RICHARD
GENETIC
F. K I T C H E N E R
EPISTEMOLOGY,
EPISTEMOLOGY,
NORMATIVE
AND PSYCHOLOGISM*
The genetic epistemology of Jean Piaget raises many important philosophical and conceptual questions. 1 But perhaps the single most important one for philosophy is this: What is the philosophical relevance of genetic epistemology? Needless to say, it has psychological and educational implications, but does it have anything of relevance for contemporary philosophy? Many contemporary philosophers would reply that it has no interesting philosophical implications for contemporary epistemology, for example, since it is merely a pyschological theory about children, whereas others might add that if indeed it does happen to have any philosophical implications, this is because it is not really a psychological theory at all but just another not so cleverly disguised philosophical theory. Thus David Hamlyn says: If I am right, however, my answer to the question, "What relevance has genetic epistemology in the more orthodox sense?", must be that in a certain sense genetic epistemology presupposes a traditional epistemological position. If it has implications for epistemology, it is in that sense and for that reason, and not because of its status as a psychological theory. 2
Such a view about the philosophical relevance of genetic epistemology is, I think, one that many philosophers (perhaps a majority) would endorse. The reasons lying behind such a view seem to be two. First, there is presumed to be a fundamental difference between philosophy and science. What philosophy is may not be all that clear, but at the very least it deals with some kind of conceptual analysis. Science, by contrast, is empirical. Thus, philosophical questions and empirical questions are (in some basic way) different. Secondly (and this may merely be a version of the first or perhaps the ground for it), if genetic epistemology qua empirical theory were relevant to philosophical issues, then some elementary fallacy would be committed. We would, for example, be deducing a normative conclusion (about some epistemological question) from factual information and Synthese 45 (1980) 257-280. 003%7857/80/0452-0257 $02.40 Copyright © 1980 by D. Reidel Publishing Co., Dordrecht, Holland, and Boston, U.S.A.
258
RICHARD
F. KITCHENER
thus would be committing the naturalistic fallacy, the fallacy of psychologism, or the is-ought fallacy. On the other hand, to infer anything about the validity, adequacy or truth of a belief or theory from an examination of its genesis would be equally bad since this would be committing the genetic fallacy. Therefore, either genetic epistemology (and any empirical theory) is begging a philosophical question by presupposing a philosophical view (and that is why it is relevant), or it is not begging a philosophical question but is committing some elementary fallacy. In either case, however, genetic epistemology really has no "legitimate" relevance to epistemology. I believe this entire line of reasoning is wrong, but wrong in important ways. The most important way it is wrong is the assumption that philosophy and science are fundamentally or radically different and thus that conceptual and empirical issues are basically different. Such a view is mistaken, but I am not going to argue against it directly. Instead, I am going to examine the second reason advanced in support of the philosophical irrelevance of genetic epistemology, namely that it either "begs the question" or commits some other fallacy. In Part I, I discuss how Piaget's genetic epistemology has empirical relevance to traditional epistemology. In Part II, I examine the claim that genetic epistemology presupposes some particular epistemology or philosophical norm in general and indicate in what sense this is correct. In so doing, I argue that Piaget's genetic epistemology has more philosophical relevance than Piaget himself is inclined to admit and that this relevance is at least partly normative in nature. Finally, I briefly consider whether any fallacy is being committed and suggest we need to reexamine the fallacy of psychologism, along with the fact-value (is-ought) distinction (especially in the context of epistemology). I. E M P I R I C A L
IMPLICATIONS
OF GENETIC
EPISTEMOLOGY
Genetic epistemology is relevant to traditional philosophical epistemology, according to Piaget, because epistemologists make empirical claims, and genetic epistemology aims to investigate such issues. Ordinarily, however, these empirical claims masquerade as something else, for example as "common sense" knowledge, a disguise that is relatively easy to spot. But more serious is the situation in which empirical claims, masquerading as philosophy, forever go undetected;
GENETIC
EPISTEMOLOGY
259
in this case what results is an endless series of philosophical speculation about what is really an empirical issue. The seventeenth century issue about innate ideas clearly falls into this category, since several of the disputed issues concerned the empirical question of whether something (an idea, concept, or disposition) existed at birth or not. (The Molyneux question is another one that readily comes to mind.) Hume's theory of causality likewise contains an empirical c o m p o n e n t - h i s theory that our idea of necessary connection is derived from the association of i d e a s - j u s t as his overall philosophical theory contains an empirical theory about the nature and workings of the mind, from which several empirical claims follow (e.g., that an idea is a faint copy of an impression). All of this is well known, of course, but what tends to be forgotten is that empirical evidence is available that ought to be weighed when Hume's theory is evaluated. Piaget claims, for example, to have empirical evidence contrary to Hume's account, 3 and several other researchers have empirical evidence in support of Hume's claims (e.g., about ideas being faint copies of sensations). We might also mention Kant's epistemology, which contains empirical predictions as well. One Of the most interesting of these is the following. Kant claims that the structure of the mind (the a priori forms of intuition, the Categories, etc.) is a priori, and this partially means temporally a priori. True, Kant characterizes 'a priori' as "independent of experience" and this phrase is open to being interpreted as "logically independent of experience." But also, quite clearly, Kant believes the a priori forms of intuition (space and time) exist before our experience of objects, 4 and seems to believe the same with regard to the CategoriesJ It would be impossible, Kant thinks, to have "experience" temporally prior to the appearance or operation of these structures. Hence, if we could determine that a human being experiences the world but does not possess the category of (say) causality (or a spatial form of intuition), this would be empirical evidence against Kant. Piaget's research can be interpreted as being fundamentally supportive of Kant's claim, depending upon how we interpret 'experience.' The infant, according to Piaget, experiences the world in some sense but not the way an adult does. Unlike the adult, the infant is a radical phenomenalist-since only sense-impressions e x i s t - and (s)he has no concept or experience of objects as permanently existing, three-dimensional, spatial entities with enduring
260
RICHARD
F. K I T C H E N E R
causal properties. Furthermore, the infant has no self-concept and does not distinguish the self from the world. Thus, the child's experience of the world is radically different from the adult's, and radically different form Kant's notion of what experience is. To this extent, Piaget's research empirically supports Kant's transcendental claims. But Piaget also believes Kant is wrong in supposing the categories do not evolve and develop, but are innate. Kant is unclear about this point and this is due to a fundamental lack of clarity about the entire concept of 'innateness,' a confusion that permeates all discussions of this topic from Locke and Leibniz (on the one hand) to the current views of Chomsky and Katz. Is the concept of 'innate' something that refers to the underlying capacity (disposition, ability, endowment, potentiality) which may require considerable stimulation and exercise before it functions to produce (say) spatial perception or language acquisition, or does 'innate' refer to the actual perception (existence of an idea, linguistic performance) at birth? Kant seems to hold both views, since in his Dissertation and Correspondence he defends the first view about what is a priori and claims that the ground of spatial and temporal intuitions alone is inborn, but the representations themselves are not. In the Critique, however, he seems to maintain the second view: spatial representations lie ready in the mind at the birth of consciousness. 6 What is meant by 'innate' is crucial, since (on the first interpretation) much of Kant's theory can be interpreted as being in agreement with Piaget's genetic epistemology and hence there would be no reason to deny, for example, that the Categories develop and thus (in the second sense of 'innate') are not innate. 7 The point, in any case, is that genetic epistemology does have relevance to Kant's philosophy and especially to the empirical issues arising out of Kant's theory. In fact, I believe it is both correct and illuminating to say that Kant's concept of a subjective deduction (of the Pure Concepts of the Understanding) is precisely what Piaget's genetic epistemology attempts to examine. The subjective deduction is concerned with the subjective (i.e., pyschological) conditions necessary for experience, with the question (as Kant puts it), "how is the faculty of thought itself possible?", s with the generative processes to whose agency human knowledge is due. 9 This could be used to summarize Piaget's genetic epistemology in a nut-shell.
GENETIC
EPISTEMOLOGY
261
Various example of Piaget's claim that philosophical epistemology makes empirical claims could be given from contemporary philosophy. But one of the most interesting examples here is that of "the linguistic theory of the a priori" originally advanced by the logical positivists. This theory, roughly put, is that the "necessary knowledge," contained primarily but not exclusively in logic and mathematics, is due to our conventions concerning the use of language. The "necessity" of certain propositions is thus due to "our determination to use words in a certain fashion" (as Ayer puts it)J ° What exactly this theory is supposed to maintain is not very clear, nor is it very plausible. But that is a story already well told by several other philosophers, u Piaget's criticism of this theory is not philosophical, however, but empirical, for he claims that this theory has empirical consequences that are false. The decisive argument against the position that logical mathematical structures are derived uniquely from linguistic forms is that, in the course of intellectual development in any given individual, logical mathematical structures exist before the appearance of language. Language appears somewhere about the middle of the second year, but before this, about the end of the first year or the beginning of the second year, there is a sensory-motor intelligence that is a practical intelligence having its own logic - a logic of action/2
There is much to criticize in this passage. Piaget is quite frankly not as clear, careful or precise as analytic philosophers would wish, and Piaget thus tends to make vague claims and to employ unclear concepts. What does Piaget mean by 'logic,' for example, and 'logical mathematical structures'? 13 However, although there is much to criticize here, the situation is not completely hopeless, as some individuals seem to think. For example, one philosopher claims that passages such as this one indicate Piaget "displays a fundamental misunderstanding" of Logical Positivism and related viewsfl Since this is a view likely to be held by other philosophers, I want to point out why it is wrong. The fundamental error in this passage (aocording to this criticism) is that Piaget's view suffers from a serious confusion of two distinct senses of 'derived': 'derivedf, meaning 'develops later in the organism,' and 'derived2', meaning 'is true in virtue o f . ' . . . By clearing up Piaget's confused use of the notions 'basis' and 'derived,' we have eliminated the force of his critique of the logicists and the positivistsJ ~
262
RICHARD
F. K I T C H E N E R
That is exactly what this author has not done. First, the author's reliance on the distinction in question is problematic: as Chisholm points out, it won't do to say "is true in virtue of," nor will it do to say "is true solely in virtue of. ''16 Furthermore, what the phrase "is true in virtue of" is supposed to mean is not indicated by the author and for good reason, for it is simply unclear what this "linguistic conventionalism" is, or was, supposed to be. Does "logical truths are true in virtue of linguistic conventions" mean that "necessary propositions" are really contingent statements about linguistic usage after all? Does it mean "necessary truths" are not propositions at all, but linguistic rules? As has become increasingly clear, neither of these formulations will work. What then is meant by this claim? To rest one's whole argument on a phrase ("in virtue of") that is itself so fundamentally vague and unclear is surely less than adequate. One way that has been suggested for explicating this view has been given by Arthur Pap in his thorough and incisive criticism of "the linguistic theory of the a priori." To say "necessary truth is in some sense the product of linguistic conventions" might be to say (according to Pap) "the existence of certain linguistic habits relevant to the use of a sentence S is a necessary and sufficient condition for the proposition meant by S. ''17 Here Pap is giving the theory (contrary to the criticism under discussion) precisely the sense that Piaget gives it, and, again, it can be claimed (although Pap does not) that this is empirically false, since (temporally) prior to the existence of such linguistic habits, according to Piaget, there is a "necessity" (or quasi-necessity) embedded in the "logic of action" (which is an algebra of classes involving seriation, inclusion, etc.). Hence the theory is empirically false. The point at issue might be put in this way. If something (language) is claimed to be a necessary condition for something else (logic), then surely it is being claimed that at all times language is a necessary condition for logic, i.e., (t) (~ Language(t)D -Logic(t)). If we instantiate for t, we get ( - Language(tr,)D -Logic(t,,)). Here then we have an empirical prediction which would be refuted if at time t,, there were no language, but at t,, there were a logic. But this is precisely what Piaget claims he has found. It is really immaterial for our purposes whether Piaget has found what he has claimed or not; that is a question about the evidential status of his claims. The point,
GENETIC
EPISTEMOLOGY
263
however, is this: Piaget's claim that this is (at least partly) an empirical issue remains plausible and, contrary to the criticism under discussion, does not seem to be an egregious error. II. DOES PIAGET'S HAVE
GENETIC
NORMATIVE
EPISTEMOLOGY
IMPLICATIONS?
Insofar as traditional philosophical epistemology makes empirical claims, genetic epistemology, since it studies the development of knowledge in the individual and in the species, is relevant to epistemology. Of course someone might respond that if, indeed, epistemologists make empirical claims this only shows that these philosophers were not careful enough in distinguishing empirical questions from logical ones. But genetic epistemology has no relevance to philosophical epistemology properly conceived and executed, which is purely logical, conceptual or normative. The real issue, it might be asked, is whether genetic epistemology has any normative implications, for this is surely what philosophical epistemology is really all about, is One way Piaget's genetic epistemology might be relevant to normative epistemology would be if it presupposed some normative account (of knowledge) and thus begged important philosophical questions. Thus it would not commit the naturalistic (is-ought, pyschologistic) fallacy, since it would not be going from pure facts to norms, but this is because the norms are already there from the beginning. Two versions of such a claim can be found among philosophers: (1) genetic epistemology presupposes a particular epistemology, and (2) genetic epistemology presupposes certain general epistemic norms. (1) Does genetic epistemology presuppose a normative epistemology? To the question, what relevance does genetic epistemology have for philosophical epistemology, one philosopher has replied: . . . in a certain sense genetic epistemology presupposes a traditional epistemological position. If it has implications for epistemology, it is in that sense and for that reason, and not because of its status as a psychological theory. 19
The epistemology that Piaget is presupposing here, according to the
264
RICHARD
F. K I T C H E N E R
author, is a Kantian epistemology (as opposed to a rationalism or empiricism). Such a claim is echoed by other individuals, including psychologists, 2° although arguments for this are strangely lacking. Perhaps it just seems intuitively obvious that every theory must have epistemological presuppositions. But to Piaget, at least, this does not seem evident, since he denies his genetic epistemology does presuppose any particular type of epistemology. Two questions are crucial here: What does genetic epistemology include, and what does 'presuppose' mean? As Piaget characterizes genetic epistemology it is a m e t h o d - the study of knowledge as a function of its d e v e l o p m e n t - a n d this method does not "presuppose" or "prejudge" (according to Piaget) any epistemological position, such as realism, empiricism, a priorism, etc. 21 Piaget even insists it does not presuppose a genetic solution, for (on the contrary) nongenetic epistemotogies (a priorism, realism, phenomenology) are compatible with the genetic method, and in fact could be verified by such a method insofar as they are accounts of how knowledge increases. 2z That is, each of these solutions might be an adequate explanation of how knowledge develops and it would be an open question, to be settled by empirical research, as to which explanation was better. The only solution it would rule out, it seems, would be a scepticism which denied that knowledge increases at all. The genetic method does not presuppose a Kantian solution, therefore, even though Piaget believes this is the best explanation of the growth of knowledge. If this answer is to work, it is essential that the genetic method be characterized in this neutral way, as "the study of knowledge as a function of its real or psychological construction, ''23 or better yet, as "the theory of the mechanisms of the development of knowledge. ''24 It cannot be described in stage-theoretical terms, nor be characterized by means of models imported from biology (a tendency Piaget constantly has), nor even described in terms of 'constructions' if that means "construction by the epistemological subject." If Piaget can consistently maintain this relatively neutral way of characterizing genetic epistemology, then it seems to me to be correct to say it is compatible with several different epistemological solutions. The preceding point has, in a way, already answered the second point concerning 'presuppose.' For if 'presuppose' is meant in the strong sense of "requiring as an antecedent logically necessary con-
GENETIC
EPISTEMOLOGY
265
dition" then Kantianism is not being presupposed by Piaget: the genetic method (if what I have said is correct) does not require a Kantian epistemology as a necessary condition. On the other hand, if 'presuppose' means something weaker, e.g., "to take for granted" or "assume", then, aside from the preceding remarks, it would be open to Piaget to answer that he was assuming a Kantian position.as an initial scientific (or quasi-scientific) theory, subject to subsequent empirical evaluation and modification. If subsequent empirical evidence confirmed or corroborated this initial theory, then the charge of presupposing it would, ! think, be undercut since (in the usual sense) he would no longer be, merely assuming it (without evidence) but advancing it as a well-confirmed scientific theory. In either case, the charge that Piaget "presupposes" a Kantian epistemology seems to me to be questionable.
(2) D~res genetic epistemology presuppose epistemic norms ? Genetic epistemology may presuppose (or entail) normative epistemology in a different way. Development is obviously teleological in nature since development is always development towards some goal (end, telos). In the case of genetic epistemology this goal is knowledge. Thus, since 'knowing' is not merely a descriptive term but entails normative evaluation and assessment, it looks as if Piaget is committed to making claims about the normative status of the development process. Marx Wartofsky puts it very nicely when he says: What is arrived at, both by the developing individual and by the species, in its cognitive growth is knowledge of the external world. Therefore, what is involved here is not simply the genesis of concepts taken as a descriptive, empirical study. Rather, the study is normative and teleological. For the claim is that the development of physical concepts is a progressive and adaptive o n e - t h a t it eventuates in knowledge; that the conceptual network and the theories thus evolved approximate more adequately to the truth. Thus, the study is not simply one which concerns the history of genesis of physical concepts, but their adequacy as wellY
Wartofsky believes it is possible, however, to separate the normative and the empirical issues. . . . apart from the normative question of truth, or of truth seeking, one may deal ~vith normal cognitive sequence as an "empirical" question. One need not, therefore, ascribe
266
RICHARD
F. K I T C H E N E R
greater approximation to truth in the later stages, or, to put it differently, the pattern of cognitive growth may not be, and need not be, interpreted as a pattern of epistemological growth. The latter would require a criterion of growth of knowledge such that one could determine a passage from ignorance to knowledge. In short, the latter would require a definition of truth and a criterion of truth, and the ordering of stages of growth would be determined by this criterion, rather than by an temporal or maturational sequences. 26
Thus one might ask how Piaget goes from a genetic psychology (which is an "empirical" study of conceptual development) to genetic epistemology (which is normative). If the latter presupposes some epistemological criterion, what is its source and status? If we say that our criterion is simply assumed, that it is the epistemological standard adopted by current science and adult common sense, then what of this standard itself? Is there any reason to believe it is a true standard, for example, or must we forever leave such a question open? Every developmental sequence is teleological in the weak sense that it has a goal (or telos) towards which it is a development. Development is thus normative in the sense that different developmental sequences can be graded in terms of their tendency to attain this goal (or not) or to approach it in certain kinds of ways (speed, probability, ease, naturalness). Thus, developmental sequences as means inherit normative evaluation in virtue of the telos or goal which is assumed. The telos itself, however, is also subject to normative evaluation and appraisal and here the normative dimension arises most clearly when one believes that telos to be good, valuable, worthy, etc., or true, valid, correct, etc. Here development is normative in the strong sense. Piaget often writes as if he had some distinction like this in mind. The telos or goal of epistemological development (or, as he calls it, "the system of reference") is simply assumed to be the normal adult (in the case of psychogenesis), or current science (in the case of the history of science). In talking, for example, about "the child's construction of reality," the 'reality' that is the system of reference is reality as postulated by contemporary science or common-sense. 27 The task of what Piaget calls restricted genetic epistemology is to investigate how such a conception of reality is attained, all the while leaving open the question of the "true" nature of reality. Thus, Piaget
GENETIC
EPISTEMOLOGY
267
can evaluate the course of development in terms of its tendency to reach these accepted epistemic norms without claiming that these norms themselves are good (adequate, etc.). In fact, Piaget often stresses the point that in studying the development of knowledge from lower to higher forms of knowledge, the standard by means of which we judge something as 'lower' or 'higher' knowledge is not to be decided by psychologists but by logicians or specialists in the respective field of inquiry. 28 Thus (on this interpretation), all development is normative in the weak (teleological) sense. But what of these epistemoiogical criteria themselves? Can Piaget have anything to say about them; for example, their epistemological adequacy or truth, or not? The safest and most acceptable response would be to reply that such "normative" issues cannot be handled by a (restricted) genetic epistemology, since it studies the development of knowledge "from a lower level of knowledge to a level judged to be higher. ''29 Certainly this is an interpretation Piaget would favor since he stresses so strongly the fact-value distinction, and claims genetic epistemology cannot answer normative questions. 3° I do not think, however, that matters are nearly as clear-cut as this, even though it certain would flatter philosophers to think so. Piaget claims, for example, that generalized genetic epistemology would study this "system of reference" itself (although Piaget's exact views about this are not very clear). 3' In fact, I think the issue is much cloudier than even Piaget admits and that he does, contrary to what he says, make normative claims in the strong sense.
Piaget's equilibration model According to Piaget, the central task of genetic epistemology (and genetic psychology) is to explain how a transition is made from a "lower" level of knowledge to a "higher" level of knowledge. This applied both to stages of scientific knowledge and to stages of individual knowledge. The fundamental concept that explains stage transition is Piaget's notion of equilibrium and equilibration. In the case of the child, for example, there are several intellectual stages (of knowledge) through which the child must pass (sensorimotor, pre-operational, concrete operational, etc.). Each stage constitutes a particular form of equili-
268
RICHARD
F. K I T C H E N E R
brium, and these stages can be characterized as producing increasingly greater intellectual equilibration. 3z For Piaget, systems are conceived to be complex structures containing possible operations and possessing goal or goal-tending states. Internal or external variables may assume a particular value that take or tend to take this system out of its goal-state, to which the system responds by a compensatory operation that counteracts the disturbance and returns the system to its goal (or goal-tending) state. These "disturbances" that produce dis-equilibrium can include not only internal stimuli (hunger) but also external stimuli (an object), and even such things as questions and intellectual problems. 33 These disequilibrating disturbances give rise to external and internal action (thinking) and action terminates when a need is satisfied, that is to say, when equilibrium is reestablished between the new factor that has provoked the need and the mental organization that existed prior to the introduction of this factor. Eating or sleeping, playing or reaching a goal, replying to a question or resolving a problem, imitating successfully, establishing an affective tie, or maintaining one's point of view are all satisfactions that, in the preceding examples, will put an end to the particular behavior aroused by the need. At any given moment, one can thus say, action is disequilibrated by the transformations that arise in the external or internal world, and each new behavior consists not only in re-establishing equilibrium but also moving toward a more stable equilibrium than that which preceded the disturbance) 4
It should be noted here that (what we might call) epistemic needs (for example, question asked of a person, or some kind of theoretical or practical puzzle) constitute one important kind of need. A question, for example, may produce disequilibrium if no answer or solution is forthcoming. What is important, however, is the fact that epistemic norms are being used here; it is not just that the individual will attempt to eliminate the question or problem in any way (such as running away from it or ignoring it) but will attempt to solve it and moreover solve it correctly. Here criteria of achievement and success are being employed (just as in Piaget's biological analogue 'adaptation' is a success concept). When a child, for example, cannot successfully answer a question, this is sometimes manifested in his speech as a contradiction (which is experienced as a state of disequilibrium), or as an inability to answer questions about the explanation offered. The degree of equilibrium or equilibration therefore seem to be its degree of success or adequacy. One stage, for example, is more equilibrated than another (earlier) stage if, for
GENETIC
EPISTEMOLOGY
269
example, it answers the questions (solves the problems) the earlier stage did and, in addition, successfully meets new difficulties. The fact that a stage is disequilibrated explains why it tends to give way to another stage in which the earlier structure is retained, but retained in a new (and more adequate) form. The tendency towards equilibrium is thus the basic principle of intellectual functioning. As Piaget has characterized equilibration and equilibrium, it seems laden with normative evaluations. In fact, the very definition of equilibration itself has epistemological norms built into it, since it is characterized in terms of compensatory or anticipatory reactions that either avert a disturbing intrusion, satisfy a need, answer a question, and so forth. If Piaget were content to employ equilibration in its strictly biological sense, then perhaps one could argue that teleonomic systems (such as thermostats or temperature regulation in the body) would not be normative in the strong sense, (that is, normative with reference to the goal), but only in the weak sense, (that is, in reference to the means). This would be true, for example, with temperature regulation, since the goal of a constant temperature of 70 ° in the living room has no normative dimension, although the means might (for example we could evaluate how well the thermostat functioned, as well as which compensatory reactions were better). Similarly with biological needs such as nourishment or oxygen; one would not be claiming these biological goals were normative in an epistemological sense (although one might argue they were normative in an evolutionary sense). Piaget may seem to be doing precisely this: assuming that the telos is the set of adult norms, and (in keeping with the aim of restrictive genetic epistemology) not claiming that these norms are themselves valid, true, adequate, etc. One could, for example, cite the following passage in support of this: Just as the body evolves toward a relatively stable level characterized by the completion of the growth process and by organ maturity, so mental life can be conceived as evolving toward a final form of equilibrium represented by the adult mind? s
But there is considerable other evidence that would argue against this "value-free" interpretation and would suggest that Piaget is making epistemological claims about the telos of development. For example, the concept of equilibrium itself is often described this way: In assimilating objects, action and thought must accomodate to these objects, they
270
R I C H A R D F. K I T C H E N E R
must adjust to external variation. The balancing of the processes of assimilation and accommodation may be called "adaptation." Such is the general form of psychological equilibrium, and the progressive organization of mental development appears to be simply an ever m o r e precise adaptation to reality. 36
Adaptation is adaptation to reality, and accommodation is accommodation to reality; they are not merely adaptation and accommodation to adult norms. When a question or theoretical problem produces disequilibrium, equilibrium is restored when the question is correctly or adequately answered or the problem actually or partially solved. But how well the question is answered and how well the problem is solved are at least partly dependent on the nature of objects and the physical structure of the world. Thus, Piaget's claim that development can be characterized as an increasing equilibration of cognitive structures entails the claim that later stages are better (independent of adult norms), since they are more equilibrated and thus more epistemologically adequate. Equilibrium and disequilibrium, in other words, are not merely subjective psychological states, nor can disequilibrium really be restored by an individual's false beliefs or psychological defense mechanisms. True enough, what counts as an adequate answer or solution is partly a function of the child's social world and cultural indoctrination, but Piaget insists with equal vigor that reality is not equivalent to "social consensus." Present adult and scientific norms are taken as a reference standard not arbitrarily, but because they are objectively better than earlier scientific norms. If this is correct, and if Piaget is claiming that development is describable as a real increase in knowledge (and not merely an increase in what adults believe knowledge to be), must he have a criterion of the growth of knowledge such that one could determine a passage from ignorance of knowledge? Must he have a definition and criterion of truth? Piaget's answer with regard to these questions is not explicitly to be found in his writing, but the type of answer available to him, and one I believe he implicitly adopts, is one that is very close to (if not identical with) that of Karl Popper, an answer that is sometimes called "fallibilism": one can never know for certain when or if one has attained absolute truth; but this is not necessary, for one can recognize progress towards truth, and thus one can know when there is an increase in our knowledge.
GENETIC
EPISTEMOLOGY
271
Popper and Piaget P o p p e r and Piaget have many similarities, an obvious one being that between Piaget's notion of equilibration and P o p p e r ' s notion of verisimilitude. For Popper, the development of science is a growth in knowledge, and m o r e o v e r a rational and objective affair. According to P o p p e r our knowledge increases when a better theory comes along and replaces an older one. What makes one theory better than another? Roughly put, a theory T2 is a better theory than (is rationally preferable to, gives us an increase in knowledge over) a theory TT if T2 explains what TT explains but in addition explains things T1 could not explain, T: explains new facts and leads to new, precise and novel predictions, and some of these predictions are verified. 37 In this case T2 would give us more knowledge than T1 gave us and m o r e o v e r Tz would be closer to the truth (has a higher degree of verisimilitude), even if Tz is later falsified and replaced by a radically different theory (T3). Thus we can tell when our knowledge has increased even if we do not (and. may never) know when we have attained "the Truth." Piaget's account, as we have just seen, is fundamentally the same with respect to what an increase in knowledge is; for him it is an increase in equilibrium, and equilibrium and verisimilitude have much in common. P o p p e r is a realist about the external world, and believes we approach the truth about these real objects as our scientific knowledge increases. Likewise, Piaget believes objects exist independently of us, but they are only known by means of epistemological categories actively constructed by the epistemological subject: To be sure, the object exists and the objective structures exist themselves before being discovered. But they are not discovered at the end of an operational inquiry (in Bridgman's sense) in the way in which Columbus discovered America during his voyage: they are only discovered through being constructed; in other words, we can gradually approach them, but have not the certainty of ever reaching them.38 Thus objects certainly exist for Piaget, but such objects, since they are only known through the subjects' actions and constructions, are really limits " e v e r tended towards but never finally achieved. ''39 Piaget's "realism" might thus be characterized as a constructive realism and it certainly has similarities to P o p p e r ' s version of realism. If this is the case, then Piaget does seem to evaluate adult norms of
272
R I C H A R D F. K I T C H E N E R
knowledge. They (like current scientific theories) are "the best guess" as to what reality is like and they can be said to be "closer to the truth" than earlier theories, as well as being more adequate, more equilibrated, etc. Piaget's theory, therefore, does seem to be normative both in the weak and strong senses. At this point, however, it may be asked whether the "Popperian" interpretation of Piaget doesn't turn Piaget into a straightforward normative epistemologist. How is any of Piaget's empirical work relevant to these issues? What does Piaget have to say over and above what Popper has to say? I want to hold off answering those questions for the moment and instead turn to a completely different line of argument designed to show that Piaget's theory has normative implications in a radical and highly controversial way, a way that one individual characterizes as being an attempt to commit the naturalistic fallacy in ethical matters and to get away with it. Such an argument can be applied, mutatis mutandis, to Piaget's epistemological theory, the result of which would be an argument that would show how normative implications follow from Piaget's theory.
Piaget and Kohlberg The theory of Lawrence Kohlberg is a cognitive-developmental stage theory of moral development, and Kohlberg considers it to be a close off-spring of Piaget's theory. Kohlberg, unlike Piaget however, believes his theory not only has implications for moral philosophy, but believes that one can go from "is" to "ought" in ethical matters: ° What this "fallacy" is and whether Kohlberg actually commits it or not remains opaque. In fact, what Kohlberg's argument is remains largely unclear. Insofar as I understand it, it seems to be the following. The fundamental premise or assumption in Kohlberg's argument comes directly from Piaget, who claims that The fundamental hypothesis of genetic epistemology is that there is a parallelism between the progress made in the logical and rational organization of knowledge and the corresponding formative psychological processfl
Let us call this the principle of isomorphism (parallelism, correspondence) between the logical and the psychological (or the logical and the temporal). There is an isomorphism, for example, between a
GENETIC
EPISTEMOLOGY
273
causal sequence of brain states (or organic causality in general) and a set of logical implications entertained in consciousness. 42 Likewise, equilibrium as a logical concept-characterized, for example, by reversible mathematical operations-is structurally isomorphic to equilibrium as an empirical biological concept (involving, say, parts and whole in dynamic equilibrium). The temporal development of the individual corresponds in some sense to a logical sequence, and so on. This principle of isomorphism is also the key assumption at the basis of Lawrence Kohiberg's attempt to "commit the naturalistic fallacy," for Kohlberg's fundamental assumption is the following: •.. an ultimately adequate psychological theory as to why a child does move from stage to stage, and an ultimately adequate philosophical explanation as to why a higher stage is more adequate than a lower stage are one and the same theory extended in different directions.43
Thus the psychologist's explanation of why all individuals universally move from one stage of morality to the next will be parallel to or isomorphic with the philosopher's justification of this higher stage of morality a s being more adequate. Such an explanation cannot be value-neutral, of course. To explain why there is a universal sequence of stages of moral reasoning, Kohlberg believes, is to explain why everyone actually moves from stage to stage. But to explain this universal temporal sequence one must explain why a later stage is better than an earlier stage; this is the fundamental assumption of any strict Stage Theory based on the underlying concept of equilibrium. A later stage is better because it is more equilibrated, and equilibration includes, as two key concepts, the notions of differentiation (of cognitive and moral categories) and integration (into new structures). These cognitive-developmental concepts are thus supposed to explain why people actually develop the way they do, and the short of it is that people develop because later stages are better, where 'better' means more equilibrated. But here it is clear that we already have normative evaluation coming in. Kohlberg's next step is to suggest that these developmentally evaluative notions-differentiation, integration, equilibration- which explain why later stages are better are really the same as (or isomorphic to) certain philosophical criteria of adequacy: namely, prescriptivity, reversibility, and universalizability. Thus, we have argued for a parallelism between a theory of psychological development and a formalistic moral theory on the ground that the formal psychological develop-
274
RICHARD
F. K I T C H E N E R
mental criteria of differentiation and integration, of structural equilibrium, map into [sic] the formal moral criteria of prescriptiveness and universality. If the parallelism were correct in detail, then formalist philosophers could incorporate an equilibrium concept as part of their normative ethical theory, and vice versa. The ultimate result would be a theory of rational moral judgment like that now present in economics, in which the theory of how people ought to make economic decisions and the way they do make decisions are very closely linked. ~
Thus a "factual" explanation, an explanation involving differentiation and integration, of why people do develop is parallel to a philosophical justification, which would be an account of why people ought to develop. Explaining why people do move from stage to stage thus involves explaining why they ought to move from stage to stage: to explain why they actually move to a higher and better stage is to explain why they ought to move to a higher and better stage. Put in different terms, the increasing logical adequacy of the sequence S~-$2-$3 (partly) explains why people actually move in the sequence S~- $2- $3. I say 'partly' because this by itself would not be enough. To say that people develop from being illogical to being logical because logic is better, or that people move from a state in which they tolerate contradiction to one in which they insist on consistency because consistency is better than contradiction is not sufficient as an explanation. Not only must these later stages be more a d e q u a t e - a logical q u e s t i o n - w e also need a purely psychological component to explain why they actually move in that sequence, and here we must bring in some cognitive concept such as desire and belie[, people must want or desire to be more consistent (logical,). Otherwise put, it must be the case that disequilibrium (logical and moral inadequacy of a certain mental structure) is motivating, that it is the basic "motor" pushing people to develop. This is, of course, exactly what Piaget and Kohlberg believeY I have been suggesting that an explanation of moral development requires a logical theory of the relative adequacies of different stages and, in addition, a psychological component concerning the cognitive states of people (their beliefs, desires, etc.). Neither of these components by themselves would be adequate: the logical component would indicate what people ought to do, but without the psychological part one could not explain what people actually do, and thus, a fortiori, one could not explain why they move from stage to stage. Likewise, if the psychological and logical aspects were at variance in
GENETIC
EPISTEMOLOGY
275
that the stages were evaluated as being logically inadequate as a sequence but people really moved through this sequence in a certain order, this would be inadequate. Suppose, for example, individuals were in fact to move from an Aristotelian (Egoistic, Phenomenalistic) stage (SO to a Newtonian (Utilitarian, Realistic) stage (Sz), to a Neo-Aristotelian (Neo-Egoistic, Neo-Phenomenalistic) stage ($3). Suppose our philosophical theory judged Stage 2 to be better (higher) than Stage 1, and Stage 3 to be worse than Stage 2. We would surely need a psychological explanation of why individuals developed in this "illogical" sequence. What would this psychological explanation look like? It might involve, if Stage Theory is correct, notions of differentiation and integration, but to introduce these explanatory concepts would be tantamount to saying people believed Stage 3 was better than Stage 2. Since this, in turn, invites the question, "why?", reasons for the "belief" would enter and thus questions of logical adequacy would again emerge. We would be back again with our logical components, debating which logical account was better. Kohlberg's claim here seems to be that either our logical (philosophical) assessment was initially wrong (e.g., our moral philosophy was mistaken) or our psychological explanation was wrong (e.g., our Stage Theory explanation was wrong, our data were incorrectly gathered, etc.). This seems to lie at the basis of Kohlberg's claim that: The isomorphism of psychological and normative theory generates the claim that a psychologically more advanced stage of moral judgment is more morally adequate, by moral-philosophic criteria. The isomorphism assumptions is a two-way street. While moral philosophical criteria of adequacy of moral judgment help define a standard of psychological adequacy or advance, the study of psychological advance feeds back and clarifies these criteria. '~
Such an account works, it should be noted, only with a cognitive psychological theory, for only in such theories can we ask for the reasons for a belief and then proceed to asses the logical adequacy of such beliefs. In a Skinnerian learning theory, by contrast, the developmental sequence S~- $2-$3 would be explained by contingencies of reinforcement, and here a request for reasons for a particular stage would be beside the point. A later stage could not be said to be higher (and better) because it is more reinforcing any more than an explanation in terms of strengthened habits would be appropriate. Likewise in psychoanalytic theory, one cannot say the Genital Stage is higher (better) than the Anal stage, since criteria of
276
RICHARD
F. K I T C H E N E R
logical adequacy would surely be misplaced on these stages. In short, cognitive theory is essential, for cognitive theory is essentially and fundamentally tied to logic and the giving of reasons. According to Kohlberg, therefore, there is an i n t e r a c t i o n between the pyschological and the philosophical. As is obvious, however, the psychological is not the purely factual, devoid of all normative appraisal. The concepts of equilibrium (along with integration and differentiation) are laden with evaluation: they are more "stable," "self-sufficient," "adequate," "not contradictory," etc. If a naturalistic fallacy is being committed, it is not that of inferring a norm from a fact; but likewise if Kohlberg is going from the is to the ought he is not disembarking from an is that most philosophers would recognize, since it is not an 'is' denuded of all evaluative components and appraisal. But, then, perhaps nothing is that barren! Normative and philosophical implications are to be found in Kohlberg's theory and, by analogous reasoning, in Piaget's. That much should be clear by now. If so, then the point of this paper has been made. Doubts that linger, linger perhaps because of the (reasonable) fear that some fallacy has been committed. In conclusion I want to briefly touch on that question.
III. PSYCHOLOGISM
Whether Kohlberg really commits the naturalistic fallacy or not is a question that remains open partly because it may not be clear what the fallacy is. But surely Piaget (and Kohlberg) have committed s o m e version of the is-ought fallacy, the most likely candidate being psychologism, the fallacy of reducing logic to psychology, or inferring the truth of something normative (or logical) from something empirical, or permitting psychology to answer questions of logic, etc. This, at least, is the fallacy Piaget is most concerned with and the one he most explicitly denies committing. In conclusion I have four much too brief and sketchy comments about this issue. First, both Piaget and Kohlberg claim that values (and norms) cannot be reduced to facts, that the normative realm is autonomous, that psychology cannot settle questions of normative or formal validity, that normative judgments cannot be derived from psychological facts, etc. In that sense, therefore, they are not committing the fallacy,
GENETIC
EPISTEMOLOGY
277
and if that is what the fallacy is, then it seems to me the question is settled. Secondly, factual considerations can be relevant to the normative, and we have just given several examples of how they can be. It is perfectly legitimate and correct to say the factual is relevant to the normative-perhaps even essential, indispensable, or necessarywithout saying the normative is derived from the "purely factual," or that the factual realm can conclusively answer normative questions. We have given several actual cases of such relevance in the work of Piaget and Kohlberg, and we could give even more compelling examples, I think, from contemporary post-positivistic philosophy of science. Actual science and its historical development is relevant to the philosophy of science-this much seems no longer controversialthe only question is how relevant? But at the very least, however, actual science places constraints on which normative philosophical account is adequate. If this is true in the philosophy of science, it is a fortiori true of epistemology in general. 47 Thirdly, the fact-value issue itself may need re-examination in order to clarify its philosophical nature. Given that facts and values are not reducible to each o t h e r - s i n c e they are different conceptsdoes it follow that they are absolutely and categorically distinct? Given one cannot derive the truth of a normative proposition from a factual one, can one derive something (anything) of importance? It has yet to be shown that the latter is impossible. Finally, our conception of what facts are may be the source of our problem and may require reevaluation. The fallacy of psychologism (or the is-ought fallacy) seems to be based upon the belief that facts are something entirely devoid of any normative aspects, that they are "brute facts." The question is whether that is so, or whether there are different kinds of facts, some relatively "brute," others not so "brute" at all. If, as was once thought, facts and theories were things that were radically different and facts were supposed to be free of all theory, and if this is now legitimately questionable, then perhaps a similar argument can be suggested with regard to facts and values. Perhaps facts are "value-laden" as well as being "theory-laden"; perhaps the fact-value distinction (like the theory-observation distinction) is a relative or contextual matter (but not absolute), etc. Sharp, razor-edged distinctions such as the analytic-synthetic, the theory-observation distinction, the subject-object distinction, the dis-
278
RICHARD
F. K I T C H E N E R
tinction between philosophy and science have a way of being dulled with the passage of time and philosophical reflection. Perhaps we should have learnt this already from Hegel and the Pragmatists. If we have to learn it again with regard to epistemology, then Piaget's genetic epistemology is as good a way to learn it as any other.
Colorado State University NOTES * An earlier version of this paper was read at a philosophy colloquium at Colorado State University. I wish to thank all the participants for their helpful comments, even though several continue to be in fundamental disagreement. I specially wish to thank Don Crosby for his detailed comments. I have discussed some of these questions in my 'Piaget's genetic epistemology' (forthcoming), in which I examine what such an epistemology is supposed to be. In a sentence, we can characterize it as "the study of the growth of knowledge as a function of its development," or as Piaget puts it, "lYtude des m~chanismes de l'accroissement des connaissances" ["Programme et m6thodes de l'6pist6mologie g6n6tique." In W. E. Beth, W. Mays and J. Piaget (eds.), Epistdmologie g~n~tique et recherche psychologique. Etudes d'~pist~mologie gdn~tique, Vol. I. (Paris: P.U.F., 1957), p. 14]. 2 D. Hamlyn, 'Epistemology and conceptual development.' In T. Mischel (ed.), Cognitive Development and Epistemology (New York: Academic Press, 1971), p. 19. 3 I have discussed the case of Hume in considerable more detail in my 'Piaget's genetic epistemology' (forthcoming). 4 For example at B41, B42, B60 of the Critique of Pure Reason. 5 Bernard Rollin convinced me of this in private conversation. 6 See N. K. Smith, A Commentary to Kant's 'Critique of Pure Reason'. 2nd ed. (New York: Humanities Press, 1962), pp. 88-98. 7 In fact in one of Kant's more obscure passages (B167) he contrasts two views of explaining the agreement of experience with the pure concepts of the understanding empiricism and preformationism-and suggests his solution is better, a solution he describes as an epigenesis of pure reason. What Kant means by this is obscure, but A. C. Ewing offers the following explanation: "The point seems to be either that we are not given innate ideas at the start, but merely possess faculties which develop themselves under the influence of the environment and relatively late in life attain to full consciousness, of the environment and relatively late in life attain to full consciousness, or that the categories are a new contribution to nature by our mind, not there from the beginning in something existing before we experience it." (A Short Commentary to Kant's Critique of Pure Reason [Chicago: University of Chicago Press, 1938], p. 131). Epigenesis, of course, is exactly Piaget's preferred biological solution. See my 'Epigenesis: the role of biological models in developmental psychology,' Human Development 21 (1978), 141-160. 8 I. Kant, Critique of Pure Reason. Translated by N. K. Smith. (New York: St. Martin's Press, 1964), p. 12 [A xvii].
GENETIC
EPISTEMOLOGY
279
9 N. K. Smith, op. cit., p. 236. As N. K. Smith points out (ibid, p. 237) this can be called a
transcendental psychology. J0 A. J. Ayer, Language, Truth and Logic. 2nd ed. (New York: Dover, 1946), p. 84. N For example A. C. Ewing and B. Blansbard on the one hand and W. V. O. Quine and A. Pap on the other. ~2 j. Piaget, Genetic Epistemology (New York: W. W. Norton, 1970), p. 41. Cf. also J. Piaget, 'Nature et m6thodes de l'6pist6mologie.' In J. Piaget (ed.), Logique et connaissance scientifique (Paris: Gallimard, 1967), pp. 93-105. ~3 I have discussed these and related issues in my 'Piaget's theory of the a priori' (forthcoming). ~4 H. Siegel, 'Piaget's conception of epistemology,' Educational Theory 28 (1978), 16--22. 15 Ibid, p. 21. 16 These points are adequately made in Chisholm's Theory of Knowledge (Englewood Cliffs, N. J.: Prentice-Hall, 1966), pp. 82-83. ~7 A. Pap, Semantics and Necessary Truth (New Haven, Conn.: Yale University Press, 1958), p. 164. 18 In characterizing philosophical epistemology as normative I mean only that it deals with the validity of epistemic claims, with issues concerning what we are entitled to have a right to claim to know, with whether our beliefs are warranted, reasonable, justified, etc. Anglo-Saxon epistemology is largely concerned with evaluating such claims and such evaluation would normally involve an appeal to epistemological standards, or norms of validity (just as logic would appeal to standards of validity when evaluating logical claims). The most radical form of such a view is R. Chisbolm's Perceiving (Ithaca, N.Y.: Cornell University Press, 1957). For a more moderate view, see R. Brandt, 'Epistemology and ethics, parallels between.' Encyclopedia of Philosophy. Vol. 3 (New York: Macmillan Co., 1967), pp. 6-8. t9 D. Hamlyn, loc. cit. 20 B. Kaplan, 'Genetic psychology, genetic epistemology, and theory of knowledge.' In T. Miscbel (ed.), Cognitive Development and Epistemology (New York: Academic Press, 1971), p. 74. 21 Introduction a I'~pist~mologie g~n~tique (Paris: P.U.F., 1950), Vol. 1, pp. 13, 24. 22 Ibid, p. 25. 23 lbid, p. 13. 24 'Programme et m6thodes de l'6pist6mologie g6n6tiqne,' op. cir., p. 14. z5 M. Wartofsky, 'From praxis to logos: Genetic epistemology and physics,' T. Mischel (ed.), Cognitive Development and Epistemolcgy (New York: Academic Press, 1971), p. 132. 26 lbid, p. 135. 27 j. Piaget, The Child's Conception of Physical Causality (Totowa, N.J.: Littlefield, Adams & Co., 1969), pp, 237-238. 28 Genetic Epistemology, op. cit., p. 13. 29 IbM (my emphasis). 30 "Insofar as any attempt to solve a logical or matffematical problem by using results borrowed from psychology is called "psychologism", we likewise condemn psychologism without h e s i t a t i o n . . . " (W. E. Beth and J. Piaget, Mathematical Epistemology and Psychology, (Dordrecht: D. Reidel, 1966, p. 132).
280
R I C H A R D F. K I T C H E N E R
3~ See his Introduction a lYpist~mologie g~n~tique, op. cit., pp. 45-48. 32 j. Piaget, 'The mental development of the child', in Six Psychological Studies (New York: Random House, 1968), pp. 3,6. 33 Ibid, p. 7. 34 Ibid. 35 Ibid, p. 3° 36 Ibid, p. 8 (my emphasis). 37 K. Popper, 'Truth, rationality, and the growth of knowledge,' in Conjectures and Re[utations (New York: Harper, 1968), pp. 215-250). 38 Principles o[ Genetic Epistemology (London: Routledge & Kegan Paul, 1970), p. 84. 39 Ibid, pp. 17, 82. 40 L. Kohlberg, 'From is to ought: How to commit the naturalistic fallacy and get away with it in the study of moral development,' in T. Mischel (ed.), Cognitive Development and Epistemology (New York: Academic Press, 1971), pp. 151-236. 41 Genetic Epistemology, p. 13. 42 The Child and Reality (New York: Viking Press, 1972), pp. 170-172. See also Pi/tget's 'Explanation in psychology and psychophysiological parallelism,' in P. Fraisse and J. Piaget (eds.), Experimental Psychology: Its Scope and Method, Vol. I. (New York: Basic Books, 1968). 43 Op. cit., p. 154. 44 Kohlberg, op. cit., pp. 224-225. 45 Such an interpretation is at variance with the views of T. Mischel in his 'Piaget: Cognitive conflict and the motivation of thought,' in T. Mischel (ed.), Cognitive Development and Epistemology (New York: Academic Press, 1971). L. Kohlberg, 'Moral development and moral philosophy,' Journal of Philosophy 70 (1973), p. 633. 47 I have discussed these issues in more detail in my 'Genetic epistemology and historicist philosophy of science' (forthcoming).
