Was Freud a Pseudoscientist?

The following is an extract from an essay titled “Are Freud’s Critics Scurrilous?”, translated and published in Le livre noir de la psychoanalyse (Editions des Arènes).


‘He thought it wrong of Rank to propagate ideas that had not been properly tested.’ (Sigmund Freud: Life and Work, E. Jones, 1957, Vol.3 p.71)

It is a pity that the word science was ever introduced into the dispute over Freud’s claims to knowledge, though it is worth remembering that the term was introduced by Freud himself and that his critics employed it in order to counter his pretensions It would spare readers much tiresome rationalisation of Freud’s deficiencies if it were clearly understood that the charge that they must meet is not that Freud was a poor scientist but that he was a tendentious interpreter of the phenomena he purported to account for. It would be more accurate to call him a pseudo-hermeneut and psychoanalysis, a pseudo-hermeneutics.

There is a question which some deem more important than the question of Freud’s trustworthiness or that of his followers and which makes any enquiry into this trustworthiness an irrelevant digression. This is whether an hypothesis is testable and therefore not impugnable as pseudo-scientific. Adolf Grünbaum applies the testablity criterion to psychoanalysis and finds that contrary to Popper’s contention that psychoanalysis is untestable and therefore pseudo-scientific it is testable and therefore is not pseudo-scientific But the testability of a theory cannot serve as a demonstration that it is not pseudo-scientific. If it could then sun-sign astrology which is for many the paradigm of a pseudoscience – and which Popper proffers as an example of a pseudoscience – would have to be denied that status for it certainly is open to empirical assessment and has even been declared falsified.


Consider the argument that since Freud manifestly changed his mind about certain issues he must be exonerated from the charge of pseudo-science. Did Hitler’s elevation of the Japanese to the status of yellow Aryans show that the Nazi version of the racist theory of history was therefore not pseudoscientific? Did those who invoke this criterion to exonerate Freud really think it pertinent to cite cases in which Freud had changed his mind without their showing that it was fresh observations that caused him to do so? And did it not occur to them to consider the more notorious examples of those Freudian theses which provoked the charge of dogmatism, e.g., the Oedipus complex?


The history of science is replete with cases where advocates of a theory have clung to it in spite of apparently falsifying data and have later been vindicated. Something more than mere tenacity is at issue. Sometimes as in the case of Freud the accusation was the stronger one that he reported his theory to have been confirmed when he must have known he was not in a position to do so.

Karl Popper himself was occasionally confused on the issue of the bearing of falsification-evasion on the pseudo-scientific status of a theory for an anecdote he relates in support of his criterion of falsification-evasion really supports a different criterion, that of treating the capacity of a theory to explain away disconfirmatory data as further confirmation of the theory. Popper recounts producing a counterexample to Adler’s theory of neurosis and of Adler explaining it away and adding that what entitled him to do so was ‘his thousandfold experience’ to which Popper replies ‘And now I suppose your experience is a thousand-and-one fold.’ (Conjectures and Refutations, K. Popper, 1968, p.35) Popper is not here reproaching Adler for evading falsification but for treating his ingenuity in explaining away apparent falsification of the theory as further confirmation of the theory (‘And now I suppose your experience is a thousand-and-one fold’) Adler is being charged not just with falsification-evasion but with spurious confirmation. The same implication follows from Popper’s complaint of ‘the stream of confirmations’ (1968, p.35) This is not just a complaint as to the untestability of a theory but as to spurious confirmation claims.


I once heard an anecdote about J. Edgar Hoover, the founder of the FBI to the effect that when he had decided to monitor the phone of someone suspected of subversion he would prepare two judgements, one headed ‘subversive’ – for cases in which incriminating conversations were overheard – and the other ‘cunning subversive’ – for the cases in which they were not.

The same practice has been imputed to Freud, but before we decide on the justice of the imputation we must be clear as to what the moral of the Hoover anecdote is. The moral is not, as a crude falsificationist might think, that Hoover ought to have declared the subject under surveillance to be an innocent, non-subversive because no incriminating conversations were overheard. That question ought to remain sub judice. What Hoover did which is reprehensible and aligns him with Freud (and Adler in Popper’s anecdote) was not that he failed to exonerate when there were no incriminating conversations but that he convicted in spite of their being none.

The parallel in the Freudian case is the objection of critics to Freud’s announcing that his theory had been vindicated by experience when the most he was entitled to assert was that it had not encountered exceptions he could not explain away. The suspicion that Freudians must allay is that the reason analysts in general had not encountered exceptions is that Freudian theory provides no clear account of what an exception would look like. In the Dora case history Freud wrote: ‘I can only repeat over and over again – for I never find it otherwise – that sexuality is the key to the problem of the psychoneuroses and of the neuroses in general.’ (SE 7, p. 115) What makes this a spurious claim rather than just a mistaken one in the eyes of Freud’s critics is that Freudian theory provides no sufficiently determinate conception of what would constitute ‘finding it otherwise’. It is therefore not surprising that Freud could claim after thirty years of practice that ‘all my experience shows that the neuroses are based on sexual instinctive forces’ (in the Three Essays, fourth edition, SE 7, p. 163). It is in its bearing on claims like this that the notion of untestability can be appropriately introduced and where its implications are most damaging.

The reviewer (for the TLS) of a volume containing Freud’s major case histories says of Freud: ‘He writes as if he had, at the back of him, a great body of tested doctrine. The result is that what seems to the uninitiated reader the most obvious shuffles and the crudest analogies are introduced briefly and, as it were, peremptorily as if the writer were a scientific man referring to something as well established as the atomic weights of the chemical elements.’ It misrepresents this objection to treat it as an objection to the untestability of the theses criticised. It seems to answer better to the notion of spurious confirmatory and instantiating reports.


But what of Grünbaum’s argument that Freud could test his infantile reconstructions and etiologies through the therapeutic effect of his patient’s acceptance of them? (The Foundations of Psychoanalysis, 1984, p.129). Therapeutic success could not confer testability on psychoanalytic theses which are not independently testable any more than the cures at Lourdes could confirm or falsify the doctrine of the Immaculate Conception.


What follows as to the scientific status of Freudian theory from the fact that its promulgators may have been disingenuous or dishonest? Morris Eagle states clearly the thesis I believe to be both mistaken and pernicious. He maintains that what matters is not ‘the methodological practices and attitudes of individual analysts (including Freud)’ but ‘the independent logical structure of psychoanalytic theory’ that is ‘whether or not certain psychoanalytic propositions can be treated as authentic hypotheses.’ (Review of Frederick Crews’ Skeptical Engagements in Contemporary Psychology, 1988, p.104). The same view is taken by Grünbaum who wrote: ”The scientific value of Freud’s hypotheses for the study of man does not turn on his own intellectual honesty or methodological rectitude, Even if all psychoanalysts were dishonest…this would not prevent non-analysts from appraising and using their theory…’ (personal communication). What is wrong with this emphasis on the logical properties of the theory is that it does not explain why one should expend energy on assessing a theory the evidence for which one has good reason for distrusting Theories are not like Mt. Everest. We don’t undertake the arduous task of assessing them merely because they are there. We want reasons for thinking they might be true. In 1913 the physician-author of a paper on Freud wrote in connection with his own conviction as to the truth of Freud’s claims: ‘To deny the evidence of these psycho-analytical findings with regard to infantile sexual phantasies is to deny the intellectual integrity of Freud and his followers.’ (M. Wright. ‘The Psychology of Freud’, Medical Magazine, 1914, p.145.) This is correct and the failure to acknowledge it and to insist on transforming the issue into one of logic deflects interest from the central question of the whether the grounds advanced for crediting Freudian theory are good enough to warrant further enquiry. Analysts themselves, including Freud, acknowledge that the evidence they are able to produce for assessment is not the basis of their conviction. This lies in features of the analytic situation which cannot be produced for inspection – imponderabilia. In the case history of the Wolf Man Freud wrote: ‘It is well known that no means has been found of in any way introducing into the reproduction of an analysis the sense of conviction which results from the analysis itself.’ (1918, SE 17, p.13 )


Is Freud a pseudo-scientist? Yes.

Is this because his theories are untestable?

No. (Though some of them are untestable.)

Is this because he arbitrarily refused to capitulate to reported falsifications? No. (Though he sometimes does arbitrarily refuse to capitulate to falsifications.)

Why is Freud a pseudo-scientist then?

The strongest reason for considering Freud a pseudo-scientist is that he claimed to have tested – and thus to have provided the most cogent grounds for accepting – theories which are either untestable or even if testable had not been tested. It is spurious claims to have tested an untestable or untested theory which are the most pertinent grounds for deeming Freud and his followers pseudoscientists (though pseudo-hermeneut would have been a more apposite and felicitous description).

Frank Cioffi is the author of Freud and the Question of Pseudoscience.

Comments are closed.