Sigmund Freud, answered from the texts and cited to the page.
The question turns on what one means by empirical. If the standard is the controlled experiment, the prediction verified in advance, the laboratory result that can be reproduced by any competent hand — then psychoanalysis does not meet it, and I have never pretended otherwise. But to demand that standard of a science of the mind is to demand that it be a different science than its subject matter permits.
Consider what the method actually does. A dreamer wakes with fragments — absurd, disconnected, apparently meaningless. The manifest dream, as it appears to consciousness, is a mass of symbols representing elements in something deeper: what I call the latent dream ideas.1 The dream-work — condensation, displacement, dramatisation — translates the latent into the manifest, and the task of analysis is to reverse that translation.2
Condensation works by finding or creating common points between dream-thoughts, so that a single element in the manifest dream corresponds to a junction of several underlying thoughts, overdetermined, carrying multiple meanings at once.3 Displacement, for its part, shifts the energic emphasis — what stands vivid and central in the manifest dream has often occupied only a peripheral position in the dream-thoughts, and vice versa, which is precisely why the dream seems strange and incomprehensible to the waking mind.4
Now, where is the empirical test in all this? It lies in the analytic process itself. Take a remembered element, track it back by free association, and the forgotten elements stir; coherence emerges from incoherence.5 The interpretation is not declared in advance and then confirmed — it is constructed from the patient's own material, material the patient did not know he possessed, and often violently resists.
When the interpretation is correct, the resistance confirms it as surely as acceptance would; the symptom lifts, the memory returns, the disguised text becomes legible. That is the test — not prediction but decipherment, the same standard by which we judge any science of meaning. I have said plainly that the dream-work is a process the like of which had not previously been discovered in psychic life, and that its mechanisms — condensation, displacement, the regressive translation of thoughts into pictures — are new discoveries that richly repay the effort.6
A theory that abandoned the seduction hypothesis when the clinical material could not sustain it, that reworked the theory of anxiety when observation compelled revision, is answering to evidence. That is what empirical work looks like in a discipline whose material is the interior life. Those who demand that psychoanalysis meet the falsifiability criteria of physics are asking it to ignore the nature of its subject.
The challenge I put to the critic is the concrete one: explain the dream, the slip, the symptom, the transference — on some other account that does less violence to the material. Dream symbolism is a reality; the difficulty is not whether symbols exist but which images are legitimately symbolic and which are not, and that is precisely the kind of question the method is designed to answer.7
Psychoanalysis is a science of the mind, but a science of meaning, and its evidence is of a kind the physical sciences do not furnish. The material is the analytic situation itself: the free associations, the dreams, the symptoms, the transference, the resistances — and above all the point at which an interpretation, offered at the right moment, lifts a repression and produces conviction, recollection, and the disappearance of the symptom.8
The proof of an interpretation is not that it can be predicted in advance but that it makes the disguised text of the unconscious legible and that the patient's own material confirms it. The resistance an interpretation meets is itself evidence: what the patient most vehemently denies is often just what the analysis has touched. I do not claim the certainty of the laboratory, and I have said plainly that the theory is built on observation and revised where observation compels — the abandonment of the seduction theory, the reworking of the theory of anxiety, the late turn to the structural model are the record of a theory answering to its material.
Those who demand that psychoanalysis meet the criteria of physics ask it to be a different science than the one its subject matter permits.9101112
The manifest dream is a mass of symbols representing elements in the latent dream ideas.On Dreams
How the latent dream ideas generate the manifest dream is discovered by psycho-analysis, the translation from the latent to the manifest is the effect of the dream work.On Dreams
The newly framed common points of condensation enter as representatives of the dream-thoughts into the manifest dream-content, so that an element of the dream corresponds to a point of junction or intersection of the dream-thoughts, and with regard to the latter it must in general be called 'overdetermined.'Wit and Its Relation to the Unconscious
It manifests itself by the fact that what occupies the center of the manifest dream and is endowed with vivid sensory intensity has occupied a peripheral and secondary position in the dream-thoughts, and vice versa. This process causes the dream to appear out of proportion when compared with the dream-thoughts, and it is because of this displacement that it seems strange and incomprehensible to the waking state.Wit and Its Relation to the Unconscious
Take a remembered element of a dream, track it back and back by free association or other method, and you will find that, at one or two removes, the remembered element stirs up forgotten elements, and ultimately brings coherence out of incoherence.On Dreams
These condensations, displacements, regressive translations of thoughts into pictures, are new discoveries which richly repay our efforts in the field of psychoanalysis.A General Introduction to Psychoanalysis
The dream symbolism, in particular, it is easy to criticise; but, after all, dream symbolism is a reality. The point to investigate is, what dream images are legitimately considered symbolic and what not.On Dreams
In this way the transference is changed from the strongest weapon of the resistance into the best instrument of the analytic treatment. Nevertheless, its handling remains the most difficult as well as the most important part of the technique of analysis.The Question of Lay Analysis
But what is the mode of procedure which is applied in your analysis, and to what treatment do you subject your patient?" III To make myself perfectly plain to you, it will be necessary for me to acquaint you with certain psychological teachings which are not known beyond the circle of analysts and accordingly, not appreciated beyond this group.The Question of Lay Analysis
He thus makes the reader suspect that in all his description of such dissociated states he has never made the attempt to find the dynamic explanation for these phenomena.The Interpretation of Dreams
He naturally has the desire, you say, to meet the challenge, to explain the slip, and hence he says the first thing he can think of if it seems relevant. But that, you say, is no proof that this is really the way the slip happened. It might be so, but it might just as well be otherwise, you say. Something else might have occurred to him which might have fitted the case just as well and better.A General Introduction to Psychoanalysis
And yet he struggles, in the very interests of his malady, against one who would help him. How improbable this assertion must sound! And yet it is so, and if we are reproached with its improbability, we need only answer that this fact is not without its analogies.A General Introduction to Psychoanalysis
The central methodological controversy of psychoanalysis and the one on which its standing in the sciences turns. Popper's falsifiability criterion made psychoanalysis the textbook example of pseudo-science; Grünbaum's Foundations of Psychoanalysis (1984) — from a philosopher sympathetic to Freud's scientific ambitions — argued that the clinical setting cannot distinguish a correct interpretation from a suggested one, so the 'tally argument' fails and the causal claims are unsupported by the evidence Freud offered. The Freud wars (Crews and others) pressed the suggestion-contamination charge into a broader attack. Defenders (Drew Westen's empirical work on psychodynamic constructs, Solms's neuropsychoanalysis) argue that many core claims have found independent support. The dispute remains genuinely unsettled.