Repetition as Failure

The next section begins with a summary of the previous one:

The discrete, the alienated and the repressed are the three cases of natural blockage, corresponding respectively to nominal concepts, concepts of nature and concepts of freedom. In all these cases, however, conceptual identity or Sameness of representation is invoked to account for repetition: repetition is attributed to elements which are really distinct but nevertheless share strictly the same concept.

and later:

In all these cases, that which repeats does so only by dint of not 'comprehending', not remembering, not knowing or not being conscious. Throughout, the inadequacy of concepts and of their representative concomitants (memory and self-consciousness, remembrance and recognition) is supposed to account for repetition. Such is therefore the default of every argument grounded in the form of identity in the concept: these arguments give us only a nominal definition and a negative explanation of repetition.

In other words, since we take the perfect mirrored correspondence of one distinct thing to one distinct concept as the natural and default state of affairs, every repetition indicates some sort of error or failure or lack in our concepts. The explanation for repetition is that something got fucked up and is abnormal.  

A critique of this "negative" way of explaining stuff runs through all of Deleuze's philosophy, and in my mind actually constitutes one of its deepest lessons.  Maybe one of the easiest ways to see the problem with this way of framing explanations is to consider the way we use the term insane to purportedly explain something.  As in, "he heard the voice of his dog telling him to kill people because he was insane".  This explanation is 99% moral judgement (which might explain your reaction to something but its not going to help at all in explaining the phenomenon in question) and 1% I don't know.  It basically takes a habitual or statistical definition of sane and normal for granted, observes that this case is different, attaches some moral stigma to it, and passes this whole works off as an explanation of what happened.  In fact, you might even go so far as to say that the goal of this type of explanation is precisely to get you to stop asking why something happened and instead to focus on thinking about how bad it was that it did.  

In contrast, a real explanation focuses on the way in which the effect was actually produced.  Because the Son of Sam really did hear the voice of his dog.  Something actually happened there.  It certainly may be out of the ordinary.  Even something we want to discourage in polite society.  But that doesn't make it less real.  And the more you try to search for a genuine explanation of how the voices could be produced, the more you realize that you're not even sure how "normal" voices are produced.  You mostly just take it for granted that the ones you hear in your head somehow represent the "real" ones out there in the world.  Calling the dog's voice hallucination is a non-explanation of how it's different from something you don't understand to begin with!  In fact, one of the very best tools in all of neuroscience has always been to watch what happens to an abnormal brain as a means of inferring what must be going on under the hood in a normal one.  Drugs and insanity are a great window onto those inner workings, but so are crazy lesions like Phineas Gage suffered.  In all these cases, the question should be the same: how did this happen?  No one would even consider it an explanation at all if you said that Gage's personality changed because his accident "drove him insane".

Real explanations try to trace the chain of causal events that produced an effect.  They are positive in the sense that they focus on a ground up explanation of how something came to be, rather than on how it differs from a preconceived normal state of affairs that we take for granted we already understand.  Sometimes Deleuze will even call these explanations or definitions "genetic" (as opposed to "nominal") because they focus on how something we actually generated, rather than just on how we recognize it and give it a name.  It might be easily mis-leading to call these explanations "mechanistic" but that's essentially how I think of them.  Not mechanistic in the sense of "literally reduced to a series of colliding billiard balls" but more as in "possessing some mechanism that doesn't take the outcome for granted but explains how it was produced".  Maybe the better terminology is Nozick's idea of "invisible hand explanations"? I'll have to re-read that bit to make sure it works here.

That seems to be the direction Deleuze shifts in with the remainder of the introduction.  Forget about the way it theoretically seems like repetition should be impossible (or could only be understood as generality) in a law-like world that we are able to fully represent.  Let's look for a positive mechanism by which repetition could actually be produced.  That would constitute an explanation of repetition.