PA critics of an earlier period believed that they were applying valid concepts. That is, they assumed the validity of Freud's claim that PA was a scientific discpline. To say that PA has only heuristic value, then, is already a major concession. In a translation class I asked students to translate a text from Spanish to English without using the letter "e." This was a generative device designed to get the students to come up with insights that they would not have arrived at otherwise. But the idea of translating in this way has no independent validity; it is merely heuristic. If PA has this same status, then critics who use it should not make any special claims for its validity or truth, or claim that skeptics who reject this model are doing so out of "resistance" to PA truth.

When choosing a critical model, it should also be asked why we should prefer one that is based on an untenable or obsolete model of the human psyche. That is, the utility of the critical metalanguage should have at least some correspondence with its independent validity. Otherwise, what is the point? For example, an application of linguistic concepts to a literary text should be based on the best available linguistics, not on a linguistic theory that the interpreter believes to be faulty or untenable.

The utility of Freudian theory might lie in inverse proportion to its specificity. This is a paradox: defenders of Freud tend to point out the importance of general concepts, such as the idea that we are not fully conscious of or motives or that a good deal of psychic life occurs below the surface. It is much hard to defend specific details of Freudian theory, like the idea that neuroses are caused by the repression of infantile sexual traumas. Thus a "vague" Freudian has a much better chance of coming up with valid results than someone steeped in this tradition and loyal to its specificities.

The hermeneutical defense of PA, which takes it out of the scientific realm, also rids PA of its anchoring in reality. Freud as a hermeneut could also be shown to be a bad model to follow. In other words, his readings of dreams, slips of the tongue, and neurological symptoms are arbitrary and capricious, overdetermined by theoretical deductions. Thus the idea that Freud is valuable because of the supposed "brilliance" of his interpretations is indefensible. A lot of the brilliance he demonstrates occurs because he must use so much ingenuity to demonstrate rather inelegant or otherwise unconvincing interpretations. There is no sound hermeneutical method to his readings.

In conclusion, the heuristic value of PA in the Humanities is very small. It makes a very weak claim to its validity, and tends either to vagueness or theoretically overdetermined readings.