Roger Carlson's Papers
Evolving Knowledge or Evolving Discourse?
Roger D. Carlson and Joel A. Nixon
Whitman College
Presented at the Western Psychological Association
Rocky Mountain Psychological Association
April 1989 Convention, Reno, Neveda
What constitutes human knowledge? Philosophers answer this question in many ways. When we say knowledge advances, by virtue of what do we make this claim? Traditionally, acceptable explanations must meet standards such as parsimony, comprehensiveness, and have heuristic value (Marx, 1963). When a theory or model "fits" the data to be explained well, when it does so simply, and when it promotes research, we consider it to be a "good theory," or a "good explanation." From a perspective that looks at the transactions involved, however, a valid explanation is a functional unit that resolves a question in the mind of a wonderer. To do so, an account must be on the same ordinal level as the query that it answers. For example, imagine a child walking into a behavioral laboratory and seeing a rat in a Skinner box. A psychology professor approaches her and asks her a question: "Why does that rat press that lever?" And she replies "The rat presses the lever to get food because he's hungry." Her explanation is at a level that she can understand and which resolves her own perplexity. At this point, at least in reference to the rat in the Skinner box, she has knowledge about the world. and the world makes sense.
As the little girl grows older, however, she has increasingly complex interactions within the context of a verbal community. She goes to school, reads books, and observes her own behavior and the behavior of others, always integrating more and more information.
Eventually she enters college and once again finds herself in a behavioral laboratory. The same psychology professor, having now secured tenure, asks her the same question: Why does that rat press that lever? Her reply is much different than it was fifteen years ago: "The rat was deprived of food for 24 hours. He was then reinforced with food pellets for successive approximations of lever pressing behavior until the desired outcome was established. The lever pressing behavior is maintained on a fixed ratio schedule of 15." Her explanation is still at a level that she understands and which resolves her own perplexity. She feels that she knows the world and the world makes sense to her. And we would all probably agree that her knowledge had increased. She would certainly think so. But how much closer to the absolute truth of the matter is she? From an ontological perspective, she has only just begun to explain the rat and its behavior. How does reinforcement work? What about drive-reduction theory? Does the animal think? Could divine intervention be involved? In fact, the relationship between her explanation and the absolute truth, if there is any absolute truth, is ultimately unknowable.
This paper attempts to answer the question of whether or not knowledge evolves in such a way that the "truths" of explanation stand independent from mere rhetoric, discourse, or communicative changes in the social domain of explainers and listeners. In other words, can it be argued that knowledge somehow stands apart from the domain of language use, and thus denotes a kind of "pure" truth, independent of the linguistic interactions and conventions of the explainer and the listener? When the knowledge of a society "advances," in what sense, if any, does this societal knowledge approach absolute truth? In many ways, a society is like our little girl. As it matures, observations and transactions within a verbal community render increasing amounts of information which are integrated and discussed. As information increases, the old answers to the questions of life seem inadequate. Accordingly, new explanations arise to resolve the consensual perplexity of the verbal community. When general perplexity is resolved, we say that knowledge has advanced. But is this consensual knowledge any different from the personal knowledge of our little girl? Is it any closer to absolute truth? Ontologically speaking, who knows?
Take, for example, the concept of intelligence. Early psychological questions concerning the nature of intelligence were answered by unitary explanations such as Spearman's "g" factor. Over the course of time, multiple notions of intelligence, such as Guilford's, or even more recently, Sternberg's and Gardener's, have been gaining acceptance. Is there an absolutely pure and true model of intelligence toward which psychology ever draws nearer? Or, do we just continually package it differently, encapsulating it in more and different ways as the societal applications of what we call intelligence change?
The philosophical works of Ludwig Wittgenstein and Michel Foucault struggle with issues such as these. For instance, Wittgenstein, in his later philosophical work (1953) argued against the possibility of having a private language. Any language, in the usual sense of the concept, must communicate, and thus involves interpersonal interaction. Linguistic conventions enjoin language users in a verbal community, and thus, people share the criteria for the appropriate use of a particular unit of language. Hence, we infer from the work of Wittgenstein that the phenomena of "explanation" and of "understanding" lie within the social domain rather than in any domain of absolute truth. Knowledge, from this point of view, is inherently consensual and transactional. Thus, to use our example, when referring to "intelligence" and modifications in the "intelligence" concept, Wittgenstein would assign the locus of change to the language using practices of people and the societal needs which necessitate a modification in the concept of intelligence.
Intelligence tests were, in fact, first developed for pragmatic reasons. When one develops an intelligence test, a question of validity also emerges, namely: "What do intelligences tests test?" The obvious answer must be "intelligence." The trouble with this conception is that intelligence, which is initially a qualataive commentary on behavior, is now viewed as a thing to be quantified. That is, instead of thinking of intelligence in the way that is implied by the sentence, "John solved his problem of having two dates on Saturday night intelligently," we become concerned with how much intelligence John has. Viewing intelligence as something that is quantifiable has obvious practical and political implications.
Foucault, in his book, The Order of Things (1970), likewise seems to construe knowledge as an inherently social/political phenomenon rather than something absolute which progresses as science modifies it. For Foucault, social and political advantages occur by creating particular discursive practices. Foucault interprets language as the idioms and taxonomy of a society and it thereby allows for the treatment of particular people in particular ways. For example, by dichotomizing the world into distinct groups such as students and professors, mental patients and doctors, and those with high intelligence and those with low intelligence, certain practices are condoned which make the former subservient to the latter in each of these cases. Hence, we infer from Foucault that knowledge, and the taxonomy of that knowledge, in a particular society, at a particular time is instrumental in that it implies certain specific qualitative relationships between people. Thus, to return to our example, when referring to intelligence and modifications in the intelligence concept, Foucault would say that change occurs in response to socio-political power relationships that evolve as a function of society's use of the idioms and taxonomies of language. In other words, "knowledge" evolves in response to socially prescribed criteria. What we come to know generally is what we want or need to know, regardless of that knowledge's position relative to any absolute truth.
What are we to make of knowledge which evolves or changes? Suppose a person who once used and understood "naive" explanations and found them satisfactory, later went on to use "sophisticated" behavioral explanations. Would we say that she now has a more "sophisticated" explanation? If so, what are we to say if this person then goes on to decide that a cognitive explanation of the rat's bar-pressing is even more satisfactory than the behavioral account? Has the person's knowledge progressed or regressed? Or, more radically, what if a rigorous, experimentally sophisticated approach to understanding the development and acquisition of knowledge was later displaced by a more humanistic explanation? Is this progress or not? Probably the most which could be claimed is that the language user's criteria for explanation and understanding have changed. What is now satisfactory as explanatory was not satisfactory before, and what was then satisfactory is not now. Which paradigm more closely approximates a true explanation is unimportant. Explainers and the language of explaining change as do the needs and purposes for explanation, but none of this necessarily constitutes movement toward "truth."
The common links between Wittgenstein and Foucault, then, emerge, in respect to extant knowledge, as language and discourse being the transactional medium in which consensus occurs, and, in respect to new knowledge, language and discourse act as functional precursors which set the occasion for new explanations to develop and to be accepted. The criteria for the "goodness" of a model or theory, then, lie in consensual linguistic transactions that arise in response to socio-political phenomena. The proximity to an absolute truth may be perceived as growing ever closer, but this relationship is actually relative and ultimately unknowable. This dramatically contrasts the common science oriented belief that knowledge possesses an inherent value and that new knowledge arises as the pure methodology of science brings us ever closer to the absolute truths of the cosmos. If the claim were made that an absolute and unchangeable truth is approximated, or even attained, Wittgenstein and Foucault would probably say that such a claim is unverifiable and temporary. In as much as human knowledge is concerned, good explanations are those that result in good understanding on the level of consensual validation within the larger sphere of societal wants and needs.
References
Foucault, M. (1970). The Order of Things. New York: Pantheon.
Marx, M. H. (1963). The general nature of theory construction. In Marx, M. H., Theories in Contemporary Psychology. New York: Macmillan.
Wittgenstein, L. (1953). Philosophical Investigations. New York: Macmillan.