It’s brave of Meillassoux to begin After Finitude with the argument from the archefossil, because it’s such a terrible argument. Indeed, Meillassoux admits that it is a terrible argument, which the correlationist will have no trouble dispatching; the reason for this, though, is that the discussion of the archefossil isn’t actually supposed to be an argument at all. When I first heard of it, it seemed to be a strange updating of Johnson refuting Berkeley by kicking a stone, with the curious addition of a complicatedly constructed hypothetical stone. But that’s not really how the discussion of the archefossil is supposed to work: the archefossil isn’t supposed to present an example of brute reality and thereby disprove idealism. It is presented and refuted as such during the course of the first chapter, but this argument is really preparing the ground for the real use of the archefossil, which is not to prove something about reality, but rather to raise a question about the relationship between thought and reality.
The point of the example of the archefossil is to “raise the question of the emergence of thinking bodies in time,” which is also “the question of the temporality of the conditions of instantiation, and hence of the taking place of the transcendental as such” (25). The archefossil is an example of ancestrality, but the real problem of ancestrality is, if space and time depend on thinking beings, how would we understand the fact that thought first arose at a specific space and time? However, this is only a question at all if there was indeed a moment at which thought first arose, that is, if thought and being are fundamentally distinct. Meillassoux accepts this inasmuch as he points out that ancestrality is only a problem for the transcendental idealist, not the speculative idealist; but he does not consider that this ancestrality is, for the same reasons, not a problem for the materialist, either.
The problem of ancestrality, then, depends on an implicit Cartesianism that runs through After Finitude, an assumption that thought and being are two different things, and that the connection or not between them is thus a problem. For a position like the one Marx sketches out in The German Ideology, in which thought is simple a particular set of relations among being, the problem of how thought relates to being is no problem; their interrelation is essential to understanding thought in the first place: “The production of ideas, of conceptions, of consciousness, is…directly interwoven with the material activity and the material intercourse of men, the language of real life.” What’s distinctive about Marx’s position is that it sees thought as active, rather than merely reflective or contemplative. Marx charges Feuerbach with failing to really break with idealism because he still conceives the material world as something that is passively reflected in though: “Feuerbach’s conception of the sensuous world is confined on the one hand to mere contemplation of it, and on the other to mere feeling.”
This description also applies to Meillassoux, as we can see from the solution he proposes to the problem of ancestrality. Meillassoux argues that science shows us a way of thinking the unthought: mathematization. But it’s interesting what he thinks it is about mathematization that renders it independent of humans: mathematics is independent, specifically, of human sense-experience. Mathemetization provides “a world capable of autonomy” because mathematized bodies “can be described independently of their sensible qualities, such as flavor, smell, heat, etc.” (115, my emphasis). What is distinctively (and, for Meillassoux, limitingly) human is sensuous contemplation; to put it another way, Meillassoux considers the human relation to the world solely epistemologically, in terms of the kind of passive experience that Heidegger calls “present-at-hand.” Both the problem of and Meillassoux’s solution to correlationism depend on this cartesian assumption that reifies humanity, as thought, as something specific and separate from being. If one doesn’t make this assumption, is correlationism an issue at all?