In 1977 Paul Ricoeur published an essay in the Journal of the American Psychoanalytic Association entitled "The Question of Proof in Freud's Psychoanalytic Writings" where he describes what constitutes a fact in Freudian psychoanalysis. His claim is that what is relevant as a fact in psychoanalysis is of a different nature than what counts as a fact in the natural sciences -- which is not an outrageous conception since what counts as a fact and as an observation in mathematics is also of a different nature than what counts as a fact in the natural sciences. Thus Ricoeur considers 4 criteria for what counts as a fact in psychoanalysis: (a) it must be a part of experience such that it is capable of being said. This means that facts in psychoanalysis are not observable behaviors, rather they are reports. Indeed, even symptoms (though they are partly observable) enter the field of analysis only in as much as they are in relation to other factors that are verbalized in the "report". This, explains Ricoeur, forces us to situate the facts of psychoanalysis inside a sphere of motivations and meaning. (b) What constitutes a fact in psychoanalysis is not only that it is something "sayable"-- according to Ricoeur's reading of Freud, it is also that it is said to another person. This other, he tells us, can be someone who responds, who refuses to respond, who gratifies, who threatens. This person may be a real existent person or a fantasy, spmeone present or lost, a source of anguish or the object of successful mourning. (c) Psychoanalytic facts concern a "coherent field of resistance that constitutes a psychic reality in contrast to material reality". Ricoeur recalls that, for Freud, in the world of neurosis it is psychical reality which is the decisive kind of reality. According to Ricoeur, the epistemological consequences of the distinction are considerable: "while academic psychology does not question the difference between what is real and what is imaginary -- insamuch as its theoretical entities are all said to refer to observable facts and , ultimately, to real [existing] movements in space and time, psychoanalysis -- on the other hand -- deals with psychical reality and not with material reality". In the end, the criterion is no longer the physically observable world, but the fact that it presents a coherence and a resistance comparable to that of material reality. Finally, (d) concerns the fact that the analytic situation selects from a subject's experience that which is capable of entering a story or narrative [whence the role of memory in psychoanalysis]. If there is truth in psychoanalysis, according to Ricoeur, it relates to psychoanalysis' ability to represent and explain such facts. Ricoeur goes through a long development (it's an interesting read), but histheoretical point (this is what we ought to keep in mind, much more so than the above) is that the degree of exactitude that is to be expected from psychoanalytic statements depends on the sort of truth that can be expected in this domain (in the representation or account of facts such as they have been described above, for instance). As Ricoeur writes, "for lack of an exact view of the qualitative diversity of types of truth in relation to types of facts, verificational criteria appropriate to the natural sciences, in which facts are empirically observable by one or more enquirers, have been repeatedly applied to psychoanalysis. The obvious conclusion has been that either that psychoanalysis does not in any way satisfy these criteria or that it satisfies them only if they are weakened. What is needed by the consideration of the types of facts psychoanalysis is concerned with is thus another form of "verification", one which Ricoeur goes on to spell out -- for instance, a distinction between "saying-true" and "being true". And in "saying-true" one finds the idea of constructing (or reconstructing) a coherent story or account from the "tattered remains of experience". (This implies understanding what makes a narration a good explanation -- in the psychoanalytic sense of the term).
Further Answer to Burnett
Analogous forms of argumentation are found in the later Wittgenstein. You witness a schoolboy raise his hand in class. You ask someone what just happened. The person starts explaining the physiological process that starts with the brain sending a message down the spinal cord and leads to muscle action. The whole theory of physiology can be laid out. But none of this will explain that the boy wanted to answer a question (or why he would want to do that) [by the way, must his motivation be reduced to the survival of the fittest in the classroom?] or get his teacher's attention since he needs to leave class to go to the washroom. The simple "fact" of arm raising is not so simple after all -- and many different facts may be involved.
I'll agree that much work (a tremendous amount in fact) is still ahead with regards to the various forms reality can take, though it seems obvious that reality goes beyond the single domain of the sensuous (perception and sensation should be distinguished, and one can be an empiricist without being a sensualist -- the failure to recognize this distinction is a major drawback of logical positivism).
Peirce (and Peircean pragmatists) agree that simplicity is better in the economy of research (in science). However, one should not refer to it as a way to block the road to inquiry. It may well be that in certain cases (if the nature of the fact actually overlaps) a cognitive account may prove to be a better account. But it seems to me that in most discussions in film studies the facts accounted by cognitivism and psychoanalysis don't actually overlap. (This is not to say that psychoanalytic film criticism is necessarily right -- not at all; nor that it is always done properly -- far from it). Moreover there is the entire issue of whether what most of us do in film studies can properly be called "science" if, as I put it in my talk, one has in mind the purpose and methods associated with the natural science. Wittgenstein, for one, saw Freud as a wrong headed scientist, but as an interesting philosopher -- one capable of explaining the human condition (or giving reasons) by way of apt images or metaphors. Think, for instance, of Thierry Kuntzel's famous study of the credit sequence of The Most Dangerous Game. Here, psychoanalysis helped Kuntzel "see" and "account" for aspects of the film that I believe are truly "there" to be seen in the film. It may be, however, that psychoanalysis is not the best account possible of these "qualities" -- though that remains to be seen.
I don't want to champion psychoanalysis (I'm happy to let the psychoanalysts deal with their epistemological problems), however I think we need to review the strategies used to pooh-pooh* it in film studies (indeed this is what Grunbaum did in reviewing Popper's argument against psychoanalysis).**
* According to Peirce: “The first order of induction, which I will call Rudimentary Induction,or the Pooh-pooh argument, proceeds from the premiss that the reasoner has no evidence of the existence of any fact of a given description and concludes that there never was, is not, and never will be any such thing.” And elsewhere: “A Pooh-pooh Argument is a method which consists in denying that a general kind of event ever will occur on the ground that it never has occurred. Its justification is that if it be persistently applied on every occasion, it must ultimately be corrected in case it should be wrong, and thus will ultimately reach the true conclusion.”
** Note that I will be away travelling as of Saturday and won't be able to answer the blog until my return in July.