Filing correspondence apart from evidence

Most reporters start an inquiry with one folder. It seems sensible at the time — everything related to the story goes in the same place, in the order it arrives. An email from a source sits next to a leaked memo, which sits next to a note about a phone call, which sits next to the actual document that started the whole thing. For the first few weeks this works fine, because you can still hold the whole shape of it in your head.

The trouble starts around week six or seven, when the folder has grown past what any one person can remember, and two very different kinds of material have become tangled together. One kind is evidence: documents, records, data, anything that supports or contradicts a claim you might publish. The other is correspondence: the conversations, requests, and exchanges that got you the evidence in the first place. They serve different purposes, and treating them as one pile makes both harder to use.

Why the two keep getting mixed up

Correspondence often contains fragments of evidence — a source might paste a figure into an email, or describe a document instead of sending it. It feels wasteful to separate the email from the fact it contains. But an email is a record of a conversation with a particular person at a particular time, carrying obligations around confidentiality and consent. A piece of evidence, once verified, stands on its own and can usually be referenced without exposing how it was obtained. Filing them together means every time you reach for the evidence, you also touch the correspondence, and every time you review the correspondence, you risk treating unverified fragments as settled fact.

"If you can't explain to a colleague, in one sentence, whether something is a source conversation or a piece of evidence, it's filed in the wrong place."

There's also a practical risk. Correspondence sometimes needs to be withheld entirely — to protect a source's identity, or because a promise of confidentiality was made. If it's threaded through the evidence file, redacting it later means combing through everything a second time, usually under time pressure, right before publication.

A structure that holds from week one

The fix doesn't need to be complicated. Two top-level divisions — one for evidence, one for correspondence — with a simple cross-reference system between them, is usually enough. When a source mentions a document in an email, the email stays in correspondence and a note is added pointing to where the document itself lives in evidence, once it's obtained and verified. The two never need to occupy the same file.

A few habits make this easier to sustain over months rather than weeks:

• Decide the split before the first document arrives, not after the folder is already full
• Treat anything unverified as correspondence until it clears verification, even if it looks like evidence
• Keep a short cross-reference note rather than duplicating material across both sides

None of this requires special software or a formal records system — a clearly labelled set of folders and a habit of asking "which side does this belong to" is usually sufficient. What it does require is deciding on the split early, before the volume of material makes any reorganisation feel like starting over.