Now we turn to the philosophical problems that have been raised concerning the scientific status of psychoanalysis. Freud always insisted, particularly in "The Question of a Weltanschauung [world-view]," that psychoanalysis is just a part of the scientific world-view, and applies to psychological phenomena the same standards that are typical of the other natural sciences. But this would seem to suggest that psychoanalysis must meet scientific standards of evidence, and this seems to be questionable. What kind of evidence does the theory have?

The evidence for the theory is clinical: By the interpretation of dreams, free association, and transference, the analyst provides the patient with a picture of childhood fantasies, anxieties, and conflicts that explain present neurosis.

Since the evidence only comes from the interpretations, and the interpretations are based on the theory, how can the evidence support the theory?

Freud: If the interpretation does not "tally with what is real" in the patient, the patient will not be helped by the interpretation. A false interpretation would be incapable of improving the patient's condition.

The problem of suggestion: Is the patient's acceptance of an interpretation caused by the influence of the analyst? Is the cure caused by the truth of the interpretation, or by the expectation of cure? How can we be sure that the truth of an interpretation is causally relevant to its cure?

One common view was originated by the philosopher Karl Popper, who claimed that Freud's theory is actually a pseudo-science, i.e. it has the trappings and the suits of science, but fails to live up to the evidentiary standards of science, because all possible evidence is already infected by the theory. Any verification of the theory is essentially a self-fulfilling prophecy.

Popper's criticism of psychoanalysis was based on a view of science that became known as "Falsificationism": Scientific theories are never truly verified. Moreover, to be always verified is not a virtue in a scientific theory. (Cf. Popper's The Logic of Scientific Discovery and Conjectures and Refutations)

Example: Is Freudian psychoanalysis true?

Q. How do we know that repressed memories of infantile sexual desires are the causes of neurosis?

A. These desires are revealed in our dreams, "Freudian slips," free associations, and other "symptoms". They are the "latent content" expressed symbolically.

Q. How do we determine the true meaning of these symbols?

A. We interpret them by applying Freud's theory.

Q. What if the patient denies the interpretation?

A. The patient is "resisting," which indicates that the interpretation is correct and therefore disturbing.

No evidence could possibly be incompatible with the theory.

The characteristic of a scientific theory is that it forbids certain states of affairs: it is incompatible with certain possible evidence. Theories that are compatible with all possible evidence are "pseudo-science." (For example, Newtonian physics makes very precise predictions, and the theory was abandoned when some of these turned out to be inaccurate. But, according to Popper, Freudians can interpret any possible evidence; they could not specify a prediction whose falsity would overthrow the theory.)

Freud's answer to challenges of this type was always the same: If the theory is false, it will not enable patients to get over their neuroses. This is a good answer to Popper. But it has strict implications for the theory: It means that belief in the theory must depend on our ability to demonstrate somehow that:

1. Psychoanalysis does in fact reliably lead to amelioration of neurosis, and

2. The correctness of the Freudian interpretation is causally responsible for the amelioration, i.e. the amelioration would not be possible with a false interpretation; moreover, non-psychoanalytic treatments would have to fail, at least in the long term, or at least to be substantially less successful than psychoanalysis. (If neurosis is caused by repression, then any treatment that does not undo the repression cannot produce a genuine cure. If the symptoms are treated, but not the cause, Freud's theory requires that other symptoms must take their place.)

Alternatives to science?

Instead of addressing these scientific questions directly, e.g. by doing epidemiological studies of the effectiveness of psychoanalysis relative to other types of therapy, many commentators have taken the view that psychoanalysis should not be judged by the standards of natural science. It is a different type of inquiry, a human science rather than a natural science, and therefore interpretation, understanding, and empathy play a role in it that they don't play in physics, chemistry, traditional medicine, etc.

Paul Ricoeur: A hermeneutic view of psychoanalysis

[Hermeneutics: Textual and symbolic interpretation; applied to psychoanalysis, it is the view that psychoanalysis is not a natural science, but a fundamentally interpretive science, and therefore has to be judged by different criteria.]

What does verification mean? What are the "facts" for psychoanalysis?

Natural science: Facts are "observables," and theories are verified or falsified by the observables that they predict or fail to predict.

Fact in psychoanalysis are determined not by simple observation, but by the entire "analytic situation". Four criteria of facts:

1. They are essentially limited to what may be said--i.e. Not the dream, or memory, or feeling, but what the patient reports of it.

2. They are essentially what is said to the analyst--that is, they are said in the context of the relationship between analyst and patient.

3. They concern psychic reality rather than material reality.

4. They are limited to what may be put in the form of a narrative; remembering something, in the analytic situation, means placing it meaningfully within a coherent narrative of one's life.

How does the theory relate to the facts?

Hermeneutical: The facts are texts to be interpreted by the interpretive theory; the fundamental notion is the semantic one of meaning. The goal is understanding through motives, thoughts, etc.

Economical: The facts are effects to be explained by the causal theory; the fundamental notion is the dynamic one of force. The goal is explanation through causes.

The economic model reflects Freud's overemphasis on the natural-scientific model of explanation.

But the hermeneutical model must always proceed from the economic.

Understanding through motives requires explanation through causes--resistance, repression, etc.

What is the truth of a psychoanalytic claim, and how can it be verified?

The truth is fundamentally one of self-understanding; the verification is in the conjunction of theory, interpretation, therapy, and "the narrative structure of the analytic experience." A good interpretation:

1. Must be coherent with the basic Freudian principles;

2. Must satisfy the "universalisable rules" for interpreting the unconscious

3. Must be successful in economic terms, i.e. It must be therapeutically successful in helping the patient to work through problems.

4. Must place the events of the patient's life into a coherent and intelligible narrative.

Consequences of a hermeneutical view:

The patient's self-understanding is the basis for cure.

This is something that only the patient is capable of judging, and something that sets psychoanalysis apart from natural science:

Self-knowledge is obviously not a factor in any causal interaction as understood by natural science;

Causal connections in natural science are deterministic, but psychoanalytic self-understanding leads to autonomy, or freedom from determinism.

Psychoanalysis therefore results in the "breaking" or the "overcoming" of causal connections.

Note that this last is essentially a bogus argument. If we say that A causes B, we mean that the removal of A will bring an end to B. If repression causes neurosis, lifting the repression should relieve the neurosis. But this is obviously not a breaking of the causal connection; on the contrary, the causal connection between A and B is corroborated-though not necessarily proven-if removing A removes B.

Note also that the hermeneutic view admits that the causal connection plays an essential role in the interpretation, i.e. that it is the causal link between the repressed ideas and the symptoms that is supposed to justify us in thinking that the interpretation is true. Therefore, in spite of the "hermeneutic" aspect of psychoanalysis, there is no escaping the fact that the interpretations depend on Freud's causal theory of neurosis.

The structure of Freud's argument for the truth of psychoanalytic interpretations

First causal inference: Improvement in symptoms is caused specifically by the lifting of repression.

(How do we know that this is not a placebo effect? According to Freud, because each symptom requires a separate repression as its cause, a separate lifting as its cure.)

Second causal inference: Repression is the necessary condition for maintaining a neurosis.

It follows from these that removal of repression leads to therapeutic success.

The "Master Proposition" or "Necessary Condition thesis" :

a. Neurosis can only be eliminated by "conscious mastery" of repression and of the repressed thoughts.

b. Only psychoanalysis generates the necessary insight into the "pathogens" of the neurosis.

If the "Master Proposition" is true, then the following consequences can be drawn:

1. The cure of neurotic symptoms vindicates the psychoanalytic interpretation.

2. Interpretations are not "self-fulfilling prophecies," i.e. their acceptance by the patient alone is not what makes them therapeutic.

3. Only psychoanalysis can produce a genuine cure.

Critical questions

Is psychoanalysis an effective treatment?

Is it more effective than other therapies?

Is psychoanalytic insight into the unconscious a necessary condition for the cure of neurosis?

Freud's arguments for the necessity of psychoanalysis lack comparative data, i.e., data on either:

the presence of psychoanalytic "causes" of hysteria in non-neurotic people

the progress of neuroses treated by non-psychoanalytic therapies, or by no systematic form of treatment, or not treated at all.

Again, this is a serious problem given that the evidence for Freudian interpretations is only the therapeutic success they are said to make possible.

Some epidemiological studies of treatment effectiveness:

Fisher and Greenberg (1977)

"(1) Psychoanalysis has been shown to be consistently more effective than nontreatment with chronic neurotic patients, and (2) psychoanalysis has not been shown to be significantly more effective than other forms of psychotherapy with any type of patient."

Lambert (1976) Study of spontaneous remission: 43% remission rate for people on the waiting list for psychotherapy

Columbia Psychoanalytic Center Research Project (1985):

Psychoanalysis produces improvement in slightly more than half of cases; it is better than no treatment at all.

General verdict: Psychoanalysis is

a. better than no treatment at all, though neurosis is frequently overcome without any systematic treatment

b. No better than other "psychotherapies" (i.e. "talking" therapies)

c. Not as effective as drug and behavioral therapies

If psychoanalysis is no better than other therapies, one can ask what it is that successful talk therapies have in common, i.e. what are the factors on which therapeutic success depends. Research has identified the following as therapeutic techniques common to successful psychotherapies:

1. Well-practiced rituals;

2. A special vocabulary;

3. A knowledgeable manner;

4. Personal charisma of the therapist.

If this is what success depends on, however, obviously the success of the therapy has no implications for the truth of the theory on which the therapy is based. In particular, Freudians can no longer call on the results of therapy as evidence for the truth of psychoanalytic interpretations of dreams and symptoms. The successes achieved by psychoanalysis, such as they are, would have to be regarded as placebo effects.

[Note: This is not to say that it is definitively established that successful psychoanalyses are placebo effects. It only means that, at the present time, there is no evidence that they are anything other than placebo effects.]

Perhaps a worse problem is Freud's own admission, later in his career (1926), that psychoanalysis is not really necessary for a cure: "As a rule our therapy must be content with bringing about more quickly, more reliably, and with less expenditure of energy than would otherwise be the case the good result which in favourable circumstances would have occurred by itself."

Free association and suggestion

How did Freud deal with the accusation, with which he was quite familiar, that psychoanalysis succeeds by suggestion, i.e. as a placebo effect? His standard answer was that free association inevitably leads from the current symptoms, including slips of the tongue and dreams, to their original cause. Since free association represents the unconscious "speaking" to the therapist, there is no way for outside influences to interfere with the chain of associations back to the original repressed desires. All that is required is that patient and analyst obey the "fundamental rule," i.e. that all conscious censorship be relaxed and that ideas are uncritically released as they occur to the patient. The unconscious censorship will still disguise the true motives, but the process of free association will gradually lead back to the repressed material.

Freud's basic defence was expressed in what has been called the "Tally Argument" (1917): "There is a risk that the influencing of our patient may make the objective certainty of our findings doubtful...If [this objection] were justified, psychoanalysis would be nothing more than a particularly well-disguised and particularly effective form of suggestive treatment and we should have to attach little weight to all that it tells us about what influences our lives, the dynamics of the mind or the unconscious....Anyone who himself has carried out psychoanalyses will have been able to convince himself on countless occasions that it is impossible to make suggestions to a patient in that way....After all, his conflicts will only be successfully solved and his resistances overcome if the anticipatory ideas he is given tally with what is real in him."

In Freud's own time, however, there was already a great deal of study of the influence of observer expectations on experimental outcomes, not only in psychology but even in studies of animal behavior, which might appear to be objective and not subject to much interpretation.

Here is how the philosopher Bertrand Russell viewed the influence of observers on animal behaviour studies:

"All the animals that have been carefully observed have behaved so as to confirm the philosophy in which the observer believes before his observations begin. Nay, more, they have all displayed the national characteristics of the observer. Animals studied by Americans rush about frantically, with an incredible display of hustle and pep, and at last achieve the desired result by chance. Animals observed by Germans sit still and think, and at last evolve the solution out of their inner consciousness."

By itself, this implies nothing about Freud. But we have to ask, was Freud aware of the influence of his own expectations, or "demand characteristics"-i.e. the influence on the behavior of subjects by their sense of what the observer demands from them? This is a particularly acute problem in psychoanalysis, since a major therapeutic role is given to "transference," in which the relationship of patient to analyst (supposedly) recapitulates earlier relationships in the patient's life, and the bond that develops becomes a source of authority for the analyst's interpretations.

Apparently Freud did not think that there was a serious risk of contaminating the evidence of free association by suggestion. Perhaps the nature of "free association" would protect against this. But did Freud really use free association? That is, were the associations of his patients really free, or did he subtly influence them in the direction determined by his own theoretical presuppositions?

Freud often asserted the "freedom" of free associations. Yet he just as often allowed a larger role for the analyst in the treatment:

"The treatment is made up of two parts--what the physician infers and tells the patient, and the patient's working over of what he has heard. The mechanism of our assistance is easy to understand: we give the patient the conscious anticipatory idea and he then finds the repressed unconscious idea in himself on the basis of its similarity to the anticipatory one. This is the intellectual help which makes it easier for him to overcome the resistances between conscious and unconscious."

In other words, Freud felt it was necessary to give his patients "conscious anticipatory ideas" to help them with their "free" associations!

This can be verified by a look at any of Freud's case studies, where, without seeming to notice that it conflicts with the idea of free association, he explicitly discusses the "anticipatory ideas" that he uses to guide their "free" associations. We saw this already in the seduction theory, when he would "warn" patients that traumatic childhood scenes were going to "emerge" from the analysis-clearly not trusting their associations to lead there of their own accord.