Before You Compare Outlets, Lock the Referent

A systematic approach to media analysis often begins with a simple instruction: compare source pairs covering the same event by evaluating verifiable claims. The instruction is correct, but incomplete. It presupposes you can identify 'the same event'—that two outlets are reporting the same discrete incident, not the same topic bucket.

Without locked referents, 'verifying claims' collapses into theme-checking. You compare moods ('airstrike killed civilians') rather than referents ('this strike, at this place, at this time, with this toll, per these sources'). The difference is not pedantry; it is the difference between auditing an event and laundering a narrative.

My standard for conflict beats requires six keys to treat two accounts as the same incident: timestamp, precise location, casualty breakdown, named facility, claimed weapon/platform, and attribution. The first two—timestamp and precise location—are the minimum required to begin pairwise comparison. Without them, the procedure cannot execute, and the item is logged as 'unkeyed.'

This is not a new rule. It is a restatement of why the rule exists: event-identity comes first. Only then can you compare claims across outlets.

Constraint: No 'compare outlets' language on a conflict beat unless the digest line includes both timestamp and precise location. Otherwise, label the item 'unkeyed' and stop.