Corpus II · US Reflection

There are two ways of being present to a thing. One can alter it, or leave it whole. This article isolates the second and gives it a name — the witness — then draws from it a criterion that rests on no prior definition: a witness is that whose withdrawal changes nothing in what it notes.

I. Two presences

A man crouches at the edge of a river and blows on the water. The surface ripples, wavers, then settles again. Remove the man: the wind would have rippled the water the same way, and it would have settled just so. His passage left nothing that the course of things would not have produced without him. The same man, an hour later, dips his hand into the river and lifts out a stone to carry away. Remove him this time, and the stone is still there. Two gestures, two presences: one withdraws leaving no trace, the other leaves a void.

This difference does not depend on the force of the act, nor on the intention of the one who acts. It rests on a single verifiable property: would the state of things be changed in the absence of the one who is present? If the answer is no, the one who was there was a witness. If the answer is yes, he was an actor — whatever name he gives himself.

The word « witness » is taken here in its literal sense, not in the weakened judicial one where to witness means to recount. The witness does not yet recount: he notes what is. He is present to what is, and his presence does not weigh on what is. It is a rare position, and the greater part of what follows consists in establishing its coherence ¹.

II. To graze is not to touch

It will be objected that every presence touches, and that to touch is already to act. But the objection commits a slip of qualification, and must be refused before it is answered. To touch and to act are not synonyms. Touching implies physical contact; acting implies no contact at all — one acts by a word, by a decision, at a distance. To say « touching is already acting » is therefore to take one verb for another. Seeing the pollen is, in that loose sense, to act; yet it touches nothing. The objection collapses not because we reply to it, but because it confused two verbs ².

What remains true is that presence often does make contact — and here a distinction the French language carries in a single word must be reconstructed in English. Between the contact that takes and the contact that leaves, we have to touch and to graze — to touch lightly, as the grazing animal crops the grass without tearing the root.

A dry finger settles on a flower, grazes the pollen, then withdraws. A few grains have clung to it; one smells them, one sets them back. The flower receives and loses pollen constantly — that is its function; the dry finger blends into that course. One has grazed. Let the same finger carry a little saliva, and the pollen fixes to it, leaves, does not return: one has taken. The gesture is identical; a single drop decides everything. This is the finest edge of the distinction: the verb is not read in the movement, but in what the movement retains.

The witness grazes the world and leaves it whole; the actor touches it and carries a piece away.

This line runs even through the senses. Sight, hearing, smell receive: light comes to the eye, sound to the ear, scent to the nose, and the thing stays where it is, whole. Taste does not: to taste requires that a part of the thing enter and be consumed. The nose leaves the pollen; the mouth takes it. A sense that seems light — taste — falls in truth on the side of taking, while the others remain on the side of grazing. The border is never the organ: it is what, of the contact, remains or disappears.

From this an unexpected consequence: to touch is not to undo. The bat strikes a thousand walls with its cry each second — for sound is indeed contact, a pressure wave that strikes and returns — and undoes no cave. Real contact alters nothing, provided it withdraws. We thought one had to avoid touching in order to undo nothing; it is enough to retain nothing.

III. To note is not to recognize