Originally discussed March 10, 2012
In After Finitude, Meillassoux argues that the most basic characteristic of philosophy since Kant has been correlationism: the claim that to be is to be the correlation of a thinker and a thing. In the first chapter, he uses the aporia of the arche-fossil to begin driving a wedge between thought and appearance.
Primary and Secondary Qualities
Meillassoux says it is time to rehabilitate the distinction between primary and secondary qualities. Primary qualities are those “which are supposed to be inseparable from the object, properties which one supposes to belong to the thing even when I no longer apprehend it.” Secondary qualities are those qualities which exist only through sense perception. The pain I feel when I touch a flame does not exist in the flame itself – it is a secondary quality. The redness of a rose only exists because of the way my eyes and brain interpret the light, so it too is a secondary quality.
It is not that secondary qualities are injected by me into the object, like a constant hallucination. There is indeed a constant link between real things and their sensations. Even so, the sensible only exists as a relation: a relation between the world and the living creature I am. The secondary quality does not exist in the object, but it also does not exist in me – it exists purely as a relation.
There is a serious problem with the very distinction between primary and secondary qualities. It was acknowledged that one set of qualities depended on a relation with a subject – so why not all the qualities? While he was not the first to use the terms, the origin of the primary/secondary distinction lies in the 17th century French philosopher Rene Descartes. For Descates, the primary qualities were those of extension, qualities subject to geometric proof. Yet, we cannot imagine an uncoloured extension – extension is still tangled up with secondary qualities. This is one of the basic reasons why the distinction was dropped.
Meillassoux offers a new idea of primary qualities – that which is mathematizable. While the sensible – what we can see can touch – really does only exist as a relation between the subject and the world, “the mathematizable properties of the object are exempt from the constraint of such a relation, and that they are effectively in the object in the way in which I conceive them whether I am in relation with this object or not.”
This sounds like dogmatic metaphysics. It is the claim that thought is capable of discriminating between the properties that are a relation, and the properties that are in the world in itself. But such a thesis has been untenable since Kant and Berkley, because “thought cannot get outside itself in order to compare the world as it is “in itself” to the word as it is “for us,” and thereby distinguish what is a function of our relation to the world from what belongs to the world alone.” We cannot know anything about the world beyond our relation to it. Even mathematical properties “cannot be exempted from the subjectivation that is the precondition for secondary qualities: they too must be conceived as dependent upon the subject’s relation to the given…” – representation for the Kantian, an act of subjectivity for the phenomenologist, or a specific formal language for the analytic philosopher. Those that acknowledge the critical turn or the end of dogmatic metaphysics will say we cannot think something without abstracting from the fact that it is we who are thinking about something.
These considerations reveal the extent to which the central notion of modern philosophy since Kant seems to be that of correlation. Correlation is the idea according to which we only ever have access to the correlation between thinking and being, and never to either term considered apart from the other. Correlationism is any current of thought which maintains the unsurpassable character of the correlation so defined.
The correlationist position is that we cannot grasp subjects and objects independently of one another. It’s always about the relation, over and above the related. Before Kant, philosophy thought substance; after Kant, philosophy thought the correlation. The question is no longer, which is the proper substrate (God, atoms, ego) but which is the proper correlation (subject and object, or language and referents, etc).
In the 20th century, consciousness and language were the principle “media” of the correlation – consciousness for phenomenology, language for analytic philosophy. Consciousness and language make the world; “from their perspective ‘everything is inside’ but at the same time ‘everything is outside…'”
Another philosopher, Francis Wolf, explains it this way: “Everything is inside because in order to think anything whatsoever, it is necessary to ‘be able to be conscious of it’, it is necessary to say it, and so we are locked up in language or in consciousness without being able to get out.” But on the other hand, thought is always pointed to an outside-of-thought or is dependent on it: “To be conscious of the tree is to be conscious of the tree itself, and not the idea of the tree; to speak about the tree is not just to utter a word but to speak about the thing.” The consequence of this is that “We are in consciousness or language as in a transparent cage. Everything is outside, yet it is impossible to get out.”
An outside world that has no relation to us has been lost – all we can experience or know of the world is its correlation with our thought. The insistence on thought being oriented to the outside is “a bereavement – the denial of a loss concomitant with the abandonment of dogmatism.”
Ok, why break with correlationism? Science can produce statements about things before there was consciousness. We can date objects absolutely; we can measure their duration based on the constant rate of disintegration of radioactive nuclei. “Thus contemporary science is in a position to precisely determine – albeit in the form of revisable hypotheses – the dates of the formation of the fossils of creatures living prior to the emergence of the first hominids, the date of the accretion of the Earth, the date of the formation of stars, and even the ‘age’ of the universe itself.” The things that we use to make these statements – isotopes, fossils of single-celled creatures, star light, etc., are what Meillassoux calls arche-fossils. Arche-fossils are the material support (or justification, or evidence) for ancestral statements, statements about the world that existed before humans or any kind of a relation or the qualities that relation would contain.
What is the meaning or significance of these statements? Ancestral statements must be a discourse whose temporality recognizes thought only as one event among others, not an origin. In other words, science can tell us about things that existed before the correlation began, which means those things had qualities that existed part from our relation to them. How can a correlationist interpret these statements?
As an aside, a dogmatic philosophy such as Cartesianism has no problem with ancestral statements. While statements like “it was really hot” are inadmissible, the mathematical measurements are perfectly acceptable to a Cartesian. A Cartesian can speak of the primary qualities at the origin of the Earth, and will accept the dismissal of the secondary qualities. While scientific statements are always open to falsifiability and are revisable, they are capable of being literally true.
Philosophers have become very circumspect where science is concerned, claiming that their work in no way depends upon or contradicts the work of the scientist. A correlationist will even accept ancestral statements – unless the statements are meant literally, without qualification. Take this ancestral statement: “Event X occurred Y number of years before the emergence of humans.” A correlationist will not dispute this sentence at all, but they will add a silent addition: “. . . for humans.” The correlationist implies that the scientist is trapped in the ordinary attitude, and that the philosopher “possesses a specific type of knowledge which imposes a corruption upon science’s ancestral statements – a correction which seems to be minimal, but which suffices to introduce us to another dimension of thought in its relation to being.” For the correlationist, there are two levels of meaning to any ancestral statement – the immediate, realist meaning, and the deeper meaning, activated by the amendment “… for humans.”
So the correlationist interprets the ancestral statement this way: “The present community of scientists has objective reasons to consider that the accretion of the Earth preceded the emergence of hominids by x number of years.” Since Kant, “objectivity” is the universalizable – “The sun heats the stone” as opposed to the “subjective,” private, “the room is warm.” Intersubjectivity replaces adequation as the marker of objectivity. “Scientific truth is no longer what conforms to an in-itself supposedly indifferent to the way in which it is given to the subject, but rather what is susceptible of being given as shared by a scientific community.”
The phrase “the present community of scientists” is absolutely vital. It indicates we do not precede from the ancestral past but the correlationist present. The “deeper” meaning that only the philosopher has access to is that what is really real is our relation now to the arche-fossil and the statements about those fossils that may be shared by all, not the world that existed before the relation that the arche-fossil reveals.
A realist will dismiss the “for humans” amendment. The immediate, realist meaning is the only meaning. For the correlationist, this produces a list of absurdities. Meillassoux lists six absurdities, but we will only speak about three here. One such absurdity is “that what is preceded in time the manifestation of what is.” This is absurd because it is allegedly impossible to speak of something existing independently of the relation that contains all qualities. Another problem the purely realist meaning creates is that the relation itself emerged in space and time, and that we can date it – but space and time are modes of human perception, and so could not exist before the relation. Finally, it is impossible for the correlationist to accept “that the fossil-matter is the givenness in the present of a being that is anterior to givenness.”
To unpack that last sentence, “givenness” refers one particular version of the correlation; it is the way the world shows itself to us. A rock is given as existing in time and with certain qualities. We can rephrase that problem as saying it is ridiculous to think that the fossil is evidence in the present of a thing that existed in a time before any qualities could exist.
Yet, we don’t measure isotopes to test their universality, we measure them “with a view to external referents which endow these experiments with meaning.” Science is, at best, only secondarily concerned with the community of scientists. Experiments are meaningful because they deal solely with the world, not with any universal community of humans. Therefore, the “for humans” codicil does not deepen the sense of the statement, it cancels it. Either the ancestral statement is only realist, or it has no sense at all. The consequence is that the distinction between “civilized” transcendental idealism and silly subjective idealism – the line between Kant and Berkeley – dissolves “in light of the fossil-matter.” Every variety of correlationism is an extreme idealism.
The Paradox of the Arche-Fossil
Meillassoux raises some possible objections to his position, but we will only consider one here. In order to recover the ancestral fossil for correlationism, a counterfactual is all that is needed: if there had been an observer, the event would have appeared in such and such a manner. Meillassoux says that this entire response involves conflating two things: the ancestral and the ancient. The distant event is indeed like the ancient event. But the ancestral event is not an ancient event. The distant event and the ancient event both take place in a world with givenness. The distant event is “recuperable as an in-apparent given which does not endanger the logic of correlation.” However, the ancestral statement is not an absence in givenness – that is, something we as one half of the correlation do not perceive, like the ocean floor – but an absence of givenness. Givenness arose in a time and place before givenness. The relation that produces qualities arose in a time and place before the relation existed.
Therefore, no counterfactual can be used to recover the arche-fossil. “For the problem of the arche-fossil is not the empirical problem of the birth of living organisms, but the ontological problem of the coming into being of givenness as such.” Science thinks a time that is indifferent to givenness. The challenge is the following: to understand how science can think a world wherein spatio-temporal givenness itself came into being within a time and a space which preceded every variety of consciousness.” Science does think the time of the emergence of the transcendental, the conditions that govern the relation between the perceiver and the world.
It is important to note that Meillassoux is not attempting to refute the list of absurdities in paragraph 17. The arche-fossil is not a refutation of correlationism, but rather a paradox at the heart of it. Qualities do only exist as a relation between perceiver and world; yet we can make literally true statements about the world before the relation existed, and we know the relation began in time. The rest of the book is an elaboration of this apparent contradiction: what are the consequences of the fact that we do have knowledge of a world absolutely separate from us, despite the fact that we are wholly limited to our own thought?