Science in Christian Perspective

EXPLANATION IN THE 
BEHAVIORAL SCIENCES
LARS L GRANBERG

From: JASA 16 (September 1964): 86-91.

Several weeks ago Senator Margaret Chase Smith of Maine stood before the National Women's Press Club and gave an impressive set of reasons why she should not be a candidate for the office of president of the United States. Among her handicaps, she said, were lack of money, no time to campaign, and no organization. Whereupon she announced she was going to run anyway. I think I understand this a bit. When the invitation came to participate in this symposium, devoted to an inquiry into the nature of scientific explanation, a number of reasons why I should decline came to mind. First, the niceties of this topic are the province of the philosopher of science, and philosopher I am not. (As a matter of fact, my old philosophy professor, Dr. Gordon Haddon Clark, would be more than a little bemused to find me in this role.) To compound the felony, the science of psychology is practiced with fierce intensity by those persons usually designated experimental psychologists, a group who scoff at the scientific pretensions of clinical psychologists. Clinicians in turn tend to divide themselves into those oriented toward experimentation and scientific verification and those tenderminded (soft-headed?) apostles of clinical intuition who feel more at home with the products of poets, philosophers, and theologians than with logic, statistics, and experimental design. This is the bottom of the barrel. So, like Margaret Chase Smith, I considered why I should decline; then, from my couch on the bottom of the barrel, I accepted. This represents a rare opportunity to have to become involved with important issues which I ordinarily ignore, to my detriment.

What I have to say will have to do with explanation in psychology. Some of it may be applicable to anthropology and sociology, but their special problems of explanation I have set aside as outside my sphere of competence.

It is now four score and five years ago since Wilhelm Wundt declared psychology's independency of philosophy and established in Leipzig, Germany, the first laboratory given over to psychological experimentation. Since then psychology has succeeded neither in living without philosophy nor in living with it. E. G. Boring points out that "often the men who cry out most loudly against philosophy in psychology are the men who regard psychology as a system and who write of epistemological matters."^l He continues, 

Ever since their foundation, the journals of psychology in all countries have been weighted down with "theoreticar' papers that are really the expositions of psychologists, untrained to philosophy but writing on philosophical matters. ²

The psychologist thus sketched consciously eschews philosophizing but indulges freely in the pastime by giving it another name. If the psychologist really

Dr. Lars I. Granberg is Professor of Psychology at Hope College, Holland, Michigan. Paper presented at the third biennial science symposium at Wheaton College, February, 1964.

The basic problem is of course the complexity of personality, especially the private aspects of personality. It is a truism that what people say does not necessarily reflect what they do, and what they do does not necessarily show how they feel about either what they say or what they do. Man is also the creature who can watch himself while he is proposing marriage to his beloved and mentally compliment himself on a neat turn of phrase even as he continues to press his case. Self consciousness, the capacity for the kind of divided consciousness which enables one to be actor and spectator simultaneously, along with the rich realm of private experience vastly complicate the application of scientific modes of study to the human person. But it seems that these inherent difficulties have been seriously complicated by psychology's legacy from the eighteenth and nineteenth century; a legacy which has resulted in three conflicting explanatory approaches.

To understand this legacy it is necessary to recall that before Wundt, psychology wag usually called "mental philosophy" and taught by philosophers. One such, the influential Immanuel Kant, laid down a set of criteria for a natural science which called for an a priori mathematical foundation and the possibility of experimentation. In view of these Kant regarded the possibility of psychology's becoming a natural science as dim. Herbart, Kant's successor, was ready to go along with the idea that mental facts could be quantified, but reiterated Kant's doubts about the possibility of experimentation. This moved Gustav Fechner, who already had a well established reputation in physics, to prove him wrong. Building upon Weber's experiments on the sensitivity of the touch and visual senses, he claimed to have succeeded in measuring sensations, albeit indirectly, through a careful measure of the stimuli that aroused them. Fechner's work helped provide impetus to the establishment of Wundt's laboratory.

Wundt's laboratory, then, was founded in a climate of reaction against philosophizing ("arm-chair speculation"); and in a spirit of determination to prove that a psychologist could be fully the scientist which the physicist had become, his students proceeded to study the phenomena of consciousness through introspection.

Barely a generation later, John B. Watson and his followers founded the school known as Behaviorism. In a mood of radical reaction they cast off the introspectionist techniques of Wundt and his followers and asserted that the only proper study for a scientific philosophy is observable behavior, which must be submitted to strict experimental procedures - in particular control of critical variables and verification of results.

The conflict between Wundt's structuralists and Watson's Behaviorists expressed a basic tension between competing philosophical roots which continues to complicate (and enrich) experimental psychology to this day.

Behaviorism finds its origins in the British associationism of Locke and his associates. It sees man as a tabula rasa (blank slate) in need of stimulation from the external world to activate him ("There is nothing in the intellect which was not first in the senses" - Locke³). He is thus reactive rather than active; almost infinitely plastic, and acquires his characteristics through the process of developing conditioned responses. The position is epitomized by Watson's much quoted boast: "Give me any healthy infant, let me control his conditioning during the first seven years of his life, and I will make of him anything you want me to: richman, poorman, beggarman . . ."⁴

The image (scientific model) of man employed here is the machine. The aim of scientific activity is seen as the discovery of regularities of behavior ("laws") through careful observation and measurement. This movement, which comprises the main stream of scientific psychology in America, is reductivistic. It has severely limited its curiosity about man to the observable (public), the regular, that which men have in common (the general) and the quantifiable. These are sought through the analysis of small segments of behavior.

The consequences of this reductivist approach to the person may be seen in the answers which follow in response to the same question:

"What Is a kiss? asks the coed of the Reductivist:

"Two moderately contracted orbicularis oris muscles in juxtaposition," he replies.

"Oh," she replies in a disappointed tone, "there seemed to be more to it than that!"

And what is a kiss when all is done? A promise given under seal - A vow taken before the shrine of memory - A signature acknowledged - a rosy dot Over the I of living - a secret whispered To listening lips apart - a moment made Immortal, with a rush of wings unseen - A sacrament of blossoms, a new song Sung by two hearts to an old simple tune - The ring of one horizon around two souls Together, all alone! ⁵

One would hesitate to claim that a kiss invariably means all these things. Is it not nevertheless indica-

3. G. W. Allport, Becoming: Basic Considerations for a theory of Personality, New Haven, Yale University Press, 1955, p. 7.
4. John B. Watson, Behaviorism.
5. Edmond Rostand, Cyrano de Bergerac, Act M

tive of something that, as Cyrano's rhapsody comes to an end, audiences customarily respond with prolonged, enthusiastic applause while the characteristic response I have observed to the reductionist definition, if any, is a brief, hearty guffaw (mainly from the male section of the audience)? What this indicates admits no simple explanation. My illustration admittedly tends toward caricature. But it does underscore the presence of complex, nevertheless real and important experiential elements in human behavior which have been largely ignored by the dominant tradition in scientific psychology. Valuable as are its emphases on caution, accuracy, careful checking and rechecking, and conservative generalization, this tradition seldom takes into account that in Gordon Allport's words "it is a prisoner of a specific period of culture, and of a narrow definition of science."⁶

Few, if any, even of the most conservative behaviorists would deny the existence of a private world within the individual. 'Their ground rules for science, however, tend to write these off as not suitable for scientific inquiry. One need not, however, minimize the necessity for scientific precision, nor need one look back nostalgically to pre-Wundtian speculation, to be uneasy at the recollection of Aristotle's remark, "It is the mark of an educated man to look for precision in each class of things just so far as the nature of the subject permits."⁷

Many psychologists feel that the associationist. tradition has sought scientific precision at the expense of an adequate conception of man, i.e., it has done violence to the nature of their subject. Characteristically those who think this way turn toward the alternate philosophical root of experimental psychology, German idealism, which, for convenience, Allport calls the Leibnitzian tradition. Leibnitz' response to the Lockian observation that there is nothing in the intellect which was not first in the senses, was "except the intellect itself". This was intended to portray man as an active agent, one who initiates behavior, exercises a selective approach to his environment, and who is modifiable only within limitations. This image of man underlies such diverse psychologies as Gestaltists, Organismic (Holistic), Personalistic, and Existentialist. Representatives of these viewpoints study behavior at the molar rather than segmented level. That is, the whole human in action, or some large portion of behavior as compared with conditioned responses.

Approximately at the same time that the structuralists and the behaviorists were struggling for ascendancy over American academic psychology, Sigmund Freud and his associates were setting the foundations for modern clinical psychology. Freud's early theories were the culmination of a century of attempt's to understand a spectacular and comparatively common form of behavior disorder known as hysteria. This is a disorder, now relatively infrequent, which gives rise to spectacular physical symptoms (blindness, deafness, paralysis).

Freud based his system of explanation on the centrality of striving. He saw men pushed from within by blind, pleasure-seeking impulses and threatened from without by an unconcerned, prohibitive society. Men needed to find defenses against the ensuing anxiety which would allow a modicum of pleasure while avoiding society's wrath. Most attempts at clinical explanation have accepted Freud's paradigm of anxiety and unconscious defenses as the key to understanding behavior. Contemporary clinicians prefer explanations grounded in unconscious dynamics.

We see, then, that contemporary psychology's efforts at explanation of behavior is complicated by three major perspectives. These are by no means totally antithetical. They do, however, represent different moods toward explanation. Each has some tendency to paint the others in the role of villain. Yesterday we heard Dr. Giles observe that what is regarded as good explanation is ultimately grounded in the fact that a community of scientists (e.g., physicists) like it. This suggests that a workable system of explanation within a science presupposes that there is within that science a well developed sense of community, Here the existence of these three moods within psychology, which generate more than a little mutual hostility, appears to be retarding the formation of the needed sense of community.

It appears to me that a constructively eclectic spirit around which this sense of psychology as a scientific community can develop, is needed in order to further good explanation- There are encouraging signs that this attitude is on the increase.

SPECIFIC COMPLICATIONS OF PSYCHOLOGICAL EXPLANATION

Scientific explanation, then, says Scriven, "Requires selecting from among the variables that are involved; those whose activities are unknown to the inquirer and crucial for the phenomena; . . . The explanation won't be an explanation unless the variables are crucial,"B (e.g., to explain soil erosion without giving due consideration to wind direction and velocity and mean annual rainfall is to fail to explain, Proper selection of variables presupposes some reasonably well established general theories (laws) to use as a guide. Explanation becomes a way of showing whether or to what degree new facts fit into accepted patterns.⁹ And at present there is a distressing paucity of general laws.

If a small parenthetical indulgence in self pity may be permitted, explanations are bound to be a good deal more difficult to come by in psychology than in the older sciences. Scriven observes that "we already have common-sensical and well supported explanations

8. M. Scriven, '%xplanations, Predictions and Laws," in Feigl and Maxwell (FAs.) Scientific Explanation, Space and Thae, Minnesota Studies in the Philosophy of Science, Vol. M., p. 211, (italics author's)
9. J. G. Kemeny, op. cit., p. 164 italics mine.

of nearly all the easy cases, and we are therefore left with the problems we have not been able to solve exactly by common sense."¹⁰

Unrealistic expectations make trouble too. We tend to measure our success against physics operating under laboratory conditions and in consequence expect exact predictions. However, in the realm of practical problems, which make up a major portion of the psychologist's province, Scriven maintains that exact predictions or faultless explanations are probably not possible inasmuch as these are rarely achieved by any sciences today. In physics, where practical problems are represented by engineering problems, meteorological problems, and problems in aerodynamics, solutions to problems are frequently compromises and approximations. He cites as an example the problem of determining how far a given missile of a particular shape will travel with a given propellant. This cannot be learned exactly from scale models since neither air molecules nor the critical mass of the propellant can be scaled down. The physicist can, to be sure, give good answers to such questions, He can even develop radio controls (as he has) which get him around the problem of prediction. Generally speaking, however, we in psychology must work with a larger number of critical variables, and are more unlikely to be able to run valid, full scale, repeatable tests. It hardly surprises that psychological prediction or explanation is currently less successful than the engineer, the aerodynamicist and even the meteorologist (% and probably will continue to be for a long time.

Explanation is further complicated by the tendency in psychological research for the data-gathering process itself to bring about or to change particular characteristics. People often change their behavior in the process of being studied (e.g., polling on attitudes toward public issues). And the publication of studies of behavior has been known to change conditions sufficiently to significantly modify or even invalidate the original study. To what degree, for example, did the early Kinsey publications modify the trends they published by 1) providing the predatory with statistics to prove to the reluctant that "Everybody's doing it!" and 2) by promoting the thesis that what is not should be?

This is a problem, but more readily surmountable than others discussed. Possibly the limits of variation may be measurable. One can also moderate this effect at least in some contexts, by the use of indirect measuring devices such as projective techniques. It should also be borne in mind, Nagel observes, that all scientific "laws" are conditional.¹¹ They obtain if such and such conditions exist. If, then, one of the conditions of the law is that actions be based upon ignorance of it, we cannot consider the law in error when actions based on knowledge of it do not conform to it.

10. Scriven, "A Possible distinction between Traditional Sclentific Disciplines and the study of Human Behavior, in Minnesota Studies in Phil. of Sci., 1, p. 334.
11. Nagel, op. cit. p. 487.

Possibly the most plaguing of all is the problem of individuality. Our present techniques are normative. They measure the ways in which people are like one another. Individual uniqueness is missed, and to degree this is crucial our explanation misses the point. A person is more than a nexus of profile percentiles. These leave out much and tell us little about how a particular position on a scale fits into the total pattern of behavior. Finally, the psychologist whose presuppositions concerning man are Christian is disturbed by an approach to the study of behavior that seems to confine man within the realm of nature. How can one adequately explain regeneration, worship, prayer, and fellowship with God as natural phenomena? Moreover, the individual's experience of God, salvation and spiritual fellowship cannot be directly dealt with. Nevertheless his reports of these things and the influence of his beliefs on his behavior can be placed, at least to a significant degree, in the realm of psychological explanation.

CONDITIONS FOR GOOD EXPLANATION IN PSYCHOLOGY

Few things bring such chagrin to the dedicated experimental psychologist as the difficulty of gaining adequate experimental controls. Nagel suggests that this concern is exaggerated. In response to the question whether controlled experimentation is a sine qua non for the achievement of warranted factual knowledge and the establishment of general laws he points out that astronomy and astrophysics did not arrive at its theory through experimental manipulation of the celestial bodies, yet have arrived at well grounded general laws. He goes on to suggest that controlled investigation, which does not involve manipulation of variables, can serve a function similar to experimentation in psychological inquiry. Controlled investigation consists of a deliberate search for contrasting occasions when a phenomenon is either uniformly present or present in some cases and not in others. This is followed by an examination of certain factors discriminated in those occasions to see whether variations in these factors are related to differences in the phenomena. (Goldfarb's studies on the effect of psychological mothering or its absence seems to illustrate this approach.) Nagel regards it as immaterial whether observed variations are introduced by the scientist or are produced "naturally' and observed (discovered?) by him.¹²

Concerning the view of controlled experimentation which requires variation in just one relevant factor, Nagel considers this idealistic and difficult to achieve even in the natural sciences. Desirable as it may be, it can hardly be regarded indispensable. Temperature change and rainfall, for example cannot be varied independently, but statistical analysis allows the investigator to cope with many situations in which he cannot systematically control all but one variable at a time.¹³

Comforting as this may be to the experimental psychologist, his troubles are by no means ended. Statistical analysis of psychological data, Kemeny notes, has been retarded by the lack of development of a mathematics for intermediate range numbers (of the order of 5,000 to 50,000). Small numbers (of the order of 5) can be solved by elementary arithmetic. Large numbers (of the order of 5 million or even billion) are amenable to treatment by calculus. There is at present no equally effective mathematics for intermediate range numbers, so that progress in the behavioral sciences in some measure awaits mathematical progress at this point.¹⁴

The subjective nature of much psychological data creates further problems for scientific explanation. Men think, wish, desire, intend, suffer pain, heat, and cold, They experience joy at the presence of a friend and heartache over personal rejection. They believe or doubt, feel guilty, seek meaning, and erect value-hierarchies. For the most part, these activities are not open to the direct observation of the scientist, but it hardly seems that one can afford to exclude them and still make claims about adequate explanation.

In their zeal to eliminate the excesses of introspectionism the early behaviorists tended to take the position that psychology should be based solely upon public data which could be observed and confirmed by two or more observers. Today's more sophisticated behaviorists recognize the need to deal with introspective reports. They do not, to be sure, treat these as the expression of private inner states. Reports purporting to represent the inner life are classed with other observable activity in which a person engages under certain conditions. As such they form part of the basis for psychological generalizations. This, Zener comments, "shifts the burden of scientific responsibility from the report of the subject to the report of the experimenter."¹⁵ Percy Bridgman presses the behaviorist at this point too: "Once the behaviorist has admitted that introspectional report is a proper subject for psychological study he must also admit, I believe, that in particular the report of his own introspectionings is a proper topic . . .16 Bridgman insists that private (inner) activity not only can, but must be made public. He considers the phenomenon of projection an effective means for unlocking the private world of the person. This gives pause to the clinical psychologist, as we shall see. Bridgman argues, however, that the inevitable result of living in society is that one learns how to interpret words referring to private experience. "When we project", he says, "we understand the action of our fellows by imagining ourselves in his place. This involves the assumption that we and our fellows are sufficiently alike for practical... purposes."¹⁷

14. Kemeny, op. cit. p. 247 ff.
15. K. Zener, "The Significance of the Experience of the Individual for the Science of Psychology," Minn. Studies in Philos. of Sci., 11, p. 367. }
16. Bridgman, The Way Things Are, p. 242 ff. 17. Bridgman, op. cit., p. 243.

In support of Bridgman's contention it is well to recall that the process of projection has a stout ally in the individual's need to communicate himself to others. John Donne is right. No man is an island - even the badly regressed schizophrenic patient. And ff the pathological effects of psychological isolation are well known, so also is the promised bliss of "dialogue", a phenomenon regarded by a troop of contemporary psychologists, philosophers and theologians as the key to self-discovery. I come to know myself, they say, in the struggle to make myself clearly known to another. (cf. Sidney Jourard, The Transparent Self).

The perturbed clinician's response is likely to be that men have not lost their need for "fig leaves" (psychological defenses). Neither do they possess the capacity fully to report their experiences. How does one adequately report all he sees as he watches a sun set in the Rocky Mountain area? Much less describe the total effect of what he is seeing upon himself? These are valid concerns, but for all of them subjective experience is considered to be so essential for an adequate view of man - hence for adequate psychological explanation - that the effort to counter these limitations must continue. Significant recent developments in methodology (Q-technique; projective technique; analysis of verbalization of therapy clients) stand to further effectiveness in ascertaining and properly assessing inner factors.

It would not do to leave this line of discussion without illustrating one of the chronic pitfalls of clinically derived explanation. I refer to the tendency to incomplete explanation so blithely indulged in by more than a few clinicians (to the despair of the science makers among us). An explanation is incomplete when it fails to explain what it purports to because the connection between a given effect and the phenomenon that is supposed to have produced it is inadequately established.¹⁸ Consider one of Freud's explanations. In The Psychopathology of Everyday Life he explains a slip of the pen be had made. On September 20th, shortly after his return from a holiday, he found that he had written "Thursday, October 20th" just below the correct date. This he interpreted as the expression of a wish not to have the next month fly by as quickly as it was going to. Freud's apparent hypothesis here is that when a person has a strong, but perhaps unconscious wish, any slip (tongue, pen, memory, etc.) will represent an expression and perhaps fulfillment of the wish. This is no model of hypothesis-making, but at that it is probably more explicit than Freud would have liked. What makes this an incomplete explanation? Let us for a moment assume that the above hypothesis is true and that Freud had the wish he reports. One can imagine a goodly number of things Freud could have done to express his unconscious wish, any of which should have been equally effective. Why then the slip of the pen? There is nothing in the specific conditions that seems to make it either

18. Scriven, "Explanations, Predictions and Laws," Minn. Studies in Philos. of Science, III, P. 200.

necessary or inevitable. Nevertheless it cannot be rejected for it is not only plausible, it has elements of probability. Hence this is called an incomplete explanation.¹⁹

What constitutes a complete answer, then? There are those who feel that a complete explanation cannot be given in science. Certainly, as Scriven comments, "there is the possibility of indefinitely challenging the successive grounds of an explanation."²⁰ However, within the context of science the word "complete" has acquired a standard use which applies to some explanations and not to others.

A complete answer has been given when the particular object has been comprehensively related to the directions that are understood ... the request for an explanation presupposes that something is understood, and a complete answer is one that relates the object of inquiry to the realm of understanding in some comprehensive and appropriate way. What this way Is varies from subject matter to subject matter . . . and what counts as complete will vary from context to context In a field . . .²¹

At the present time, particularly in clinical psychology, explanation sketches, such as the Freudian example given above, are the rule rather than complete explanations. An explanation sketch is "a suggestion of general outlines of what, it is hoped, can eventually be supplemented so as to yield a more closely reasoned argument based on more explicit and more testable explanatory hypotheses.²² To move in this direction, we need to learn how to get beyond the obstacles discussed above.

WHAT OF THE FUTURE?

Kemeny, taking the historical approach is cautiously optimistic. He reminds us that in the history of science many avenues of research turn out blind alleys during the early stages of a science, and suggests that for a long time to come the really fruitful results in the behavioral sciences are likely to come in areas offering no immediate benefit to mankind. Then he makes the interesting comment that some mathematicians believe that the inspiration from the physical sciences which has for so long given impetus to progress in mathematics has about run its course, but that the next great development in mathematics is likely to be sparked by the behavioral sciences. So much the better for the latter, for, as indicated above, the behavioral sciences are handicapped by deficiencies in mathematical knowledge.²³

19. Colodny, op. cit., p. 15
20. Scriven, op. cit. p. 201.
21. Scriven, op. cit., p .202
22. Colodny, loc. cit.
23. Kemeny, op. cit., p.256 ff.

created for us in psychology. If we do we shall have no need to regret the fact that humans are both complicated and selfconscious and its attendant consequence that psychology will never be lige Newtonian astronomy.

Scientific explanation in psychology is presently at a low level of completeness, tending toward explanation sketches, which are rich in experimental potential, but incomplete as stated.

Greater scientific completeness of explanation in psychology seems to require a more adequate conception of man, a broader view of science; a sense of com munity among psychologists; a large assist from the mathematicians; data-gathering techniques which will not alter the situation in unexpected ways; and a better approach to individuality. Good progress has been made, and there is good reason for optimism,

I should like to beg the indulgence of the Christian public. One of the persistent pressures it places on the Christian psychologist is that he should keep jumping from his empirical data to theological ends. That is, there is a tendency to try to make a given postulate or conclusion in the realm of the person or behavior fit a theological statement far too soon, and to opine darkly concerning the psychologist's orthodoxy if he does not do this. This makes it difficult to build from sparse, moderately validated data to ultimate frames of reference in the solid, step-by-step way that makes for genuine growth in knowledge. It also leaves little margin for the 'trial and error and the inevitable fits and starts that accompany progress. We do see through a glass darkly. We do know in part. All of us. My plea, therefore, is that God's revelation of himself in the Scriptures not be reduced to a Procrustean bed into which data must fit on demand. :

E. G. Boring, Psychology and Science., Selected Papers, NY: John Wiley, 1963

Percy Bridgman, The Way Things Are, Cambridge, Mass.: Harvard University Press, 1959

G. Burnistan Brown, Science: Its Method and Philosophy, NY: W. W. Norton, Inc., 1950

Robert G. Colodny, Frontiers of Science and Philosophy, University of Pittsburgh Press, 1962

H. Feigl and M. Scriven, Foundations of Science and the Concepts of Psychology and Psychoanalysis, Minnesota Studies in the Philosophy of Science, Vol. 1. University of Minnesota Press, 1956

H. Feigl, M. Scriven, and G. Maxwell, Concepts Theories and the Mind-Body Problem, Vol. 11, Minnesota Studies, 1958

J. R. Kantor, The Logic of Modern Science,, Bloomington, Ind.: Principia Press, 1953

John G. Kemeny, A Philosopher Looks at Science, Princeton, NJ: D. Van Nostrand Co., Inc., 1959

Edward H. Madden, The Structure of Scientific Thought: An Introduction to Philosophy of Science, Boston: Houghton Mifflin, 1960