Method

Where the phrase comes from

In December 2016, before a change of administration in the United States, a group at the University of Toronto spent a day copying federal environmental data onto servers the government did not control. They expected pages about climate science to stop being served. The day was run with the Environmental Data Governance Initiative and the Technoscience Research Unit, much of what they copied went to the Internet Archive, and similar events followed across the United States over the next year. The organisers called the work guerrilla archiving.

Community memory projects now use the phrase for work that is older than the term, which is gathering and keeping what official archives were never going to gather, without waiting for a budget or a mandate. In both cases the reason to act quickly is the same. Material that nobody copies this week may not be there next week.

Why this archive is one

A placard is cardboard and a marker pen. It is carried for an afternoon and then it is thrown away, and graffiti lasts until somebody paints the wall. Album 011 is about writing that went up around Jantar Mantar and was gone inside a day. The photographs in it are the reason those words can still be read.

The official record of a protest holds none of this. It is a crowd estimate and a case number. What the protest actually said was handwritten by somebody who went home afterwards and threw the cardboard away.

So the collecting is done by the people who were standing in the crowd rather than by anyone sent to do it. Photographs arrive through a form on a phone, often the same evening. Nobody commissioned them, and no institution decided in advance that they were worth keeping. They stay the property of the people who took them, and this archive holds them rather than owns them.

Where this differs from the original

The phrase covers a range of practice. Three differences matter here.

• Copies in more than one place are what make a record hard to erase, and that is the part of the practice people usually mean by the name. Every published photograph is in version control beside the code that arranges it, with its history, so the archive exists on the web host and again in the repository. One command brings down every photograph and the record of the decisions behind them. A copy at the Internet Archive is the next one, which is where the 2016 event sent what it rescued.
• The work is slower than the name suggests. Faces of people in the crowd are covered before anything is published, a curator approves every photograph, and a removal request is honoured without argument. A fast sweep that asked nobody would collect more and protect people less. The guidelines set out how all of that works.
• Toronto was copying material that already existed and was about to be taken away. Almost nothing here had been recorded by anybody before it was photographed, which puts this closer to community documentation than to a rescue.

What happens to a photograph

• Somebody at a protest photographs a placard and sends it through the form, naming the movement it belongs to and saying how they want to be credited.
• It goes into a review queue. Nothing publishes on arrival.
• Faces in the crowd are covered. Faces printed on the placards stay visible, because those are the artwork and covering them would destroy the thing being kept.
• A curator approves it, rejects it, or sends it back when the covering landed in the wrong place.
• What is approved gets a number, a page of its own and a place in a movement album. What is rejected never enters the record.
• Anyone can ask for a photograph to come down, for any reason or none. The form for that asks for no explanation.

What the archive keeps about contributors, and for how long, is set out on What we keep.

Add to it

This works only if the people who were there send what they photographed. If you have pictures of placards, banners, graffiti or anything else people made and carried, they belong here more than they belong in a camera roll.