This distinction between structural demarcation and a-structural exchange poses a series of problems for the notion of discussion. From the perspective of demarcation, not only is it impossible for the philosopher to converse meaningfully not only with the partisans of the doxa, but also with the scientist and the artist, nor even with philosophers working out of other problematics.
The incommensurability of the non-referential concept with the doxa, with the function, and with percepts and affects seems absolute. The same applies, internally to philosophy, to the discontinuity between different philosophical systems.
Deleuze and Guattari condemn « discussion » as a narcissistic and sterile activity, an empty social ritual to be avoided at all costs:
Every philosopher runs away when he or she hears someone say, “Let’s discuss this.” (WHAT IS PHILOSOPHY?, 28)
1) One should be wary of taking such statements at face value, as enouncing some sort of general law about an abstraction called « discussion », about which one can say something in isolation from any specific context or concrete arrangement. Deleuze and Guattari do not seem to be speaking concretely.
What Deleuze and Guattari getting at here is the existence of incommensurability, which is very often ignored, denied or repressed. Many people proceed in a discussion as if you were using the same words with the same meanings and go off on a tangent based on the use by their interlocutor of a few key words that they recognise and employ themselves. This is a bad way to handle incommensurability, by autistic denial.
2) Nor should one understand this statement of Deleuze and Guattari as propounding a general rule: always avoid discussions, never discuss.
Deleuze and Guattari reject discussion and oppose the solitude necessary for creating concepts to the sterility of gregarity:
Communication always comes too early or too late, and when it comes to creating, conversation is always superfluous (WIP?, 28).
Yet we feel that there is something one-sided about this proposition. All forms of dialogue seem to be equated, and equally excluded: discussion, communication, conversation are treated as synonyms. General rules are propounded: « flee discussion » and « conversation is always superfluous ».
This is very strange in a book stemming from the decades long collaboration (or creative dialogue) between two thinkers in whose professions dialogue plays an essential role: a philosopher (pedagogical and collegial dialogues) and a psychoanalyst (analytic session, therapeutic dialogue).
In contrast, exchange and conversations is held to be both possible and fruitful between the creative disciplines of philosophy, science, and art, by way of their accompanying a-structural « shadow ».
In an interview that Deleuze gave after the publication of WHAT IS PHILOSOPHY? he distinguishes between « conversation » and « discussion », and attributes a creative role to conversation:
Discussion amounts to wasting a lot of time on indeterminate problems. Conversation is something else entirely. Conversation is quite necessary. But the slightest conversation is a highly schizophrenic exercise happening between two individuals with common resources and a taste for ellipse and verbal shortcuts. Conversation is composed of immobility interspersed with long silences; it can give you ideas. But discussion has no place in philosophical work (TWO REGIMES OF MADNESS, 384, translation modified by me).
There can be no general rule, as most concrete cases, according to Deleuze and Guattari’s own principles, are mixtures. In this case, we are confronted with mixtures between « discussion » (bad) and « conversation » (good). The idea that as a general rule discussion or dialogue is narcissistic, whereas a monologue or a conversation between friends is not, is quite doubtful.
There is also the problem of the criterion. How are we too decide when the other is merely « discussing » and not « conversing »? Are we not in danger of a dogmatism that legitimates its refusal to listen or to be open to critique as a creative posture?
Further, are we not in danger of rationalising and universalising a characterological particularity? Deleuze was perhaps an aristocrat at heart, and may have had difficulties with discussion that were sociologically and psychologically based rather than being philosophically grounded.
If we apply the distinction (discussed here by Badiou) between the three types of logics (classical, intuitionist, and para-consistent), we may say that « discussion » is inherently unstable and unsatisfying. This is because it tends to oscillate between the classical logic (following the law of the excluded middle) of binary demarcations (called by Deleuze in DIALOGUES the « forced choice ») and the intuitionist logic (allowing for the existence of other values than true or false) of a plurality of opinions condemned to « tolerate » each other.
Conversation is « schizophrenic » because it adds to this mix a para-consistent logic (which allows for the existence of contradictions). Here conversation is ellipse, and an animated exchange is compatible with immobility and silence.
A good conversation thus combines the three logics, making sharp distinctions and clear demarcations, (provisionally, where necessary), and allowing for a multitude of nuances and intermediate positions, while remaining half-plunged in a chaos where creation and destruction are maintained as one.