Extract From: Gergen, K. (1984). Experimentation and the Myth of the Incorrigible. In V. Sarris, A. Parducci (Eds.). Perspectives in Psychological Experimentation: Toward the year 2000 (pp. 27-42). New Jersey: Erlbaum.

Experimentation and the Myth of the Incorrigible

Kenneth J. Gergen

During the past century psychologists have participated in what might be considered one of humankind's greates, intellectual adventures. They have, in J. L. Austin's (1962) terms, joined in the "pursuit of the incorrigible," or the "always true," a pursuit that has challenged thinkers from Heraclitus to the present. Perhaps the major stimulus for recent pursuits can be traced to philosophers of science who came to believe that it is possible to discern within the variegated activities of the natural sciences a common pattern of acquiring knowledge. As maintained when those rules of knowledge acquisition were properly distilled, the resulting elixir would transform the character of human life. Natural scientists might employ such rules to determine what forms of inquiry were productive and thus accelerate manifold the impressive advances of the centuries preceding. And, within other spheres of inquiry, including the sociobehavioral sciences, the adoption of such rules would insure progress no less significant than the harnessing of electrical energy, the discovery of genetic transmission, or the smashing of the atom. In Bertrand Russell's (1956) terms, it was hoped that one day there would be a "mathematics of human behavior as precise as the mathematics of machines." I am speaking here, of course, of the positivist-empiricist movement toward a unified science, a movement generally committed to the belief that when properly employed theoretical language can act as a representation of the contours of nature and can be constrained through rigorous assessment of such contours. With steady increments in the objective certainty of theoretical language, humans might become the arbiters of their own destiny.

In large measure, it is this context of optimism that has served as the progenitor to the experimental tradition in psychology. Whereas positivist-empiricist philosophy demonstrated the possibility of utopian ends, it was the experimental [p. 28] method in psychology that was to serve as the major means to such ends. It was the experiment that was to enable investigators to trace precise causal sequences among events, to assess magnitude and directionality of effects, to examine the effects of factors both in isolation and combination, and to rule between competing explanations for given phenomena. Over time, concern with this means to knowledge, its improvement and sustenance, has largely replaced the concern with ends. As Wertheimer demonstrates in the initial chapter of this volume, immense strides have been made in terms of experimental proficiency, control of extraneous variables, precision in manipulation and measurement, safeguards over subjects' welfare, and associated statistical procedures. Further, powerful institutional frameworks have developed to enhance both quality and quantity of experimental products. The major periodicals within psychology are largely dedicated to reporting experimental findings, and advancement in the discipline is largely based on one's capacity to contribute to the corpus of experimental literature. Rigorous experimental contributions thus serve as professional talismans, inviting good fortune and warding off the evil spirit of professional failure. However, far less attention has been given to the efficacy with which such means are enabling the discipline to achieve its estimable goals. Are its methods indeed moving the discipline toward objectively secure knowledge?

It is at least sobering to note that in this later respect a number of prominent scientists, moved to review the advances in their domains, have shown considerable dismay. For example, as the editor of a seven-volume study of the state of psychological knowledge, Sigmund Koch (1971) has summarized: "consider the hundreds of theoretical formulations, rational equations, mathematical models of the learning process that have accrued; the thousands of research studies. And now consider that there is still no wide agreement on the empirical conditions under which learning takes place (p. 693)." In the field of personality psychology, Lee Sechrest (1976) has compared the major issues of study over a 10-year period and asked: "Now why have the themes changed? If it were because issues have been resolved, because important phenomena are now so well understood that they no longer merit attention, it would be cause for encouragement -rejoicing perhaps. Alas, on cannot escape the conclusion that investigators ran out of steam, that issues were abandoned, and that problems were never resolved (p. 1)." Similarly, in reviewing the nearly 300 studies on individual versus group risk taking, Dorwin Cartwright (1973) has commented on the: "ever increasing pile of issues in cognitive psychology which we weary of or become diverted from but never really settle (p. 289).

Of course, such discontent may be attributed to the still less than perfect methods at our disposal, to the lack of funding, to insufficiencies in investigatory [p. 29] ingenuity, and the like. As is said, it is a young science and one should not anticipate great strides at this stage. Yet, such undaunted optimism might be difficult to sustain if a serious assessment were made of developments within the philosophy of science since the bold empiricist years of the 1930s. Such assessment indicates a full-scale deterioration of the metatheoretical launching pad from which experimental psychology was thrust into the world. A brief summary of developments proves useful at this juncture.

Positivist-Empiricist Metatheory: The Erosion of Confidence

First, no satisfactory account has been discovered within the positivist-empiricist domain for the emergence of theoretical insight, postulates, or hypotheses. As widely recognized, one cannot induce general principles from observation; scanning the world in itself produces no ideas or theories to test. In effect, the fundamental basis for what we take to be knowledge does not grow logically from the soil of nature itself (i.e., empirical observation) but from some other source. The grounds for suppositions about the world do not thus appear to be objective, but subjective. And if the initial grounds for scientific propositions are essentially subjective, some might even say mystical, what is the warrant for believing that the experiment will restore rationality and objectivity? For after all, what is revealed as fact by the experiment depends on one's framework of interpretation. It is precisely this framework for which the grounds remain elusive.

With the publication of Popper's Logic of Scientific Discovery, it also became apparent that the empirical confirmation of theory could not furnish the basis for a progressive science. The location of confirmations does not act as a significant crucible for a theoretical account. The most important question is whether a given theory can withstand attempts at falsification. Thus, it might be argued, the common attempt to furnish experimental verification in psychology could not itself yield an accumulation of knowledge. Yet Quine (1953), along with Duhem (1906) long before, also cast significant doubt on the process of falsification. As argued, any theory contains a host of unstated, or auxiliary, assumptions. Whenever data are brought to light that appear to falsify the theory as manifest, the theorist can dip into the nether region of unstated assumptions for distinctions or rationale that can be used to extract the venom from the fangs of falsification. Theories in this sense are not vanquished through falsification but progressively elaborated. It is this fact that enables members of the Flat World Society, dedicated to sustaining a belief in the flatness of the Earth, to remain steadfast in spite of what "everybody knows" to be convincing evidence to the contrary.

Quine (1960) has also challenged another shibboleth of traditional empiricist metatheory in his examination of the indeterminacy of theoretical language. [p. 30] Precisely what are the empirical referents for theoretical terms, asks Quine; in order to test a theory one should be able to specify unequivocally the range of particulars to which it applies. But, argues Quine, the relationship between theoretical terms and particulars is inherently cloudy. If I point to an object and say that if it falls into my classification of desk, for example, to precisely what aspects of the object am I referring? Do I necessarily mean its color, the length of the legs, the number of corners at its perimeter? If I removed a leg, would it continue to be a desk? If I burned it, would the embers also constitute a desk? Thus, to speak of desks does not tell the listener precisely about anything in particular.

The argument has further has further been made by Hanson (1958) and others that observation is inherently theory laden; that is, one cannot undertake the process of systematic observation without certain ground rules for what constitutes a phenomenon. Such ground rules ensure that the evaluation of a theory proceeds in ways that are already theoretically constrained or contaminated. In this way if theories determine what counts as data, competing theoretical accounts may be empirically incommensurable. The phenomena supporting one theory simply do not count as phenomena relevant for its potential competitors. Thus, for example, in psychology, one may be unable to make empirical comparisons between operant theory and phenomenological theory, as the former fails to recognize the "phenomenological field" as an entity and the latter may hold that "reinforcement contingencies" are irrelevant to action unless they are constituents of the phenomenological field. In this sense the theories are empirically incommensurable.

It is in this context of generalized discontent with traditional empiricist metatheory that the writings of both Kuhn (1962) and Feyerabend (1976) could flourish. As Kuhn argued from a historical perspective, theoretical shifts in the science do not represent steady accretions in knowledge. Theories have not come to represent with increasing fidelity the contours of nature. Rather, the shift from one theory to another approximates a Gestalt shift, in which the world is seen through differing theoretical lenses. On this account scientific knowledge does not accumulate through continued empirical work; theoretical sifts over the centuries are not improvements in degree but alterations in kind. To this, Feyerabend has added his diatribe against any codified methods of the kind proposed by his positivist-empiricist predecessors. For Feyerabend knowledge is best achieved by a form of theoretical and methodological anarchy. Even mysticism is to be given serious attention as a candidate for physical explanation. The works of both Kuhn and Feyerabend have been widely discussed and need no further amplification at this juncture. Both have proved vulnerable to wide-ranging criticism. However, of particular significance is the fact that their voices have been among the loudest on the contemporary scene. If positivist-empiricist philosophy continued to demand broad confidence, such audacious voices would be reduced to a murmer. The broad audiences that they command are thus salient [p. 31] indicators of the deteriorated condition of the metatheoretical base on which rests the experimental research tradition in psychology.

Yet, one might continue to reply, in spite of the pervasive criticism of the traditional views of scientific activity, and in spite of those nagging doubts arising within the discipline itself one might hold steadfast to the hope for progress through empirical test. After all, the wide-ranging disillusionment in philosophical circles has scarcely had a dampening effect on the physical sciences. Perhaps the early account of scientific progress was inaccurate or misguided; perhaps the next time around philosophers will get it right. In the meantime natural scientists have continued their efforts undaunted, and the result has been an impressive array of technological advances. Let the common faith not be shaken, it may be argued; if we but continue along the route of rigorous, critical probing of human activity, we may ultimately hope for results of more telling significance.

Experimental Science and the Problem of Phenomenal Instability

There are few who would care to play out the heroic role implied by these latter sentiments more than the present author. There is a certain existential enchantment in launching oneself into the open space of a groundless faith. Yet, several concerns that have emerged within my own work over the past decade militate against such blind commitment. These are not the primary concerns of philosophy of science more generally but ones that grow out of an appraisal of the specific problems encountered in studying human activity. It seems to me that there are certain peculiarities of our craft that, if not in kind, certainly in degree, separate most investigation within the sociobehavioral sciences from that undertaken in the natural science domains. Most of my thinking about these problems has emerged within the context of social psychological research devoted to such topics as aggression, moral behavior, emotion, attitude change, rule following, personal dispositions, and a host of cognitive processes relevant to human interaction. ....

.... On the level of momentary action one's body is in continuous motion; patterns of action are undergoing continuous alteration and certain recurring patterns dictated by biological necessity. Patterns of inhalation, ingestion, and excretion, for example, must occur on a periodic and repetitive basis. However beyond a handful of essential functions, biology seems to [p. 32] furnish the individual with the capacity for immense variability of action. Or to look at it in slightly different form, biology seems to place limits on human functioning, but within these limits there is latitude for infinite variation. Change also seems a paramount feature of life across periods of history. As Feyerabend (1976) (borrowing from Lenin) characterized it: "Accidents and conjectures, and curious juxtapositions of events are the very substance of history, and the complexity of human change and the unpredictable character of the ultimate consequences of any given act or decision of men its most conspicuous feature (p. 17)." If such characterizations seem compelling, then the traditional belief that, when properly conducted, scientific research enables us to proceed ineluctably toward the truth is rendered problematic. One may seek to comprehend the past with increasing clarity and to grasp contemporary pattern in transition. However, regardless of methodological precision one does not thereby accumulate knowledge in the traditional sense or increase one's capacity to predict and to control in the distant future.

As this line of argument was developed with relevance to the range of human activity, one is moved to question its generality. In what degree does it apply to the kinds of processes of interest to investigators in learning, cognition, developmental, and clinical psychology? No blanket judgments appear possible in this matter; the range of topics treated in those domains is immense and study of historical alteration is limited only to a few specialized domains (e.g., concept of childhood, hysteric symptoms, madness). However, one may employ a broad rule of thumb that would furnish an approximate answer in each case: To the extent that any pattern of activity is not demanded by the structure of the nervous system, such activity may be considered historically situated.... One's possible range of responses in each case would seem limited only by the human imagination. From this perspective, empirical methods such as the experiment take on the character of historical [rather than transhistorical] markers (Gergen, 19873, 1978). Except in delimited cases, these methods primarily serve to inform us about various aspects of contemporary life. Whether the experiment is a particularly effective vehicle for furnishing such information may be considered moot. The essential problem is the extent to which it is fruitful to adopt a theory of knowledge based on the stability or reliability of events in confronting creatures whose activities demonstrate precious little of either.

Meaning Systems and Scientific Metatheory

Given this major theme of human change, let me embark on the first of two significant variations. The first has to do with what may be considered a second fundamental characteristic of the organism, namely that of symbol generation and manipulation, along with the associated capacity for employing these [p. 33] symbols in the process of communication. As it is usually argued, the vast share of human activity appears to rely on the manner in which people cognize, symbolize, or conceptualize the world. It is not the stimulus of another's bodily movements to which people respond, as it is generally argued, but the meaning or conceptualization of these movements (Collingwood 1946; Schutz, 1962; Winch, 1958). The same movement of a hand rapidly moving toward one through space may thus be conceptualized, for example, as an act of aggression (a blow), a signal of comradeship (a slap on the back), a practical joke (an attempt at sudden surprise), or a sign of inept gregariousness. The response to the stimulus is not, thus, produced by the stimulus itself but by the particular meanings that the individual happens to assign to it. And, it would appear, people possess the capacity to generate multiple conceptualizations and to communicate these conceptualizations to others.

To the extent that one accepts this characterization of human functioning, support is rendered to the initial argument posed previously; that is, to the degree that human conduct is based on processes of conceptualization or meaning systems, presently existing patterns of activity must be seen as historically situated..... To illustrate, consider the work of Seligman and his colleagues on learned helplessness (Abramson, Seligman, & Teasdale, 1978; Seligman, 1975). As maintained in this case, when people confront situations in which they believe themselves to be helpless or without control over their outcomes, they become depressed and inactive. Yet, to the extent that such patterns do exist, they would appear to depend on one's particular way of conceptualizing helplessness or lack of control.... Depression and inactivity as responses to helplessness or loss of control would not thus appear to be constituent parts of "human nature"; rather they would appear to rely on conventional meaning systems that are subject to all manner of alteration across time.

Yet the problems posed for the behavioral scientist by human capacities for symbolization and communication are even more problematic than suggested thus far. Traditional science is predicated on the assumption that scientific knowledge does not itself alter the phenomena of interest; observation of the planets does not alter their trajectory nor does knowing antibiotics alter [p. 34] their effects on bacteria. It is not being argued in the present case that research on human action need necessarily alter its character; typically investigators attempt to rule out such effects in a variety of ways. However, what is essential in the present case is the possibility that the knowledge resulting from behavioral science inquiry at any given point may so affect the culture that further inquiry would not yield the same results at a later juncture... We offer to the society ways of comprehending human conduct or forms of understanding. As these forms are communicated to the culture, they carry the potential for alteration of human conduct. After immersion in Freudian theory, a parent can scarcely remain unaffected in the treatment of children; after exposure to Schachter's labeling theory of emotions, one is not likely to view others' reports of their feeling with the same degree of credulity, and so on.

.... Such formulations themselves may become absorbed by the culture and affect it in ways that blunt their prognostic capacities. In effect, to the extent that scientists are investigating matters of critical concern to the society, and the society is open with respect to communication, then the scientist may continuously engage in a game of hide-and-seek with patterns of conduct.

The Pseudo Objectivity of Behavioral Description

....

Experimentation in an Interpretive Science

.... [p. 41]

Conclusion

The deep and pervasive yearning for incorrigible truths has led in the present century to a virtual deification of methodology [(i.e., empirical measurement or processing procedures)]. It is this belief in truth through method that has also served as a hallmark of the behavioral movement in psychology. Yet after a long and valiant struggle to vindicate this faith, it is becoming increasingly apparent -both within philosophy and psychology- that methods are not sacred but profane constructions. That which is derived from methods is fully dependent on the interpretive perspectives brought to bear. And whereas the greatest share of professional attention has been diverted to method, the range and potential of interpretive perspectives has languished. Theoretical work must cease to play handmaiden to [emprical-observational] methodology. Rather, it is an auspicious time for reversing the tide of history and placing theoretical work in the center of disciplinary activity. Methods may then serve the ancillary purpose for which they are so well devised.