How a record is sourced

The rule

No claim about a living person enters the record without a source — and for the highest-risk records, sources are demanded at the door rather than at review.

Reporting a problem law is one of those. So is flagging a challenger. Both are negative public claims about real people, and enforcing citations only at approval time leaves uncited claims sitting in an inbox where they can be read, forwarded, and eventually mistaken for findings.

What counts

Something a stranger who distrusts us can go and check:

• a roll call
• a committee calendar or a rule
• the filed text of a bill
• a statute citation
• a floor transcript
• an agency letter

What does not count

• another page on this site
• a summary nobody linked
• "it is well known"
• an assertion by somebody we like

Provenance

For anything downloaded, there is a provenance record: what was fetched, from where, and when. Voting records, district maps, photos and transcripts all carry them.

So a disagreement about our data is settled by looking at what was actually pulled, instead of becoming an argument about whose memory of it is right.

State the act, not the character

A record says what a person did: held the bill at a named stage, on a named date, by a named mechanism. Not what they are, not what they wanted, not who they were working for.

Motive is an argument. Procedure is a record. The movement publishes the second, and the discipline is what makes the first unnecessary — a documented pattern of acts speaks without anybody having to characterise anybody.

Gaps are published

Sometimes a bill is known to have been held and not reliably known by whom. That is recorded as a gap — died at this stage, unattributed — rather than assigned to whoever is most convenient.

A named gap is a real finding, and it is an invitation: it tells a citizen with access to an archive exactly what would close it.