When Wish Replaces Thought, page 28
The psychoanalyst would, for example, make predictions concerning the fantasy life of children who will later become philosophers; in practice this will, of course, take the form of prediction of the types of content of childhood fantasies that will be reported by adult philosophers.8 Consider the following fantasies and feelings of childhood: A child wonders whether his choosing not to listen to a particular game played by his favorite baseball team was of sufficient causal importance to account for his team's losing. The same child feels discomfort at division problems that leave an inelegant-feeling remainder. The same child takes an inordinate amount of time putting away finger paints because he, unlike the teacher who wishes merely that the fingerpaints be put on the shelf, feels that the paints should be placed in (an intuitively perceived) order (red, orange, yel- low--or yellow, orange, red-but not red, yellow, orange). This child wonders whether one could freeze a piece of ice to a temperature sufficiently low that it would, when dropped in the ocean, freeze the water around it, which would freeze the water around the water, until the oceans were frozen. The same child feels intuitively that no two things could be exactly alike because different "causes" generate different effects (and all causes differ at least with respect to time/space).
Now these are hardly exceptional fantasies, and they are found to a greater or lesser extent in most children. The Freudian would, however, make the prediction that these sorts of fantasies will be far more frequently reported by individuals who become philosophers, mathematicians, and theoretical scientists, as well as by members of other groups who lack the interest (accountants) or intelligence (stockroom organizers) than by, say, actors. The Freudian would suggest that the reports of childhood fantasies of historians and novelists would differ markedly from those most frequently reported by the philosophers.
Even if this were all true it would not, in and of itself, do much to corroborate Freudian theory. All it would indicate would be the rather unsurprising fact that the adult mind reflects the same propensities found in the childhood mind; the same can be said of skin color. The fact that Freudian theory is concordant with childhood-adult similarity, like the fact that Freudian theory is concordant with a physicalist's view of the world, and with physiological explanations, merely allows it to stay in the game.
What is crucial is that the Freudian claims to predict not merely that there will be a consistency of childhood and adult mind and behavior, but that there will be correlations between childhood familial experiences/ situations and type of mind; the Freudian claims not merely that such correlations will be surprising rather than obvious and trivial, but also that they will make sense only in terms of Freudian explanation and the factors that it sees as crucial.
The Freudian claims, in other words, that the group of individuals who become philosophers will report weaning experiences, toilet-training experiences, attitudes about the sufficiency of the mother's love and the father's strength, and the like that are specifiably and significantly different-quantitatively and/or qualitatively-from those reported by members of other groups. Whether the early experiences of the members of the two groups differed in reality or only in perceptions of the members is not ascertainable from report alone, but this is not of great importance (since it is the individual's view of the world that is seen by the Freudian as crucial).9 In any case, such predictions are not merely falsifiable, but are translatable into correlational terms that could satisfy even the operationalist. If the Freudian specifies the reported childhood experiences and specifies the adult cognitive propensities, then the Freudian cannot be accused of making predictions that owe their correctness to biased selection, the charge that is so convincingly leveled at the clinical evidence that constitutes almost the entirety of the contemporary defense of Freudian theory.
In all of his explanations and predictions the Freudian, like anyone else mak ing a prediction, invokes a ceteris paribus clause that enables him to ignore factors that are irrelevant to his prediction. He does this with physiological predispositions, certainly when they are hereditary and often even if they result from some fetal environmental experience. He does the same thing with social factors. Thus, the fact that a particular low-income group may have a much higher murder rate than does a high-income group does not cast doubt on any Freudian explanation of a murder committed by a member of the low-income group. The Freudian does not claim sufficiency for his explanation, but only necessity (or at least increased probability). To put this another way: The Freudian does not claim to explain why the members of the low-income group commit more murders (which demands a sociological explanation), but why those members of the lowincome group who do commit murders do so when the vast majority of other members of the same social group do not; this can not be explained in terms of social group differences. (It is worth remembering that the dichotomizing of "social" and "psychological" is always a heuristic oversimplification. Social factors, for example a welfare system that drives the father out of the house, can play a great role in the development of personality.)
COMMENTS ON SOME SPECIFIC METHODOLOGICAL CRITICISMS OF FREUDIAN THEORY
Much of the philosophical criticism of Freudian theory argues that such theory, while perhaps not, strictly speaking, unfalsifiable, is so "open textured," so replete with escapist subclauses and ubiquitous theoretical constructs, that it is in practice untestable. For example, one can hardly deny that Freudians have on many occasions seemed to explain "opposite" effects in terms of identical causes and identical effects in terms of "opposite" causes.
However, it is well to remember that, on the level of gross observation, "same" causes do seem to generate "opposite" effects, and vice versa, and that science is an attempt to explain observation, gross observation at first. We do observe that the child of one overly aggressive father is overly aggressive, while the child of another overly aggressive father is passive ("same cause, opposite effect"); we do observe that one overly aggressive child has an overly aggressive father while another overly aggressive child has a passive father ("opposite cause, same effect"). Clearly it is to be hoped, and is usually the case, that a Freudian explanation would specify causes and effects far more precisely than this, but I would suggest that even an explanation this gross is not, as critics often assert, meaningless by virtue of unfalsifiability. This criticism would be valid if "overly aggressive" and "passive" subsumed all possibilities, if every father was seen as either "overly aggressive" or "passive." If this were the case, assertions of a relationship between paternal aggression/ passivity and child aggression/ passivity would be unfalsifiable and we would have no reason to believe that paternal aggression/ passivity had anything to do with child aggression/passivity. But clearly this is not the case. Clearly, the Freudian conceives of most fathers and most children as being nei ther "overly aggressive" and "passive." The meanings of "overly aggressive" and "passive" are in theory operationally specifiable (as scoring very high or very low on an objective test measuring "aggressiveness" or as meeting behavioral requirements such as hitting and kicking, on the one hand, or extreme uninvolvement, on the other). These terms do not, therefore, subsume the behaviors of all individuals, but only the behaviors of individuals at both extremes of a continuum.
This can be seen more clearly if we consider the concept of "reaction formation," a concept that has been criticized in similar terms. Again we might point out that even gross observation does indicate that (A) people do "protest too much" and defend by denial, suggest that (B) this is an empirical observation demanding explanation, acknowledge that (C) if the concept of "reaction formation" "explained everything," it would explain nothing, and emphasize that (D) the concept does make risky predictions that are falsifiable and testable.
Thus, even if one did explain the alcoholism of one individual in terms of the same causes as the rigid temperance of another (with "reaction formation" included in the latter explanation), his explanation would be of a gross sort, but it would be better than nothing (which is perhaps the strongest argument Freudian theory has going for it). The prediction is made that the configuration of causal factors specified by the Freudian is highly correlated with both alcoholism and rigid temperance, but not with the moderate use of alcohol that is found in, say, 80 percent of the American population. Much Freudian explanation is of this type. The psychoanalyst would, knowing only that an individual's toilet training was unusually difficult, predict that the individual would have difficulty in the area of cheapness/ generosity and would be unusually parsimonious or unusually extravagant. Likewise, the psychoanalyst would, knowing only that an individual had an unusually difficult time with weaning and toilet-training, predict that the individual would have difficulty in the area of dominance/ submission and would see far more of reality in these terms than would one who did not have such difficulty in infancy. The psychoanalyst would, in other words, predict the area of life and type of concern that will dominate the individual's life, rather than the direction of concern or behavior.
To be sure, such explanations are incomplete and remain so until the direction of concern or behavior is specified. To be sure, the explanations are incomplete and remain so until the distinctions necessary for prediction of direction (cheapness or generosity) as well as area (cheapness/generosity) can be made. But even without this capacity the explanation does make risky predictions of unexpected correlations and can justifiably be termed scientifically meaningful.
It is worthwhile discussing two additional Freudian terms that are often incorrectly asserted to be unfalsifiable and guilty of other methodological sins: "overdetermination" and "latent homosexuality."
The concept of "overdetermination" is, like "reaction formation," valid only at a certain, and in practice, inevitable level of ignorance. Since every nonrandom event can (in theory) be shown to be different from every other event (if we accept our intuition that "every difference makes a difference'), and since, therefore, every such event has a different causation, the Freudian term "over determined" reflects our (inevitable) ignorance of the complete nature of the reality we attempt to explain. Sufficiently precise description of an action (and the thoughts in the actor's mind) would (in theory) enable us to present a cause that "entailed" the action, and only the action, we wished to explain. Thus, when we say that a painter's painting a nude is overdetermined in that it serves a number of needs, we mean that any one of the needs is sufficient to explain the artist's painting a nude. But this explanation is sufficient only because we describe in only the grossest of terms (painting a nude) the effect we wish to explain. If we were to describe the painter's thoughts and the minute details of the painting with sufficient explicitness, we would see that one, and only one, causal configuration of needs possessed the necessity and sufficiency to explain the effect. This point is so absurdly theoretical, given the reality that faces the psychoanalyst, that the psychoanalyst is well within his rights to ignore it altogether.
Similarly, the criticism that the concept of "latent homosexuality" is without value because it is a concept that is untestable is clearly a criticism that is without merit. Let us imagine a test administered to sixteen-year-old boys with no homosexual experience. Let us say that those boys who score above eighty turn out, with great statistical significance, to become adults whose sexual preference is solely homosexual. Let us say that those who score between sixty and eighty do not tend to participate in homosexual encounters but do remain unmarried and do tend to exhibit the defensive behavior that the Freudian associates with suppressed homosexual tendencies. It seems to me that this is the sort of thing the Freudian has in mind when he speaks of "latent homosexuality" and this is perfectly valid even from a purely operationalist point of view. Such a test may or may not exist, but the mere fact that we can imagine it is sufficient to demonstrate the methodological validity of the concept of "latent homosexuality."
Most of the philosophical criticism of Freudian theory is of the type we have discussed. When such criticism is meant as criticism of what Freudians do in practice, it is unexceptionable. The objections I raise in this essay refer only to the suggestion that the criticisms expose a logical or theoretical flaw in Freudian theory. Such criticisms are common. Arthur Pap found fault with the concept of unconscious feeling, Sidney Hook objects that the Freudian can not describe a child who does not have an Oedipus Complex, and Karl Popper implies that, because the Freudian will say that the man who drowns a child suffers from a repression of some element of his Oedipus complex while the man who saves the child acts out of sublimation, Freudian theory is capable of explaining everything and is therefore unfalsifiable.
Pap argues that the common concept "unconscious feeling" (as in "he unconsciously dislikes Mr. Smith, though he thinks he likes him") is meaningless because terms like "feeling" and "dislike" are by definition conscious states of awareness. He suggests that we are better off using the term "unconscious disposition"; thus, when we observe that Mr. Jones feels that he likes Mr. Smith, but acts as if he dislikes Mr. Smith, we are observing that Mr. Jones has a disposition, of which he is unaware, to act toward Mr. Smith as would one who consciously disliked (felt a dislike for) Mr. Smith. I would suggest that if one conceives of the unconscious as ultimately a physiological entity that, like the central nervous system, subsumes a great number of physiological entities, and if one imagines that Mr. Jones's (unconscious) disposition to act as if he dislikes Mr. Smith represents an environmentally generated physiological state that is identical to that found in one who consciously dislikes Mr. Smith (except with reference to the physiological factors relevant to conscious awareness), then the term "unconscious dislike" does have meaning.
I think that it is possible to describe a child who lacks an Oedipus Complex, and that this is what the Freudians have done in response to Sidney Hook's criticism. I would, however, suggest that Hook is quite correct in suggesting that the virtual nonexistence of individuals without Oedipus Complexes would render the concept virtually worthless as an explanation of anything were not described more specifically. (It would be like an explanation of nightmares that saw them as caused by life; one can describe a person who lacks life, and it would be true that a life is a necessary condition for a nightmare, but the explanation that said no more than that nightmares are caused by life would be somewhat unsatisfactory.) I would suggest that the Freudians would be better off by taking a different approach; they would acknowledge that virtually all people have Oedipus Complexes, but that "Oedipus Complex" is merely a convenient term that subsumes a number of factors (desire for mother, fear of father, anger towards father, and the like) whose relative strengths will affect the individual's adult thought and behavior. We do not have units for measuring these factors, but we can certainly speak of a relatively great fear of the father and a relatively weak fear of the father, we can specify the adult behavioral realities we are attempting to explain, and we can predict correlations between various combinations of these factors and adult thought and action. No one criticizes as meaningless the sociologist's statement that stratification "causes" alienation simply because all societies are stratified, because stratification cannot be counted, or because alienation is difficult to define precisely. One understands that the sociologist, in saying that stratification "causes" alienation, is simply oversimplifying the results of investigations that demonstrate that the strength, degree, and type of differentiation subsumed under the term "stratification" are correlated with such measures of alienation as suicide.
Similarly, Popper would be right to be dissatisfied with the "explanation" that did nothing more than say that one man drowned the child as a result of repression while another saved the child as a result of sublimation. But surely the Freudian would say something more; he would claim that he could, at least, differentiate the childhood factors tending to lead one person to develop a style in which repression is the primary defense from those tending to lead another person to develop a style in which sublimation is the primary defense.
So many psychoanalytic theorists have responded to the demand for falsifiability and testability by arguing that psychoanalytic theory requires "postdiction" rather than prediction, and so many psychoanalytic theorists have defined "postdiction" in so many different ways that it is necessary to address the point here.
(1) There is no problem when by "postdiction" the theorist means a predic tion of what a patient will report about his past. There is no logical distinction between such postdiction and prediction. This invocation of postdiction has the advantage of stressing that the psychoanalyst does not (necessarily) claim to be able to predict (for example) which four-year-olds will become adult homosexuals. (The theorist might argue that while the seeds of homosexuality have been sewn by this age, practical problems of observation preclude even statistically significant predictions of which four-year-olds will become adult homosexuals.) This theorist argues that the evidence for the correctness of his theory of homosexuality is not a test of the prediction of which four-year-olds will become homosexual, but the postdiction of what adult homosexuals will report about their childhoods. If the theorist is capable of predicting that his homosexual patients will more often report a particular configuration of childhood factors and a particular childhood familial situation (which the theorist specifies) than will his heterosexual patients, then he gives us a good reason to accept the correctness of his theory of homosexuality. (I ignore practical problems of self-fulfilling prophecy.)
