The Folder That Stops You Losing the Thread

There's a familiar moment, usually somewhere around week six of an investigation, where a reporter looks at their own notes and realises they can no longer say with confidence what they've actually confirmed. Interviews, documents, half-checked leads, and stray observations have all merged into one undifferentiated pile. Nothing is wrong with any individual piece — the problem is that the pile has stopped being legible as a whole.

The fix isn't more research. It's a small amount of structure applied to what already exists. The simplest version of this is a status marker attached to every substantive claim or document in the file: established, likely, unconfirmed, or contradicted. Four words, consistently applied, do more for a stalled investigation than another week of digging.

Why four categories and not more

It's tempting to build an elaborate tagging system, but the value of this approach comes from how quickly it can be scanned. A folder with dozens of nuanced status labels takes as long to interpret as the disorganised pile it replaced. Four categories are enough to separate what you could publish tomorrow from what still needs one more phone call.

• Established — verified by at least two independent sources or a primary document.
• Likely — supported but resting on a single source or an inference.
• Unconfirmed — noted, but not yet checked against anything.
• Contradicted — conflicting accounts that need resolving before use.

Making it a habit, not a project

The convention only works if it's applied as you go, not retrofitted once a month. Five minutes at the end of each research session, spent labelling whatever was added that day, is enough to keep the whole file honest. The alternative — a single exhausting weekend of retroactive tagging — tends not to survive contact with a deadline.

What this buys you, beyond tidiness, is the ability to see your own argument clearly. A file where everything reads as "established" is usually one where the status markers haven't been applied honestly. A file with a visible mix of categories is one you can trust, because it shows its own uncertainty rather than hiding it.