Freud told us that the theory of repression is the cornerstone of the entire psychoanalytic theory of unconscious motivations. And he claimed that his clinical evidence furnishes compelling support for this cornerstone. Thus, I was able to scrutinize the logical foundations of the psychoanalytic edifice by examining Freud's clinical arguments for the repression-aetiology of the psychoneuroses, and for the cardinal causal role of repressed ideation in committing parapraxes (”slips”) and in dreaming. The upshot of this scrutiny was that the reasoning by which he thought to justify the very foundation of his theory was grievously flawed.
Plainly, this conclusion leaves quite open whether some other, genuinely probative evidence will turn out to lend significant support at least to the repression-aetiology of psychoneuroses, which is the major pillar of the Freudian structure.
Though I have given a critique of the basic pillars of psychoanalysis, it might be asked: Why its anachronistic focus on Freud's reasoning to the exclusion of the modifications and elaborations by those post-Freudians whose doctrines are recognizably psychoanalytic in content rather than only in name? Latter day psychoanalytic theoreticians that come to mind are the very influential Heinz Kohut, who pioneered the socalled “selfpsychology”, and the so-called “object relations” theorists, who include not only the leading Otto Kernberg but also Harry Guntrip, W. R. D. Fairbairn, Donald Winnicott, and others. Thus, Heinz Kohut, for example, downgrades Freud's Oedipal, Instinctual factors in favor of pre-Oedipal, environmental ones as the sources of the purported unconscious determinants of personality structure. More generally, insofar as these post-Freudian theories are indeed recognizably psychoanalytic, they do of course embrace some version of the repression-aetiology. And, furthermore, they rely epistemically on free association in the clinical investigation of purported pathogens and other unconscious determinants of behavior, while lifting repressions to effect therapy.
But, I submit, precisely to the extent that these outgrowths of Freud's ideas are thus recognizably psychoanalytic in content as well as in method of inquiry and therapy, my epistemic critique of Freud's original hypotheses applies with equal force to the aetiologic, developmental and ther-apeutic tenets of these successors. How, I ask, for example, can Kohut possibly claim better validation for his species of unconscious determinats than Freud can for the sexual ones? Moreover, it is just ludicrous to pretend with Flax (1981, p. 564) that my focus on Freud in appraising psychoanalytic theory epistemically is akin to the anachronistic procedure of “throwing out physics because there are unresolved problems in Newton's theory”. For this purported analogy suggests misleadingly that the epistemic difficulties which beset Freud's original formulations have been overcome by the much vaunted post-Freudian formulations of self-psychology and object relations theory. And it overlooks the logical incompatibility of the most influential of these versions: As Robbins (1980, p. 477) points out, Kohut's and Kernberg's views are “fundamentally antagonistic” to one another, being rooted in a schism between Melanie Klein and W. R. D. Fairbairn. (See Grünbaum, 1983b for details).
True, there are elements in some of the post-Freudian theories that give less emphasis to repression, both aetiologically and therapeutically, than the received doctrine. For example, self-psychology gives significant aetiologic weight to the absence of empathic mirroring in early childhood. Yet, as Morris Eagle has shown (private communication, and in press), these ingredients of the post-Freudian theories are at least as flawed epistemologically as the repression model that was found seriously wanting in this essay. Hence it is futile to adduce these modifications, as the disciples of self-psychology and object relations theory are wont to do, as improvements upon Freud's original hypotheses, whose articulations were more lucid and more amenable to scrutiny.
Indeed, there is not even agreement among the post-Freudians in regard to the probative value that may be assigned to One and the same case study material; While Kohut claimed clinical support for his theory from his re-analysis of Mr. Z.-a patient whose prior analysis had been a traditional one—the contemporary Chicago analyst Gedo (1980) harshly discounts the scientific quality of Kohut's case study material. And he concludes (p. 382) that the “theoretical inferences” drawn by Kohut from his clinical observations “fail to carry scientific conviction.” A similarly negative assessment is reached by the psychoanalytic psychologist F. J. Levine (1979), an ardent exponent of psychoanalytic methods of investigation and therapy. On the other hand, Ferguson (1981, pp. 135–36) believes that Kohut's case history of Mr. Z. is “a crystalline example of the fact that a progressive theory change [in L. Laudan's sense] has taken place in psychoanalysis.” But, Ferguson then seems to damn it with faint praise, saying “the case of Mr. Z. provides something of a ‘confirming instance’ of the new theory.”
No wonder that the psychodynamically-oriented psychologists Fisher and Greenberg (1977) reached the following verdict: “The diversity of the secondary elaborations of Freud's ideas is so Babel-like as to defy the derivation of sensible deductions that can be put to empirical test” (p. ix).
I am indebted to Benjamin B. Rubinstein and to Rosemarie Sand for their valuable comments on the first draft. And I thank the Fritz Thyssen Stiftung of Cologne for its support of my research.