The Foucauldian genealogy is an unmasking of power for the use of those who suffer it. It is also directed against those who would seize power in their name. As Francois Ewald pints out, there are three, not two, parties to every power struggle: not only those who exercise power and those who would exercise it in their place, but also those on whom it is exercised. Because one speaks against power, one does not necessarily speak with those who suffer it. Hence Foucault's concerted attack on all forms of interpretation and representation: on the use made of the linguistics of Saussure and Jakobson, on psychoanalysis, on Marxism. For the interpreter, things are never what they seem. People never say what they mean or mean what they say; they never know what they want or what they are doing.
For Foucault interpretation is reduction, repression, obliteration of fact, discourse, and desire. It is a technique of knowledge; it is also a technique of power. Interpretation requires specially qualified interpreters, representatives. The dialectic is an interpreter's weapon for the seizure of power.
But Foucault shows that truth does not exist outside power, still less in opposition to it. Each society has its own regime of truth: the types of discourse accepted as true, the mechanisms that make it possible to distinguish between truth and error. ... The will to the power of truth is a pitiless tyrant; it requires a singular and total devotion. It is a service that has tempted the European mind since Plato. Nietzsche gave the first signs of its possible end; he also provided a way out, which is called genealogy. Genealogy was a 'grey' activity, but it was also a gay science, a science of the hypothetical. That gaiety, that love of hypothesis, pervades all Foucault's work. He is the reverse of a guru, a teacher, a subject who is supposed to know, though he would, in all modesty, be flattered if, without excessive seriousness, he were compared to a Zen master, who knows nothing. For him uncertainlty causes no anguish:...
A love of hypothesis, of invention, is unashamedly, a love of the beautiful. What drew Foucault to the case of Pierre Riviere was not the mass of official documentation, but the beauty of Riviere's memoir, a beauty that shamed the dreary prose of the educated experts who busied themselves around him. It was a daring, provocative remark, suggesting that beauty of expression in an indication that what is being said is worth listening to. The question of Foucault's own style is not insignificant. It is not so much that Foucault writes well _...It is rather that he writes with ostentatious brilliance: his writing betrays a quite shameless delight in its own skill that calls to mind the sumptuous prose of our own pre-classical period....To write in this way is no affectation or self-indulgence. It is, if it requires justification, functional. Like all style, it is both natural and cultivated:a natural mode of expression for a writer striving to renew contact with a pre-rational world of communicating Reason and Folly and a conscious rejection of the language of Reason that seeks by its grey, measured, monotonous tones to give an impression of authority, objectivity and truth. (221-4)
All of the above is a quote from Alan Sheridan's Foucault The Will To Truth (p. 219-224)
For a genealogy of the movie Eclipse see here.