We briefly referred on the episode to the fact that, as for Marx, for Lacan, all ostensibly theoretical talk is really tainted in some way. Whereas for Marx, we're really just repeating, or perhaps reacting to in some more complicated way, the ideology of those in power. Lacan, following Freud, looks for a psychological explanation, for an underlying meaning or meaning structure that is in some way responsible for what we're really saying, whether we know it or not.
Fink deals with this in Ch. 9 of his book "The Four Discourses." These are:
1. The Master's Discourse. This is discourse ruled by the master signifier, which has no literal meaning. From p. 131:
The master must be obeyed--not because we'll all be better off that way or for some other such rationale--but because he or she says so. NO justification is given for his or her power: it just is... The master must show no weakness, and therefore carefully hides the fact that he or she, like everyone else, is a being of language and has succumbed to symbolic castration: the split between conscious and unconscious brought on by the signifier is veiled in the master's discourse...
2. The University Discourse. Where the master doesn't care about knowledge, this does; it provides "a sort of legitimation or rationalization of the master's will." He doesn't actually dismiss what he considers genuine science here; that would be in category 3. The bulk of academia will go here, though, and just as for Marx, be used to support the political power structure despite any pretense at objectivity.
3. The Hysteric's Discourse. From p. 133-4:
The hysteric goas at the master and demands that he or she show his or her stuff, prove his or her mettle by producing something serious by way of knowledge. The hysteric's discourse is the exact opposite of the university discourse... a hysteric gets off on knowledge... Lacan finally identifies the discourse of science with that of hysteria.... The hysteric pushes the master--incarnated in a partner, teacher, or whomever--to the point where he or she can find the master's knowledge lacking...
4. The Analyst's Discourse. From p. 135:
Object (a), as cause of desire, is the agent here... The analyst plays the part of... pure desiring subject, and interrogates the subject in his or her division, precisely at those points where the split between the conscious and unconscious shows through... As it appears concretely in the analytic situation, a master signifier presents itself as a dead end, a stopping point, a term, word, or phrase that puts an end to association, that grinds the patient's discourse to a halt... it could be... a reference to the death of a loved one, the name of a disease, or a variety of other things... While the analyst adopts the analytic discourse, the analysand is... backed into the hysteric's discourse... The analyst, by pointing to the fact that the analysand is not the master of his or her own discourse, instantes the analysand as divided between conscious speaking subject and some other (subject) speaking at the same time through the same mouthpiece... Clearly the motor force of the process is object (a)--the analalyst operating as pure desirousness.
In the descriptions above I've left out more than I've included, and particularly with #4, it's difficult to understand what kind of discourse it is without having in mind Lacan's view of how the subject arises through alienation. Basically, there are four elements involved (each of which corresponds to one of the above, in the order I've given them): The master signifier (language itself as an agent, as our "self"), the "other" (also, of course, a product of language), the unconscious, and the intervention of an analyst or other individual. One of these being dominant makes the type of discourse what it is. One of the points in laying out these types is to say that there is no "metadiscourse" that escapes the limitations of the four listed. From p. 137: "Psychoanalysis' claim to fame does not reside in providing an archimedian point outside of discourse, but simply in elucidating the structure of discourse itself."
We didn't bother to go into this much on the podcast. I find it not particularly convincing. Whatever the psychological motives behind an utterance, we can't reduce the meaning to those motives. We can still try to evaluate the content itself, and if somehow our judgment is undermined by psychological and social factors, well, that's something we'll have to look out for, but the possibility is not a cause for global skepticism.
For another, probably clearer account of the four discourses, read the Wikipedia entry.