Four books stand out as paramount contributions to the emergence of this still-contested but; in my view, warranted judgment:

1. Priority in time belongs to the late Henri Ellenberger's The Discovery of the Unconscious (1970), whose long and learned chapter on Freud demolished the myth, carefully nurtured by Freud himself and his Boswell, Ernest Jones, of the master's utter originality, his facing up to disturbing truths unearthed in his clinical practice, and his solitary defiance of his contemporaries' prudish hypocrisy. By displaying Freud's all-too-human opportunism and disingenuousness and by bringing him down from the clouds into 19th-century intellectual history, Ellenberger tacitly invited other scholars to inquire whether the vast cultural success of psychoanalysis rested on any actual discoveries.1

2. One such scholar was Frank Sulloway, whose monumental Freud, Biologist of the Mind (1979) went farther than its author himself initially intended, or even realized, toward dismantling Freud's claims.2 Sulloway's Freud is an ingenious plagiarist, a dogged and ruthless self-promoter, and a habitual devotee of crackpot ideas and premature conclusions. After Sulloway, it became harder to avoid perceiving that Freud's conveniently unexaminable case material always fit perfectly with whatever notion he had most recently pressed into service from unacknowledged and often questionable sources.

3. Adolf Gr�nbaum's formidable The Foundations of Psychoanalysis (1984) demonstrated that "clinical evidence," the purported engine of reliable psychoanalytic knowledge, could not underwrite that knowledge even in principle. The problem of contamination through therapist suggestion, Gr�nbaum showed, is pervasive and intractable, and even uncontaminated clinical data, if any such could be found, would necessarily lack the causal import that Freud and others have ascribed to them.3

4. The fourth classic of Freud revisionism is the book now under review. To call it a classic, however, is more a prediction than a statement of settled fact. Although Malcolm Macmillan's Freud Evaluated has been in print since 1991, and although it has been highly praised by Eagle (1993) and Spence (in press), among others, its influence thus far has been slight. I indicate later why this is the case, why I am confident of a very different future for this book, and, most important, why Freud Evaluated must be painstakingly studied by anyone who aspires to make pronouncements about the good, the bad, and the ugly in Freudian thought.

First, precisely because Macmillan is still largely unknown to American readers, a word of introduction is in order.

EVALUATION IN HISTORICAL CONTEXT

By itself, the failure provides no guide as to where the fault lies. Perhaps the original facts were inaccurately described or the original theoretical terms inadequately formulated. Would it not be sensible to see how those terms or statements were arrived at? Was there a worthwhile theory to begin with? . . . In brief, historically based evaluations help us to establish what has to be explained and whether any explanatory effort is justified. We are also placed on more certain ground in deciding which kinds of evidence should count as confirmatory and which as disconfirmatory. (p. 4)

COMPLEXITIES REDOUBLED

ILLUSIONS OF DETERMINISM

If the supporters of the suggestion theory are right, all the observations made at the Salpetriere are worthless; indeed, they become errors in observation. The hypnosis of hysterical patients would have no characteristics of its own; but every physician would be free to produce any symptomatology that he liked in the patients he hypnotised. We should not learn from the study of major hypnotism what alterations in excitability succeed one another in the nervous system of hysterical patients in response to certain kinds of interventions: we should merely learn what intentions Charcot suggested (in a manner of which he himself was unconscious)to the subjects of his experiments�a thing entirely irrelevant to our understanding alike of hypnosis and of hysteria. (Freud, 1888/1966, pp. 77-78; italics added)

If Freud Evaluated poses a cautionary moral, it is that Freud's fatal error lay exactly here. For, in Macmillan's words, "Freud was to be as wrong about hysteria as he had been about hypnosis" (p. 72)-- and in just the same manner. Although he made token efforts to reason his way around the obstacle posed by suggestion, he refused to take the phenomenon seriously:

When Freud came to treat his own patients, he never accepted that influences transmitted unconsciously from him to them had important effects upon what they claimed to recall about the origins of their symptoms. His view was that the important determinants of remembering were internal, part of the very fabric of the patient's thoughts, and as impervious to outside influence as the processes determining the phenomena of hypnosis and hysteria at the Salp�tri�re were supposed to have been. (p. 73)

The price of this mistake was a record of tragicomic blundering that Macmillan traces from Freud's cases of "Elisabeth von M" and "Dora" through his most arcane feats of system building and those of his successors, who themselves have evidenced a nearly total indifference to suggestion.7

MARKS OF PSEUDOSCIENCE

1. Hypothetical entities or processes should be characterized; that is, they ought to possess attributed properties that lend themselves to confirmation outside their immediate role in the theory at issue. If they lack this quality, "their referents are the very relations they are supposed to explain" (p. 193); they are only placeholders for mechanisms that may not exist at all.10 This is just what we regularly find in the case of psychoanalytic postulates. A term like repression, Macmillan notes, points to no independently known reality but merely gives a name to the questionable survival of traumatic memory traces in an unconscious that itself remains uncharacterized. Moreover, incompatible burdens are placed upon the term, indicating that the theory behind it is fatally muddled." When repression is then invoked as an explanatory factor in new contexts, true believers may feel that fresh territory is being conquered, but the scope of Freud's circularity is simply being widened. The same flaw of empty conceptualization appears in virtually every feature of his system, from the preconscious through the ego, introjection, the death instinct, and so forth.

2. A theory should not create its own facts. Psychoanalysis, however, does so at every turn. For example, repression is invoked to account for the delayed effect of childhood trauma in producing adult psychoneuroses, but the only reason for believing that such an effect occurs is a prior belief in repression. A dream is regarded as a disguised representation of its latent content, the dream thoughts, but such thoughts can be detected only by Freudian dream interpretation. So, too, castration threats, real or fantasized, supposedly trigger the onset of the male latency period, but the latency period is itself a pure artifact of the theory. Or again, Freud invoked penis envy to explain female submissiveness, masochism, and incapacity for cultural strivings, but in this instance the theory and the "facts" alike derived from cultural prejudice.

3. A theory should have testable consequences; "only if the facts [to be independently verified can be deduced from the fundamental statements of the theory can we say that they are explained by it" (p. 168). Notoriously, however, Freudian tenets are scarcely challenged, much less refuted, by unexpected outcomes. The vagueness of the theory is such that it can withstand almost any number of surprises and be endlessly revised according to the theorist's whim, without reference to data. Indeed, as Macmillan emphasizes, Freud drew on the same pool of evidence in offering three incompatible etiologies for homosexuality (pp. 352-353), and he did the same in proposing three incompatible paths for overcoming narcissism (pp. 35~359). Throughout his whole career of lawgiving, the linkage between evidence and theory was established by rhetorical guile and nothing more.

4. A hypothesis should be treated as such; that is, its adequacy ought to be methodically tested. Instead, Macmillan shows, Freud habitually offered postulates that he labeled as hypotheses but treated as firm expectations or even as certainties. Understandably, premature closure about one issue left him vulnerable to the same mistake with the next one. For example, all the while that he was pretending to be alarmed at his reluctant clinical discovery of sexual factors in hysteria, he was importing the conclusions he had already erroneously reached about the sexual roots of the (nonexistent) actual neuroses.

5. Finally--though this list could be considerably extended--heed must be paid to the difference between necessary and sufficient causes. An assertion that factor x causes effect y in neurotic Group A is vacuous if one merely establishes the presence of factor x in typical members of that group. Even on the most optimistic interpretation (that x is necessary to produce neurosis), x cannot be regarded as a sufficient cause unless, at a minimum, it is shown to be absent from nonneurotic Group B. Never once in his psychoanalytic career, however, did Freud conduct such a demonstration or publicly indicate that it was called for.12 On the contrary, he consistently maintained that all the reassurance of correctness he required was the stream of confirmations that flowed from clinical experience--in other words, from Group A alone. At his most scrupulous, he was content to find a few cases in which the positive correlation he was seeking appeared, however momentarily, to obtain. A palm reader or faith healer could have done as well.

NOTES

1. Concurrently with the appearance of Ellenberger's masterwork, Frank Cioffi began publishing trenchant and cogent articles and review essays (Cioffi 1969, 1970, 1972, 1974, 1983, 1985) that have had an enormous impact on Freud revisionism. As we will see, Macmillan's achievement is very much within Cioffi's tradition--that of combining epistemic interrogation with close study of Freud's (and his movement's) conduct in the face of challenge. Nearly all post-1970 arguments to the effect that psychoanalysis is a pseudoscience--as opposed to a mere theory that has not yet found corroboration--belong to "the school of Cioffi" His pungent essays and reviews on Freudian topics eminently merit collection within a book.

2. Sulloway's revised, far harsher, view of Freud is expressed in Sulloway (1991; 1992, preface).

3. It is true that Gr�nbaum, as a matter of logical possibility, left the door ajar for some future extraclinical corroboration of unspecified Freudian ideas, and he continues to do so (Gr�nbaum, 1993). But a dispassionate reader would conclude, as Gr�nbaum himself frankly indicated in a recent article (Gr�nbaum, 1994), that no such turnabout is likely. The other revisionists I have cited would probably go farther and maintain, as I do, that limited cash and patience can be put to better uses than the further testing of propositions that have remained unconfirmed after a century's effort and that arose (as we will see) from demonstrably faulty inferences.

4. This new edition is especially welcome because until now the scope of Macmillan's achievement has been obscured by frustrations and vicissitudes of the publishing trade. Twelve years in the writing, Freud Evaluated languished for another 5 years while American university presses were rejecting it as too long, too specialized, or too critical of the beloved genius whose ideas it addressed with so little obeisance. Finally. Macmillan had to settle for a publishing house, North-Holland, that printed only 1,200 copies, set the price at a forbidding $140, and made only the most negligible gestures toward advertising and distribution. Furthermore. North-Holland provided no copy-editing or proofreading and no galleys for Macmillan to check, with the result that the work abounds in vexing typographical errors and infelicities that could lead a reader to believe, mistakenly, that the author's intellectual labor was equally careless. It is not surprising that to date only a few hundred copies of the book have been sold or that many scholars who consider themselves Freud experts remain unaware of its very existence.

5. This point is now familiar from any number of recent discussions, but it was broached by Macmillan nearly two decades ago (Macmillan, 1977). Our current "recovered memory movement," with its perverse homage to the mistakes of Freud's seduction theory, could never have gotten launched if the world had attended to Macmillan's astute article instead of to later and more ideologically driven accounts.

6. In passing, we may note that Macmillan's expos� thus casts an eerie light on Freud's continuing "interdisciplinary" vogue within our universities. It seems that Freud is treasured for the very qualify--unbounded hermeneutic license--that signaled, and compounded, the failure of his medical and scientific pretensions as he first conceived them. We should not wonder, then, that academic apologetics for psychoanalysis have taken an increasingly postmodern turn, "problemitizing" the very idea of factual truth and scientific rationality. For a hostile review of Freud Evaluated that epitomizes this trend, see Leys 1992).

7. It would be incorrect, of course, to maintain that psychoanalysis has paid no attention to unconscious influence; "transference" looms large within the theory. As that concept is typically employed, however, it serves to put the phenomenon of suggestion even farther out of sight than it would otherwise be. Emotionally charged reactions to the therapist are treated as stemming not from the therapist's (possibly misguided or offensive) ministrations, but from projections on the patient's part. Thus, Freud's peculiarly narrow idea of psychic determinism reigns unchallenged--and continues to do so when countertransference, or the contribution of the therapist's own infantile fixations, comes to the theoretical forefront. When each member of the therapeutic dyed is regarded as a prisoner of the distant past, attention to the effect of contemporary expectations and demands becomes impossible.

8. Unfortunately, most of these assumptions have been making a sinister comeback as a number of psychoanalysts have become converts to the recovered memory movement. This demoralizing turn of events is most clearly manifested in a special issue of Psychoanalytic Dialogues (in press); it is further discussed in Crews et al. (1995, pp. 1~29).

9. Although Macmillan is not primarily concerned with Freud's scientific ethics, he occasionally shows how Freud distorted other researchers' work (e.g., pp. 308-310) and failed to distinguish even minimally between his own associations and those of his patients (e.g., pp. 256-261. For further questioning of Freud's integrity and competence, see Esterson (1993) and Crews et al. (1995).

10. Macmillan's views in this area constitute an extension of his compatriot W.M. O'Neil's exposition of the difference between real and empty assertion in psychological theorizing (O'Neil, 1953). It should be emphasized that O'Neil's criteria are neither eccentric nor positivistic; they merely articulate demands that are tacitly exercised in the daily practice of science.

11. As Macmillan remarks, "For an affect to be converted and for its idea to become unconscious the separation of the two has to be complete or near complete. However, for abreaction to take place, the idea has to be recovered with its affect still attached. Symptom formation thus requires repression to separate the idea from its feeling but symptom removal requires they remain attached" (p. 161).

12. The closest Freud came to such recognition was in an early letter to Wilhelm Fliess announcing that neurasthenia in men is caused by masturbation--a practice, he assured Fliess, completely lacking among "the circle of one's acquaintances" who have not contracted the neurosis despite having been "seduced by women at an early age" (Freud, 1985, p. 40). One strains to imagine the interviews that could have assured Freud of his correctness on this point.

REFERENCES

Cioffi, F. ( 1969). Wittgenstein's Freud. In P. Winch (Ed.), Studies in the

Cioffi, F. ( 1970). Freud and the idea of a pseudo-science. In B. Borger & F. Cioffi