Many readers of Deleuze project a false unity over his philosophical evolution, behaving as if he never changed his views. A particularly persistent perspective reads the whole work in terms of DIFFERENCE AND REPETITION and LOGIC OF SENSE, ignoring the places where Deleuze explicitly states that he had moved on from that problematic. His problematic changed over time, and so not only did his terminology evolve but also his concepts. Particular cases of this are the relative importance of the concepts of difference and of the eternal return in his later philosophy.
I am glad that Deleuze no longer talked in those terms. The Deleuzian version of the ethical imperative is no longer to live life so as to will its return an infinite number of times but rather to live life so as to construct the body without organs the riches in intensities and connexion. None of this mistaken claptrap about return.
I see no reason to insist on the Eternal Return as a central Deleuzian concept. Firstly, the term is virtually absent in the books after DIFFERENCE AND REPETITION and LOGIC OF SENSE.Secondly, in RHIZOME it is assimilated to the fascicular system aborting explicit unity but maintaining it at a higher level:
« Nietzsche’s aphorisms shatter the linear unity of knowledge, only to invoke the cyclic unity of the eternal return, present as the nonknown in thought » (ATP,6).
Deleuze’s system is constantly changing, and I do not think it correct to treat Deleuze’s later philosophy as a philosophy of difference (but rather of multiplicity) nor a philosophy of Return (but rather of consistence).
In the later work, all the mystagogical mantras about a « return » of difference have been jettisoned. Which is just as well as they were incoherent, and a falsification of Nietzsche. The terminology has changed, but so has the concept, and I think this is an improvement.
Changing the name of a concept is not an empty gesture but changes its dimensions as well the concept. The « return » is a bad name and DIFFERENCE AND REPETITION is full of inflated empty rhetoric to get across the notion of a « return of difference », an ultimately incoherent idea, and an unnecessary part of the concept of a deterritorialised time. So the usage is quite different. Both these concepts of difference and return came under attack from Badiou and Laruelle, and rightly so, but Deleuze had already moved on. It would be a rather strange affair if the philosopher of becoming and transformation always had the same ideas, and never changed.
Note: I am indebted to a conversation with Wayne Brooks for helping me clarify my ideas on this point.