The Archive as Form

The archive is not a window onto the past. It is a machine, and what it was built to do decides what can be seen through it.

I spent years in the state archives of Tunis and Nantes, and the romance of the place is real: vast, dust-smelling, badly catalogued, holding the traces of millions of people, and you never know what a day’s call slips will bring up. I have pulled a Jewish man’s claim for damages on a radio the French army requisitioned during the war; reports on Arab children in Sfax and Sousse questioned for handing out Italian Fascist pamphlets they could not read; a clerk’s note on Maltese merchants hawking light machinery in the street beside Bab el Khadra; the correspondence of men who found that the only safe ground from which to oppose French rule was Geneva, or Istanbul, or Berlin. Each one comes up like a face turned out of the soil.

And not one of them carries any meaning on its own. Each was collected by someone, for a purpose (a police officer’s procès-verbal, a security report, a diplomatic cable, a clipping filed by an information officer), and nearly everything I read reached me already filtered through the eye of the state. The state was a colonial one, and a colonial file is a record of what the administration needed to be true. The archive not preserves not the past but the part of the past that served the people keeping the archive.

So the meaning is not in the fragments. It is in the form someone gives them: the order, the selection, the connective tissue that turns a box of unrelated documents into a story with a shape. That assembly is an act of creation, and it was the part of the work the discipline could least admit to, because history—as discipline—wanted to be a science of the found rather than an art of the made. We buried the most imaginative thing we did under an apparatus designed to make it look inevitable.

Which leaves the question I have not worked out: If the meaning is made and not found, and if everything in the archive was written by the people who won, then what can never be seen from inside it? The subaltern leaves no file of her own; she surfaces, when she surfaces at all, as the object of someone else’s report. You can read the colonial archive against its grain, but you are still reading the colonial archive, and the grain is most of what there is.

I used to think the answer was better records, more complete ones. I don’t anymore. A perfect archive of a violent state is a perfect record of its violence, not a passage around it. The fragments will always need a hand to give them a form, and the form will always carry the hand. What I am left with is less a method than a suspicion: that the honest historian and the honest novelist are doing far more of the same work than either trade will say out loud.