimental science and some of which aredictated by the fact that behaviourist psy-chology is psychobiology. It concerns it-self with biological aspects of behaviour,with ways in which behaviour contributes toadaptation and ways in which adaptive andmaladaptive consequences reflect back onbehaviour in their turn. It relates dis-coveries about behaviour to discoveries madeby other branches of biology. Consequently,it is bound to accept the guiding principlesthat all biology must accept until such timeas they are overthrown, including the theoryof evolution and the repudiation of vitalism.

So, John will end up with a conception ofbehaviourism that is so broad that, asseveral writers including Locke (1971) haveobjected, the term may become "meaninglessand unnecessary." But, if we are to be

guided by a survey of the various positionsthat have been given that name or that haveaccepted that name, we have no alternative.If behaviourism is neither a sinister con-spiracy to subvert society nor the sole hopeof solving all society's problems, if it ismerely the scientific study of behaviourfrom a psychobiological point of view, thenbehaviourism is not only compatible with,but inclusive of, what many, if not most,contemporary psychological researchers aredoing, including many who polemiciseagainst behaviourism. But is this pointworth dwelling on? If somebody is so hide-bound and incapable of keeping up withlinguistic change that he prefers to call atoilet a privy, is this not harmless but ratherpathetic? Nevertheless, confusions overthese points are unfortunately causingtrouble for psychology. They are unneces-sarily disturbing the public relations of psy-chology, and they are leading psychologiststo spend time on acrimonious wrangles thatwould be more fruitfully devoted to the con-structive advancement of knowledge.

The malaise that is at present afflictingpsychology, and from which it has actuallynever been free, is at least partly verbal inorigin. A few years ago, an official of aCanadian Provincial government, a formerprofessor of engineering, made an unkindpublic attack on psychologists, because hehad gathered the impression that they nolonger concern themselves primarily withthe psyche. This was, of course, exactly asif a psychologist had criticised engineers onthe grounds that they no longer confinetheir attention to engines. If etymology isto be taken seriously, then psychologistsshould be studying one thing only and thatis the physiology of respiration, because theclassical Greek word psykhe, derived from averb meaning "to breathe or blow," seemsoriginally to have meant breath or air. Itmust have been noticed quite early that airis necessary to life and capable of settingobjects in motion, and, until the work ofTorricelli and Pascal in the 17th century, itwas believed that air had no weight andwas thus not a true form of matter. Theword had come to stand for a much subtlerand sophisticated concept, though still re-taining traces of its original connotations,by the time of the first Western book onpsychology, Aristotle's Peri psyhhes, morefamiliarly known by its Latin title De

BEHAVIOURISM? COGNITIVE THEORY? 73

anima. We all know the rest of the story.Psychology was a sobre, solid branch ofphilosophy until the last century or so, dur-ing which the term has come to stand fora wider assortment of activities than anyother occupational label, with the possibleexceptions of "civil servant" and "enter-tainer."

One jazz critic called the late Duke Elling-ton a "master musician, master psycholo-gist, master choreographer." This is fullyconsistent with what the word "psychology"means to many laymen, namely the art ofinfluencing people for their benefit, forone's own benefit, or perhaps for mutualbenefit. It is an art in the original senseof "skUl" or "craft". The term refers to aquite different kind of art when applied, asit frequently is, to the graphic delineation ofmental states and of characters by novelists,poets, dramatists, actors and painters. Andfinally, the word "psychology" is used todenote a variety of scholarly pursuits, some-times classed as "science," concerned withthe formulation of general principles.

There are often quarrels over what kindsof inquiry can be properly called "scienti-fic." We often hear that contemporarypsychologists, especially those of a be-haviourist bent, have an excessively narrowview of science. Such complaints are us-ually traceable to some current of thoughtoriginating in continental Europe, and it isseldom realised that words like the Frenchscience or the German Wissenschaft, withtheir equivalents in other European lang-uages, do not mean what the word sciencemeans in English. They are best translated"scholarship." For example, in the contin-ental sense, history, art criticism, and eventheology are classed as "scientific" pursuits.In English-speaking countries, the word"science" came to be confined to circum-scribed forms of inquiry that concentrateon observable events and the associationsamong them, with such methodological ap-purtenances as the confrontation of alter-native hypotheses and the segregation ofeffects of different causal factors throughexperimental or statistical control.

Psychologists that purport to be scienti-fic, whether in the strict sense or in thewider sense, have at one time or anotherhad three distinct subject-matters. One con-sists of conscious experiences. Anotherconsists of behaviour. The third con-

sists of mental events and processes, notnecessarily conscious. This seems to be thesubject-matter of psychoanalysis, Chom-skyan psycholinguistics, and some contemp-orary information-processing approaches topsychology.

Whenever one studies the observablephenomena of human or animal behaviour,one is compelled to refer to unobservableevents and processes inside the organism.But throughout history, there seems to havebeen a division between people who areprimarily interested in what is experienced,whether exteroceptively or interoceptively,and those who are inclined to attribute agreater degree of reality and importance tohidden entities whose existence is inferredfrom what can be experienced. This divi-sion between positivists and realists hasperturbed psychology from time to time, andthe present seems to be one of the times. Itis an issue over which many of the greatestminds have exercised themselves, withoutachieving any resolution, from the disagree-ments that divided Parmenides and Platofrom the Atomists and Epicureans throughthe medieval debates over the problem ofuniversals and the disputes between theRationalists and Empiricists in the 17th and18th centuries. The only new contributionto be found in recent psychological litera-ture is the practice of bombarding those onthe other side of the fence with epithetslike "trivial," "superficial," "simplistic," and"naive" (e.g., Chomsky, 1968, passim). Wemust leave it to John Q. Graduate's gen-eration to decide whether this innovationconstitutes an intellectual breakthrough orenhances the dignity of our discipline.

The legitimacy of having some peopledevote themselves to the scientific study ofbehaviour is nowadays often challenged,and many psychologists show an embar-rassed reluctance to acknowledge that thisis their speciality. Chomsky's argumentthat to define psychology as the science ofbehaviour is like "defining natural scienceas the science of meter readings" (1968, p.58) has proved curiously seductive. Deese(1972, p. 12), for example, goes one betterand says that it is also like defining chem-istry as "the science of observing changesin the colour of paper."

These analogies are singularly ill-chosenon at least two counts. First, meters, like

74 D. E. BERLYNE

litmus paper, would not exist if there wereno natural scientists. Secondly, nobody inhis right mind is interested in meter read-ings lor their own sake. The physicistexamines them in the hope of learning aboutthe ultimate, imperceptible constituents ofmatter or about other perceptible pheno-mena, such as thunderstorms and spectrallines. Human behaviour, in contrast, exist-ed long before there were any psychologists,and the troublesome practical and socialproblems that it causes, apart from othernatural sources of interest in it, make iteminently worthy of our attention.

Deese, elsewhere (p. 99), refers to the"realisation that behaviour is only the out-ward manifestation of what really counts."What really counts is presumably what goeson in people's minds, what they think andfeel. But this surely depends on circum-stances. Let us suppose that treatment bya clinical psychologist or a psychiatristcaused a patient to behave with perfectrationality and to produce all the verbaland facial expressions of euphoria, whilestill feeling anxious and depressed. Thepatient might well regard the outcome asunsatisfactory. But if warfare and violentcrimes, acts of prejudice and acts of in-justice, continued to be committed by peoplewhose minds were devoid ot animosity orerroneous belief (as seems to be more orless the case with some psychopaths andsociopaths), we could hardly feel that theessentials had been dealt with.

Perhaps the most bizarre reproof of allthose that are leveled at the investigator ofbehaviour is the frequently encountered as-sertion that what he is doing is severelylimited and that he leaves out a great deal.Surely, it is in the nature of any branch ofscience or any intellectual activity to belimited and to leave out a great deal, withthe exceptions of Aristotelian and Hegelianphilosophy. It has often seemed unfair to methat botanists are spared some of the cen-sure that falls to the lot of the psychologistconcerned with behaviour. To read botani-cal writings, one would think that the worldis one large greenhouse. They leave onewith the impression that animals and min-erals either do not exist at all or are ofminor importance. Animals play the brief-est of walk-on parts when they nibble atvegetable matter or when their carcasesprovide nutriment for saprophytic fungi.

The mineral world is absent except when itprovided substrates for sessile plants. Thewhole extraterrestrial cosmos is simply asource of sunlight to support photosynthesis.But in his off-duty hours, the botanist ispresumably as inclined as any man to ad-mire the starry sky or to take his childrento the zoo. He knows that the study ofplant life, or even of a small specialisedniche of it, is enough to occupy a busy life-time. He does not have the time to in-vestigate fundamental problems in zoology,geology, and astrophysics, as well as inbotany, and, if he tried to do so, it wouldall be done badly. But he is not preventinganybody else from pursuing these other con-cerns.

John Q. Graduate Student will certainlyfind much to astonish and perplex him as hedelves into the methodological controversiesof mid-20th-century psychology. And, as heturns more specifically to the critical com-ments that behaviourism and neo-behaviour-ism attracted in the early 1970's, his cre-dulity will be strained still further. Forexample, he will find behaviourism fre-quently contrasted with cognitive theory,generally to its detriment. But, if there isanybody who is entitled to be called thefather of cognitive theory, it is surely Tol-man, who was an avowed behaviourist.

John will find the following statement,from an article by Powers (1973, p. 352), farfrom unique: "Behaviourists have rejectedpurposes or goals in behaviour because ithas seemed that goals are neither observ-able nor essential." The point, it must benoted, is not that behaviourists have givenunsatisfactory accounts of purposes andgoals but that they have refused to counten-ance them at all. John will, of course, bewell aware that the word "purpose" appearsin the title of Tolman's book (1932) andthat many of Hull's theoretical writings(1930, 1931, etc.) were devoted precisely tothe analysis of purpose and of goal-seeking.

Locke (1971) questions Wolpe's behaviour-ist credentials because his therapeutic tech-niques presupposes the patient's ability tothink and image, which are conscious pro-cesses. Yet, behaviourist analyses of con-sciousness include two long articles byLashley (1923), briefer treatments by Tol-man (1932, Ch. XIII) and by Dollard andMiller (1950, Ch. XII), and a major book byJ. G. Taylor (1962). Even Watson's post

BEHAVIOURISMf COGNITIVE THEORY? 75

tion, with its elaborate theory of thinkingand its epiphenomenalism, acknowledgesthe dependence of at least some behaviouron internal events that are accompanied byconsciousness.

John's puzzlement over these and similaroddities will bedevil him until he ultimatelydiscovers two historical quirks that seem tohave been in large measure responsible forthem. One is the failure of surprisinglymany polemicists to realise that Watson'sbehaviourism faded out by about 1930 andwas replaced by a variety of neo-behaviour-isms differing from it quite radically (seeBerlyne, 1968). The second is the widespreadtendency during the late 1960's and the early1970*8 for behaviourism to be equated withthe views of B. F. Skinner, or more fre-quently, with some not altogether accuraterepresentation of Skinner's views. Thingsreached the point where any objection toSkinner was regarded as a refutation ofbehaviourism and any attempt to defend be-haviourism was assumed to mean agree-ment with Skinner's assertions. Yet, Johnwill have gained enough historical per-spective to recognise Skinner as, in manyways, a highly unusual behaviourist and anextremely atypical neo-behaviourist. He willfind the widespread notion that Skinner andbehaviourism are synonymous a verystrange one.

HUIXFinally, John's historical researches will

bring him up against the crowning and mostbaffling mystery of all, namely how onearth anybody in 1975 could have had thenerve to produce a paper with the title ofthis one. John will know that Hull and hisassociates were responsible for a neo-behav-iourist current that was at one time muchmore influential than any other and cer-tainly more ambitious. It was marked bythree stages. First, its finest moments arefelt by many to reside in Hull's theoreticalpapers of the 1930's and the Yale group'sforays into social psychology and psycho-pathology in the 1940's. Next, Hull (1943,1952) concentrated on the construction of anelaborate, systematic theory of behaviour.And in the final phase, beginning in Hull'slast years but embodied particularly in thecontributions of Spence (e.g., 1956) and hisformer students after Hull's death, a fewspecific learning situations were investi-

gated with a view to maximum quantitativerigour. Some commentators feel that thisfinal phase threw out the baby and kept thebath water. Bath water is certainly not tobe disparaged. Civilisation would be at alow ebb without it. But there are more in-teresting commodities, including babies.

The reasons why Hullian neo-behaviourismhad largely fallen out of favour by thefourth quarter of the twentieth century willnot be hard to find. Hull's postulate systemturned out on examination to fall woefullyshort of his ideals of consistency and un-ambiguous testable prediction (see Koch,1954). Considering what Newton had ac-complished with three laws of motion andone principle of universal gravitation, hishope of accounting for the whole of mam-malian behaviour with 16 to 18 postulatesmay not have seemed unreasonable at thetime. But a generation later, the whole en-terprise appeared ingenuous. Psychologicalprocesses, it became evident, do not possessthe manifest tidiness of the Italianate gar-den or the Keplerian solar system but theluxuriant hidden order of the tropical rainforest or the central nervous system. Hull'sattempts at experimental control of learnedbehaviour have been surpassed by Skinnerand his associates and his attempts attheoretical rigour by the mathematicallearning modellers and exponents of com-puter simulation. Many experimental find-ings have come to light that his line ofthinking would be hard pressed to encom-pass. And the weight of evidence has gra-dually tipped the balance against his moti-vation theory and his conception of rein-forcement as drive reduction (Berlyne,1967).

So, to advocate a return to Hull, as mytitle might suggest I am doing, would be toinvite ridicule. His work grew out of aparticular conjuncture of historical circum-stances. It played its part in the develop-ment of our discipline, but there were goodand clear reasons for its supersession.

My title is, however, easily explained. Itstems from two common human fallings,namely a tendency to treat serious mattersfacetiously and a tendency to lose control -ofone's language at times of exasperation.

But I would like simply to point out threefeatures of Hullian behaviour theory thathave vanished from the psychological scene

76 D. E. BERLYNE

and to suggest that this is a pity. The re-ference is to Hull's earlier rather than laterwork, to the questions he asked rather thanto the answers he gave, and to his aimsrather than to the ways in which he soughtto fulfil them. These features are absentalike from Skinnerian radical behaviourismand from the writings of most present-daycognitive theorists and humanistic psycho-logists.

IntegrationFirst, one of Hull's prime objectives, in

his heyday, was integration. His theorydrew sustenance not only from Watson, Pav-lov and Thorndike, its lineal ancestors, butalso from psychoanalysis, from the main-stream and offshoots of Gestalt psychology,and from other sources. He intended it toencompass not only all forms of animal andhuman learning, but all human behaviourfrom child development to thought pro-cesses, from social interaction to psycho-pathology. He aspired to furnish a store ofconcepts and principles that could synthesizethe various fields of psychology and facili-tate the revelation of unity underlying div-ersity that has always been a central fun-ction of science as of art. Such ambitionshave now all but faded away. Skinnerianspreserve the hope of accounting for virtuallythe whole of behaviour with the help of asparse explanatory armoury, but not theeagerness to benefit from what otherschools of thought and experimental ap-proaches may have to offer. Most experi-mental psychologists, including most con-temporary heirs of the Hull-Spence tradi-tion, are content with circumscribed theore-tical models that are meant to fit stringentlylimited situations.

Many writers are expressing uneasinessover the fragmentation of psychology into agravel of esoterically narrow specialities.But the remedy that is all too often pro-posed is the wrong kind of integration,namely the combination of the empirical in-vestigation of behaviour with other acti-vities, such as the study of conscious ex-perience or the composition of inspirationalliterature. Estimable though these otheractivities may be, their aims and methodsare so different from those of the empiricalinvestigation of behaviour that attempts atamalgamation can only work to the mutualdetriment of alL

Some of the factors that discourage ef-forts after synthesis and integration areclear. Many have been frightened off bythe grandiose failure of Hull's system build-ing, as of the equally comprehensive at-tempts of his predecessors like Wundt,Freud and McDougall. Data and theoreticalideas have proliferated so feverishly thatnobody can have the time to absorb morethan a small fraction with accuracy, letalone the panoramic vision required forsuccessful theoretical synthesis.

And, then, there is the bogey of "re-ductionism." A symposium on the short-comings of reductionism was published afew years ago under the editorship of Koest-ler and Smythies (1969). In looking throughthis book, I have not been able to find anystatement of precisely what the term de-notes. But one is left with a distinct im-pression that it means something wrong andevil, and many texts and instructors areevidently transmitting this impression to awhole generation of psychology students.

The term "reductionism" seems to standfor several distinguishable positions in re-cent philosophical and psychological litera-ture. Sometimes, it is a synonym for posi-tivism, the view, which was discussed a littlewhile ago, that scientific theories boil downto, or ought to boil down to, statementsabout observable events. Secondly, theword is sometimes applied to the view,which many people still feel it necessary toinveigh against, that every stimulus ele-ment, response or intraorganismic eventacts in complete isolation from any othersthat are occurring at the time time. Afterthe belabourings it has received at thehands of the Gestalt school and of manyearlier and later writers, this is surely thedeadest of Victorian draft horses, if indeedit ever lived. Pavlov is perhaps held upmore often than anybody else as the arch-spokesman of "reductionism" in this sense,which is ironic. Pavlov became famous,long before he worked on conditioned re-flexes, for his work on circulatory anddigestive processes (for which he wasawarded the Nobel Prize for Physiology orMedicine in 1904). The main innovativefeature of this work was his advocacy of invivo experiments with chronic preparations,i.e., of studying the functions of particularorgans or systems in the intact animal. Heinsisted on this kind of technique (which

BEHAVIOURISM? COGNITIVE THEORY? 77

was also used in his later research on"higher nervous activity") because he main-tained that the activity of any part of theorganism is profoundly affected by what isgoing on in other parts and by the overallstate of the organism (Asratyan, 1949)!And surely, every undergraduate whosestudies have carried him as far as analysisof variance has discovered that independentvariables can have significant interactions!

But in the sense that is most pertinenthere, "reductionism" seems to mean whatused to be called the "unity of science" andregarded as an obvious ultimate goal of thewhole scientific enterprise. One distinguish-ed philosopher of science, Nagel (1961, p.338), defines "reduction", in the sense, as"the explanation of a theory or a set of ex-perimental laws established in one area ofinquiry by a theory . . . formulated forsome other domain," Some of the mostcelebrated triumphs in the history of scienceseem to conform to this definition, includingNewton's reduction of celestial mechanics toterrestrial mechanics, Darwin's reduction ofbiology to Malthusian economics, the reduc-tion of much of chemistry to quantumtheory in the twentieth century, and thereduction of genetics and cellular biology tomolecular biochemistry in the last 20 years.

Several philosophers of science, notablyHarae (1972), have recently been emphasis-ing the essential role in science of struct-ural explanations. These are explanationsreferring to relations among micro-compon-ents or, in other words, entities belonging toa more fine-grained level of analysis thanthe phenomena to be explained. They de-pend, in other words, on reduction. Theirpsychological importance is confirmed bythe latest work of Piaget's group on thedevelopment of causal explanation in thechild (Piaget & Garcia, 1971).

MotivationSecondly, as Spence once said, a frequently

overlooked difference between Hull andSkinner is that Hull was really more in-terested in motivation than in learning.Attempts at defining motivation invariablyend up as enumerations of motivational pro-blems or phenomena. Like other writers, Ihave presented my own lists elsewhere(Berlyne, 1965, Ch. 9; 1970, 1973). It mayalso help to bring in the distinctions thathave become apparent between the diffuse

systems of the brain that subserve affectiveand attentional processes and the specificsystems that control the fine details of per-ception and bodily movement. Finally, it isworth bearing in mind that motivationalprinciples form part of what is needed toderive statements about performance fromstatements about competence.

But for our present purpose, the mostinstructive reference point may be the lawgoverning insanity pleas in criminal cases.Forensic practice is undergoing overduechanges in English-speaking jurisdictions.But at one time, there was a great deal oftalk about "irresistible impulse." The keyquestion, it was suggested, was whether theaccused would have done it with a policemanat his elbow.

We ought similarly, I suggest, to form thehabit of asking what a subject would dowithout a psychologist at his elbow. Wehave vast bodies of enlightening literatureon what people see or hear when a psycho-logist tells them what to look at or listento, on the course taken by their thought pro-cesses when a psychologist tells them whatto think about, on what they recall when apsychologist tells them what to reproducefrom memory or has told them what tolearn. I would be the last to insinuate thatresearches with verbal instructions do notilluminate psychological functioning outsidethe laboratory. But they are a little like in-vestigating the flight of birds by launchingthem into the air with a catapult.

An adequate theory of motivation mustidentify the conditions that generate plea-sure and discomfort, that incite people andanimals to action or reduce them to quies-cence, that give something reward value orpunishment value, that determine what goalor purpose will be adopted. It must analysethe part played by these factors in spont-aneous, as well as instructed, behaviour. Itmust have something to say about the formsof motivation, bound up with basic biologi-cal needs, that human beings share with ani-mals, as well as about the uniquely humanforms of motivation that underlie socialinteraction and social organisation, workand play, philosophy, science and art.Some researchers are doing one or other ofthese. Few are doing all of them. Thecompliance with which human beings andcomputers will accept instructions is induc-ing most specialists in cognitive processes to

78 D. E. BERLYNE

ignore motivational questions completely.Skinnerians are gathering copious informa-tion on how the scheduling of reinforcingevents affects behaviour. But they are neg-lecting such questions as why some eventshave reinforcement value while others donot, and how the effectiveness of a rein-forcer may vary with an organism's moti-vational state.

Symbolic ProcessesThirdly, the neo-behaviourists of the Hul-

lian current began to build up a theory ofcognitive processes, although they were notpartial to that particular term, that avoidedthe two untenable extremes that are pre-valent in today's psychology. On the onehand, there are some who overlook the vastand important differences between behav-iour that is governed by beliefs, thoughts,and rationally formulated intentions and be-haviour that is not. Consequently, valueslike freedom, dignity, and reason are slight-ed. On the other hand, there are those whosee cognitive processes everywhere, in de-fiance of the evidence and of Occam's razor.Enthusiasm for the cognitivist cause is alltoo often accompanied by uncritical ad-vocacy of psychophysical interactionism, offreewill, and of vitalism, in utter disregardof the difficulties that have been found withthese positions over the centuries. Anybodyis, of course, entitled to uphold these posi-tions if he can point to deficiencies in thecriticisms that they have repeatedly attract-ed. But he is surely not entitled to writeas if the counter-arguments had never beenbrought up at all.

The contention that behaviour depends oncognition encourages neglect of the need tostate explicitly the principles that lead tothe one from the other. What may seemreasonable assumptions about the connect-ions between what people say and whatpeople do can be highly dubious. Specialistsin social psychology (Fishbein & Ajzen,1974), in market research (O'Brien, 1971),and in psychological aesthetics (Berlyne,1971, 1974), are beginning to realise thatnonverbal responses are not always relatedto verbally expressed judgments in the waysone would expect.

Above all, both overestimation and under-estimation of human rationality and abilityto act autonomously, regardless of environ-mental influences, can have undesirable

social and political effects. If we fail toappreciate the contrasts between behaviourresulting from informed, rational convictionand behaviour that reflects unthinking, un-critical submission to environmental andsocial pressure, we may weaken resistanceto the blandishments and manipulations ofadvertisers, politicians and partisans of ob-scurantist ideologies. But if, on the otherhand, we do not face the evidence that muchbehaviour is automatic and unthinking, wemay do too little to guard against thehuman weaknesses, whether inherited orlearned, that conduce to guillibility, destruct-iveness and cruelty.

We need theoretical analyses that takeaccount of the immense, socially and biolo-gically crucial differences between thoseforms of behaviour that are voluntary, rat-ional and accompanied by awareness andthose forms of behaviour that are not. Weneed defining criteria that enable thesekinds of behaviour to be differentiated byan external observer. We need to distinguishthe occasions when Occam's razor leaves usno grounds for inferring the interventionof cognitions or thought processes and theoccasions when we are compelled to assumethat they are at work. Finally, we need torelate the higher, characteristically humanforms of behaviour to the lower and moreprimitive forms from which our psychobio-logical point of view compels us to assumethat they have developed.

Hull and those influenced by him built upa conception of mediating symbolic pro-cesses that was on its way to fulfilling theserequirements (see Berlyne, 1965). It merg-ed smoothly with the analyses of signs andsymbols that American philosophers of thesemiotic movement, from Charles SandersPeirce to Charles Morris, had worked outover the decades. It also had much incommon with the theories of ideationallycontrolled behaviour produced by Piaget andby Soviet psychologists under the joint in-fluence of Pavlov's writings on the secondsignal system and Vygotski's research onthe intellectual development of the child.Like the Hullians and the semioticians, theseEuropean theorists have viewed symbolicmediators as derivatives of overt acts,rather than wrapping them in impenetrablemystery, thus paving the way for a psy-chology that is both humanistic and scien-tific.

BEHAVIOURISMf COGNITIVE THEORY? 79

Since the decline in popularity of Hull'sneo-behaviourism, there has been growingevidence that he was on the right track inthese respects. Several lines of researchare beginning to call unmistakably for amulti-level view of learning and behaviour,according to which behaviour-controllingmechanisms of increasing complexity aresuccessively superimposed on one anotherin the course of evolution and in the courseof individual development. These mech-anisms obey common principles to some ex-tent and distinct principles to some extent.And symbolic or cognitive functionsare found only at the highest levels.Support for such a multi-level viewcomes from data on the ontogenetic deve-lopment of learning that have been re-viewed with reference to animals by Thomp-son (1968) and with reference to children byWhite (1965), from data on the successiveappearance of different forms of learning inthe course of evolution that have been re-viewed by Voronin (1972) and especially byRazran (1971), and from data on the suc-cessive involvement of different levels ofthe brain in the course of a single piece oflearning that have been gathered by Olds(1973).

ENVOI

But as I explained earlier, it must not bethought that I am recommending a back-to-Hull movement. Psychologists as a groupare extremely fashion-conscious. Some readtheir technical journals in the same spiritas other people read Vogue. That manydisputes now dividing psychologists areessentially rehashes of debates that havegone on for centuries, or in some cases formillennia, is hardly an original observation.In psychology, as in clothing, there is alimited number of possibilities. Nether gar-ments must be based on the trousers prin-ciple, the skirt principle, or the loin-clothprinciple, and, in each case, there is a finitenumber of discriminable graduations be-tween floor length and zero. There is con-tinuous oscillation among the possible alter-natives, but there has to be some passageof time before what was once grotesquelyfrumpish can reappear as the refreshinglyunconventional.

And then, there is the compelling argu-ment used by Freud (1912) in his article onwhy the psychoanalyst should not take notes

during the therapeutic session: if the pat-ient has said anything of any importance,the therapist is bound to be reminded of itsooner or later.

REFERENCESAstratyan, E. A. /. P. Pavlov. Moscow:

USSR Academy of Sciences, 1949. I. P.Pavlov. Moscow: Foreign Languages Pub-lishing House, 1953.

Berlyne, D. E. Structure and direction inthinking. New York: Wiley, 1965.

Berlyne, D. E. Arousal and reinforcement.In D. Levine (Ed.), Nebraska Symposiumon Motivation, 1967. Lincoln, Nebraska:University of Nebraska Press, 1967. Pp.1-110.

Berlyne, D. E. Behaviour theory as per-sonality theory. In E. F. Borgatta & W.W. Lambert (Eds.), Handbook of person-ality theory and research. Chicago: RandMcNally, 1968, Pp. 629-690.

Berlyne, D. E. Motivational problems. InJ. Linhart (Ed.), Proceedings of the Inter-national Conference on Psychology ofHuman Learning. Prague: Institute ofPsychology, Czechoslovak Academy ofSciences, 1970, Vol. 1. Pp. 155-176.

Berlyne, D. E. Aesthetics and psychobkhlogy. New York: Appleton-Century-Crofts,1971.

Berlyne, D. E. The vicissitudes of aplo-pathematic and thelematoscopic pneuma-tology (or The hydrography of hedonism).In D. E. Berlyne & K. B. Madsen (Eds.),Pleasure, reward, preference. New York:Academic Press, 1973. Pp. 1-33.

Berlyne, D. E. (Ed.). Studies in the newexperimental aesthetics: Steps towardsan objective psychology of aesthetic ap-preciation. Washington, D.C.: Hemis-phere, 1974.

Boneau, C. A. Paradigm regained? Cog-nitive behaviourism restated. AmericanPsychologist, 1974, 29, 297-309.

Briskman, L. B. Is a Kuhnian analysis ap-plicable to psychology? Science Studies,1972, %, 87-97.

Broadbent, D. E. In defence of empiricalpsychology. London: Methuen, 1973.

Chomsky, N. Language and mind. NewYork: Harcourt Brace, 1968.

Deese, J. Psychology as science and art.New York: Harcourt Brace, 1972.

Dember, W. N. Motivation and the cognitiverevolution. American Psychologist, 1974,29, 161-168.

Dollard, J., & Miller, N. E. Personality andpsychotherapy. New York: McGraw-Hill,1950.

Fisbein, M., & Ajzen, I. Attitudes towardobjects as predictors of single and mul-tiple behavioural criteria. PsychologicalReview, 1974, 81, 59-74.

80 D. E. BERLYNE

Fodor, J. A. Psychological explanation.New York: Random House, 1968.

Freud, S. Rathschlage fur den Arzt bei derpsychoanalytischen Behandlung. ZentraVblatt fur Psychoanalyse, 1912, 2, 483-489.Recommendations to physicians practisingpsychoanalysis. In Collected papers, Vol.2. London: Hogarth, 1924.

Gardiner, W. C. An invitation to cognitivepsychology. Monterey: Brooks/Cole, 1973.

Giorgi, A. Psychology as a human science.New York: Harper & Row, 1970.

Harre, R. The philosophies of science. Lon-don: Oxford University Press, 1972.

Hull, C. L. Knowledge and purpose as habitmechanisms. Psychological Review, 1930,37, 511-525.

Hull, C. L. Goal attraction and directingideas conceived as habit phenomena. Psy-chological Review, 1931, 38, 487-506.

Hull, C. L. Principles of behavior. NewYork: Appleton-Century, 1943.

Hull, C. L. A behavior system. New Haven:Yale University Press, 1952.

Koch, S. Clark L. Hull. In W. K. Estes,et al., (Eds.), Modern learning theory.New York: Appleton-Century, 1954.

Koestler, A. The ghost in the machine.London: Hutchinson, 1967.

Koestler, A., & Smythies, J. K. Beyond re-ductionism: New Perspectives in the lifesciences. London: Hutchinson, 1969.

Kuhn, T. S. The structure of scientific re-volutions. Chicago: University of ChicagoPress, 1962.

Lashley, K. S. The behavioristic concept-ion of consciousness. Psychological Re-view, 1923, 30, 237-272; 329-352.

Locke, E. A. Is "behavior therapy" behav-ioristic? Psychological Bulletin, 1971, 76,318-327.

Miller, C. A., Galanter, E., & Pribram, K. H.Plans and the structure of behavior. NewYork: Holt, 1960.

Nagel, E. The structure of science. NewYork: Harcourt Brace, 1961.

O'Brien, T. Stages of consumer decisionmaking. Journal of Marketing Research,1971, 8, 283-289.

Olds, J. Brain mechanisms of reinforce-ment learning. In D. E. Berlyne & K. B.Madsen (Eds.), Pleasure, reward, prefer-ence. New York: Academic Press, 1973.

Palermo, D. Is a scientific revolution takingplace in psychology? Science Studies,1971,1,135-155.

Piaget, J., & Garcia, R. Les explicationscausales. Paris: Presses Universitaires deFrance, 1971.

Powers, W. T. Feedback: Beyond behav-iorism. Science, 1973,179, 351-356.

Quine, W. V. O. Linguistics and philosophy.In S. Hook (Ed.), Language and philo-sophy. New York: New York UniversityPress, 1969.

Razran, G. Mind in evolution. Boston:Houghton Mifflin, 1971.

Rowan, J. Review of I. L. Child: Human-istic psychology and the research tradition.British Journal of Psychology, 1973, 6},

Ryle, G. The concept of mind. London:Hutchinson, 1949.

Spearman, C Psychology down the ages,Vol. 1. London: Macmillan, 1937.

Spence, K. W. Behavior theory and con-ditioning. New Haven: Yale UniversityPress, 1956.

Taylor, J. G. The behavioral basis of per-ception. New Haven: Yale UniversityPress, 1962.

Thompson, W. R. Development and the bio-physical bases of personality. In E. F.Borgatta & W. W. Lambert (Eds.), Hand-book of personality theory and research.Chicago: Rand McNally, 1968.

Tolman, E. C. Purposive behavior in ani-mals and men. New York: Appleton-Cen-tury, 1932.

Voronin. L. G. Stages in the evolution ofhigher nervous activity. In L. G. Voronin(Ed.), Analis assotsiativnoy deiatel'nostigolovnogo mozga, No. 1. Moscow: Mos-cow University. Press, 1972. (Englishtranslation in Soviet Psychology and Psy-chiatry, 1973,11, 93-112.)

Watson, J. B. Psychology as the behavioristviews it. Psychological Review, 1913, SO,158-177.

Watson, J. B. Psychology from the stand-point of a behaviorist. Philadelphia: Lip-pincott, 1919.

Watson, J. B. Behaviorism. New York:People's Institute, 1925.

White, S. H. Evidence for a hierarchical ar-rangement of learning processes. In L. P.Lipsitt & C. C. Spiker (Eds.), Advances inchild development and behavior, vol. 2.New York: Academic Press, 1965.