Most legal AI systems operate on a shared assumption: that first-contact information arrives organized enough to reason from. It does not. At first contact, participants communicate through partial narratives, unresolved timelines, secondhand facts, and incomplete documentation.

In almost every system, that ambiguity is compressed into a summary and discarded. The uncertainty, the competing plausible readings, the unresolved intent present at that moment — all of it disappears. What survives is a conclusion.

Once that compression occurs, later reasoning inherits the conclusion rather than the conditions under which it emerged.

The Antecedent Methodology is built around a different premise. The ambiguity present at first contact is not noise to be resolved. It is part of the record. Preserving the conditions under which interpretation forms — not just the interpretation itself — is the work.

Most intake systems preserve conclusions.
Antecedent preserves the conditions under which those conclusions emerged.