← Research directorate

PROGRAM 02 / ESTABLISHED 2008

PARALLAX

Perceptual integrity

Agreement is not evidence of an intact reality.

Distinguishing what a system observes from what its instruments have learned to agree upon.

01 / Approach

PARALLAX constructs observation networks with deliberately dissimilar failure modes. Witnesses retain the sequence of how an observation became credible, including intervals in which evidence was unavailable. Contradiction is preserved as a measurable condition rather than averaged out.

02 / Selected finding

The P-09 chamber trials detected a common-mode omission in all six observers by comparing their timing residues. Every individual feed appeared internally complete.

03 / Evidence boundary

A detected gap does not disclose its contents. Reconstructing an event from the shape of its absence remains an inference and must be labeled accordingly.

04 / Ongoing inquiry

The current instrument series tests witnesses that cannot exchange raw observations. Their disagreement is the only signal they are allowed to share.

Associated records.

D-0 / PUBLIC EDITIONS
PX–FN–023Field note

The dissenting witness

Five instruments agreed on an empty chamber. The sixth reported that it was unable to observe.

2026.09.02PARALLAX
PX–TB–026Technical brief

The observer gap

Six instruments returned a complete account. The interval between them did not.

2026.08.21PARALLAX
PX–FN–019Field note

On absent evidence

Why a missing observation must remain distinct from an observation of nothing.

2025.11.17PARALLAX