It was 1986 and I was taking a course with Henry Kyburg, Jr at the University of Rochester where I was enrolled to do my PhD in philosophy. One day Henry gave me a book which was titled Foundations of Psychoanalysis: A Philosophical Critique authored by Adolf Grünbaum. Henry said to me, “You are a psychoanalyst. See, what you can do with this book.”
At that time I was not even familiar with the name of Adolf Grünbaum and had no idea of the enormity of the task that was entrusted to me by Henry. When I read the book for the first time I felt overwhelmed. This was the most severe and most sophisticated criticism of the scientific status of psychoanalysis. Apart from the erudition of the author, his logic seemed impenetrable to me and his arguments appeared invincible. The book also hurt me somewhere. I felt a pain as if something very precious was being taken away from me.
I was initiated into psychoanalysis at a very young age. I started my training when I was not even 21. My personal experience with psychoanalysis had convinced me of its truth and efficacy at least in some very important domains of human mind. But as a student of philosophy I also knew that psychoanalysis can ignore these criticisms only at the peril of its own existence as a scientific discipline. I realised that my love for psychoanalysis is not sufficient to save it from the ever-increasing criticisms of its scientific status. I also realised that logic has to be answered with logic. So, the work began. It was not an easy task to defend psychoanalysis against the most sophisticated philosopher of science of the century. At times I despaired; but Henry won’t let me leave the project. Once he said to me, “Do you believe in psychoanalysis?” I said, ‘Yes, of course, I do.” Henry was a very quiet person. He simply said, “Then defend it.” So, this book is a defence of psychoanalysis against two of the most severe objections against psychoanalysis. It started as my PhD dissertation and the research work was carried further when I went to the University of Pittsburgh as a Fulbright fellow for post-doctoral work.
Actually it was Karl Popper who set the ball rolling. Popper was against inductive logic and argued that the proof of the inadequacy of the inductive criterion of demarcation is that it accords scientific status to psychoanalysis which, in his opinion, was no better than astrology and Marxism and all three of them, according to him, are pseudo-sciences. Popper claims that his own criterion of falsifiability does not accord scientific status to psychoanalysis because psychoanalytic hypotheses are not falsifiable.
Grünbaum’s entire endeavour in Foundations is to prove Popper wrong. His agenda in the book is to show that psychoanalysis satisfies Popper’s criterion of falsifiability which he does by listing various psychoanalytical hypotheses that have testable consequences. Then he proceeds to show that psychoanalysis fails to satisfy the inductive criterion of scientificality. He argues that the inductive criterion requires more than conceivable falsifiability for a theory to be confirmed. It requires positive evidence for the credibility of a theory and the evidential credentials of psychoanalysis, according to Grünbaum, are miserable. He argues that the methodology of psychoanalysis is faulty, its evidence unreliable, and the reasoning on which its arguments are based are fundamentally logically flawed.
Grünbaum’s attack on psychoanalysis is two-fold: he argues that the arguments supporting the hypotheses of psychoanalysis are both (i) unsound and (ii) invalid; unsound because the clinical data on which they are based are contaminated by the suggestive influence of the analyst, and invalid because they fail to meet the criterion of eliminative induction in establishing their causal hypotheses. The second charge is true for (a) repression aetiology, (b) Freud’s dream theory, and (iii) his theory of paraprexes.
My book, The Scientific Status of Psychoanalysis: Evidence and Confirmation, deals with two objections – the one related to the charge of unreliability of clinical data which we shall call the ‘suggestibility charge’, and the other related to the invalidity of the argument leading to repression aetiology. The suggestibility charge is like a ghost. It has haunted psychoanalysis since its birth. It is this old charge that has been presented by Grünbaum in a logically sophisticated way.
To prove the suggestibility charge Grünbaum formulates an argument based on the last lecture of Freud in his General Introduction to Psychoanalysis, titled Analytic Therapy, in which Freud has directly addressed the suggestibility charge against psychoanalysis. Grünbaum considers Freud’s argument in Analytic Therapy as the best defence against psychoanalysis. So, he reconstructs Freud’s argument and dubs it ‘The Tally Argument”.
The premises of the argument, he calls ‘the necessary condition theses’ or NCT in short. He also calls it “Freud’s Master Proposition”. Then Grünbaum proceeds to show that the Tally Argument is false by showing that the consequences of the argument are false. With the support of the Tally Argument gone, Grünbaum claims, psychoanalysis is without any defence against the suggestibility charge. Hence, it faces an epistemic liability of enormous proportion. The Tally Argument ascribes to Freud extremely strong and rigid premises in the form of a deductive argument. The argument goes like this:
The Tally Argument:
(1) The psychoanalytic method of giving interpretation etc. is a necessary condition for obtaining insight into the causal mechanism of one’s neurotic symptoms.
(2) This insight is a necessary condition for the cure of one’s neurotic symptom.
These two premises are called by Grünbaum “the necessary condition theses” of Freud or NCT in short.
Let us call a patient who has been cured by psychoanalytic treatment P. Given the existence of such a patient P, Grünbaum derives two conclusions from the above premises:
- The psychoanalytic interpretations given to P by his analyst are indeed correct and, as Freud puts it, “tally with what is real” in P.
- Only psychoanalytic treatment could have wrought this cure of psychoneuroses.
This argument is a perfectly valid argument and appears to be in consonance with psychoanalytic theory. Grünbaum says that with the help of this argument Freud aims to solve a number of problems related to psychoanalysis. If P is cured by psychoanalytic treatment, then the interpretations given by his analyst must be true leading P to have insight into the causal mechanism of his symptom. In other words, they must tally with what is real in P. So, the clinical data are saved. Similarly, the clinical hypotheses also must be true. If the hypotheses are true, its methodology also must be true etc. So, if the Tally Argument is sound, Freud could defend his theory with the help of the success of his therapy. “Magnificent, if true!” exclaims Grünbaum.
But unfortunately, Grünbaum claims, the Tally argument is false because its consequences are false. Let us see how. The Tally Argument appears to be a very simple argument. But a deeper reflection will show that it ascribes a very strong claim to Freud. Logically speaking, if X is a necessary condition for Y, then Y cannot occur when X is absent. So, the claim that Grünbaum ascribes to Freud is that not a single case of cure of psychoneuroses can occur without insight or without psychoanalytic therapy.
It is easy to falsify such a strong claim. And Grünbaum did just that. He showed that contrary to Freud’s claim, cure of psychoneuroses have occurred without psychoanalytic treatment. There are cases of spontaneous cure of neurotic symptoms which, Grünbaum claims, Freud initially denied but later admitted, falsifying his own NCT. He also cited cure by rival therapies, like behaviour therapy etc., which also falsifies Freud’s NCT. If NCT is false then psychoanalysis is left with no defence against the suggestibility charge and it faces an epistemic liability of enormous proportion.
The Tally Argument is deeply embedded in the theoretical structure of psychoanalysis. Grünbaum discusses extensively how important it was for Freud to enunciate his NCT and how deeply the failure of NCT affects the credibility of psychoanalysis as a science. If the Tally argument is correct, then Grünbaum’s criticism of Freud could not be denied. However, I found it entirely unbelievable that Freud would deny that even a single case of psychoneurosis could be cured spontaneously or with the help of other forms of therapy. Questions arose in my mind. Did Freud really make such strong claim? Did he really deny that there could be spontaneous cure of neuroses? I decided to check for myself. And results led me to challenge the Tally Argument.
Philosophically the reconstruction of an argument of a philosopher is best supported by textual evidence. I have with the help of extensive textual support shown that contrary to Grünbaum’s assertion from the very beginning Freud has never made such strong claims and never denied spontaneous cure in case of psychoneuroses. The Tally Argument is a brain child of Grünbaum, not of Freud. It was my task to prove this with the help of textual evidence and also with the help of the logical formulation of psychoanalytic theory. I formulated the important hypotheses of the theory in the form of theorems and derived logical consequences from them and showed that the strong claims of the NCT are not supported by the logical consequences of the theory. If Tally Argument is not Freud’s argument, its falsification has no bearing on the suggestibility charge against psychoanalysis.
But then what was Freud’s argument? I have given an alternative formulation to Freud’s argument which merely claims the insufficiency of suggestion to bring about durable cure. The argument is weak and may not be able to tackle the effects of all forms of subtle suggestions. There is no doubt that in order to give a credible answer to the suggestibility charge one must go deeper into the nature of suggestion and suggestibility. The suggestibility charge is a complex charge because it is so amorphous. It must be given a definite shape in order to be logically answered. In chapter 4 of my book, I have attempted this difficult task.
While attempting this task a fascinating aspect of suggestibility occurred to me. On the very first day of my personal analysis my analyst had said, “Be absolutely frank and honest and whatever you say here would remain confidential.” I remembered how this instruction influenced me to overcome my resistances and be truthful in my analytic sessions. Similar things have happened with my own patients and I am sure that this is a regular occurrence in analytic scenario. Patients often tell us, “Since I am supposed to be truthful and honest here I would like to tell you that … etc. etc.” Thus, suggestibility is also a factor that purifies the data instead of contaminating it. If the charge of suggestibility is true then the purification of data in the state of suggestibility also is true. In other words, suggestibility can be utilised both to contaminate the data as well as to purify the data. So, I decided to check for experimental evidence and readily enough found quite a few results where a truth-demand has resulted in greater veridicality of responses from subjects. In spite of that I have suggested certain measures to be adopted with regard to the clinical data and clinical situations that would help in minimising the effects of suggestion.
However, one may object that the very presence of the analyst is an uncontrolled variable in the analytic situation and the clinical situation will always contain this uncontrolled variable. I consider this problem not unique to psychoanalytic situation but common to almost all experimental situations. Our current scientific practice is to accept tentatively the findings of studies which may contain subtle sources of error bur more or less meet the standard criteria of scientific experimentation. For example, we do accept the results of the studies on smoking and cancer though all of them contain an ineliminable source of error, namely, heredity. The practice is rational but is in need of some theoretical support.
Henry Kyburg Jr’s theory of error provides such a justification. He advocates that it is possible to test the predictions of a theory in a probabilistic way even though all sources of error in an experimental situation may not have been controlled. We need to find out the long run frequency of error of a theory T, we can find out the probability of correct prediction of that theory. If that probability of correct prediction is within ‘acceptable range’ we can accept the theory in our rational corpora; the theory is eschewed otherwise. Kyburg believes that all our observations are fraught with errors. Even our observation in the area of physics cannot be free from error. Unless we have a principle of justification, we are not theoretically justified to accept the results of any scientific investigation. Kyburg’s theory provides that justification.
It was Grünbaum’s objection against Freud’s repression aetiology that had bothered me most. Like Grünbaum, I too believe that repression is the cornerstone of the theory of psychoanalysis. All other psychoanalytic hypotheses are based on the theory of repression. Grünbaum says that even if we suppose for the sake of argument that the data are free from contamination, the argument of repression aetiology is fundamentally flawed. He has reconstructed Freud’s argument which is based on Freud and Breuer’s joint publication Studies on Hysteria. Anna O – a patient of Josef Breuer – was relieved of her neurotic symptoms one by one as she could remember the apparently forgotten incidents with which her symptoms began. On the basis of this both Breuer and Freud arrived at the conclusion that the memory of those incidents was repressed and lifting of repression is the cause of cure. Grünbaum claims that in arriving at this conclusion they failed to eliminate the following two rival causal hypotheses: (1) repression could be the maintaining rather than the originating cause of the symptom. The original event may have caused the symptom and repression just maintained it. So repression could be just the maintaining cause devoid of all aetiological significance; and (2) the cure could be due to suggestion.
There is no doubt that this distinction between the originating and the maintaining cause is an artificial one. Can we say that a particular behaviour is a symptomatic behaviour if the behaviour is not maintained at all? Repression does not occur simultaneously with the occurrence of the event. So there will be a state of affairs where the event has occurred giving rise to the symptom but since repression has not occurred yet, the symptom is not maintained. But in such an event is there any symptom? This argument is simply unacceptable.
Now, how can we eliminate the rival candidate of suggestion? For one thing, Grünbaum has already said that the data are accepted to be free of contamination. So, no question of suggestion leading to cure arises. Even then one can deal with the problem. Since Freud’s time a lot of work in the area of suggestion has been done and we have acquired quite a bit of knowledge regarding what suggestion is capable of doing and what it cannot do. We now know that even a very strong suggestion given during hypnosis is remembered later to be a case of suggestion and has no effect on the real life of the person. A number of other findings like direct suggestion is stronger compared to indirect suggestion etc., also help in eliminating the suggestion hypothesis. With the elimination of these rival causal candidates one can give a valid formulation to Freud’s repression aetiology.
In defending clinical data it is not my intention to decry the need for experimental verification of the psychoanalytical hypotheses. Nor did I want to say that the clinical set up be the typical arena of the testing of psychoanalytical hypotheses. But in my mind there is no doubt that the clinical data are important. They are our source of knowledge and insight into the deep workings of human mind. Therefore, a summary rejection of the clinical data is a great loss not only for psychoanalysis but for any science claiming to deal with human mind and society.
I have tried to answer two basic objections raised by Grünbaum. But a lot yet remains to be done. Grünbaum’s objections against the method of free association, against Freud’s theory of dreams, paraprexes etc., remain to be answered. I hope psychoanalysts will take up the challenge and will help in establishing what is true about psychoanalysis.
Pushpa Misra was a Fulbright Fellow for the year 1993–1994 for postdoctoral research at the University of Pittsburgh, and holds a Master’s degree in psychology from Calcutta University and a doctorate in philosophy from the University of Rochester. Her academic interests range from psychoanalysis to philosophy of science, applied ethics, and literature. Her book, The Scientific Status of Psychoanalysis: Evidence and Confirmation, is published this week by Karnac.
Reviews and Endorsements
‘Given the centrality of the so-called Tally Argument attributed to Freud in Adolf Grünbaum’s critique of the scientific status of psychoanalysis, Pushpa Misra’s claim that it is “the brainchild of Grünbaum, not of Freud” represents an important contribution to the debate over the nature and foundations of psychoanalysis.’
– Donald L. Carveth, PhD, RP, FIPA, Emeritus Professor of Sociology and Social & Political Thought, Senior Scholar, York University, Canada