Skip to main content

0018 · A relation the message puts in the past

Memory namespace · version 18 · source

A sixth reason: the message puts the relation in the past​

I used to live in Amman. asserts the relation. It is positive, it is not hedged, it is not reported, and the polarity refusal added in 0011 lets it straight through. The quote is verbatim, the span reproduces it, the relation is in the vocabulary — and what is stored is a fact claiming the person lives in Amman now.

Why this is the worse of the two presences​

A claim the message did not assert is a presence rather than an absence, which is what made it worse than the three reasons before it. This one is worse again: a denial that slips through is a fact about something the person went out of their way to say was untrue, and it tends to read oddly to whoever sees it. A past tense that slips through reads perfectly. It is the answer to "where does this person live", it is cited, it is current, and the only thing wrong with it is a date nobody recorded.

Why refusing rather than storing an interval​

The obvious alternative is to store it with a closed validity: the relation ended at the message. That half is knowable. The other half is not — the sentence says nothing about when it began, and valid cannot be left unbounded below, because every read is a containment test. An unbounded start answers "did they live there in 1990" with yes, on evidence nobody has, and a fabricated start is a date this system invented and would then cite.

So the claim is kept here with its words and its own reason. Nothing false is stored, the rate is countable, and the sentences are the corpus a later decision would be made against.

What this deliberately does not do​

Using the past tense to CLOSE a fact already held. If memory holds an open lives_in(user, Amman), this sentence is testimony that it ended and the honest response is to close the interval rather than to refuse. That needs an evidence model for an ending — the receipt comes from a different message than the one that opened the fact — and a read that survives a second evidence row, which the current inner join does not. It is held separately rather than assumed here.

What it runs​

  • ALTER TABLE rejected_claim DROP CONSTRAINT rejected_claim_reason_chk
  • ALTER TABLE rejected_claim ADD CONSTRAINT rejected_claim_reason_chk CHECK
  • COMMENT ON CONSTRAINT rejected_claim_reason_chk ON rejected_claim

SQL​

{schema} is replaced with the validated namespace name when the migration is applied.

49 lines
-- ── A sixth reason: the message puts the relation in the past ─────────────────────────────────
--
--  `I used to live in Amman.` asserts the relation. It is positive, it is not hedged, it is not
--  reported, and the polarity refusal added in 0011 lets it straight through. The quote is verbatim,
--  the span reproduces it, the relation is in the vocabulary — and what is stored is a fact claiming
--  the person lives in Amman now.
--
--  ── WHY THIS IS THE WORSE OF THE TWO PRESENCES ───────────────────────────────────────────────
--
--  A claim the message did not assert is a presence rather than an absence, which is what made it
--  worse than the three reasons before it. This one is worse again: a denial that slips through is a
--  fact about something the person went out of their way to say was untrue, and it tends to read
--  oddly to whoever sees it. A past tense that slips through reads perfectly. It is the answer to
--  "where does this person live", it is cited, it is current, and the only thing wrong with it is a
--  date nobody recorded.
--
--  ── WHY REFUSING RATHER THAN STORING AN INTERVAL ─────────────────────────────────────────────
--
--  The obvious alternative is to store it with a closed validity: the relation ended at the message.
--  That half is knowable. The other half is not — the sentence says nothing about when it began, and
--  `valid` cannot be left unbounded below, because every read is a containment test. An unbounded
--  start answers "did they live there in 1990" with yes, on evidence nobody has, and a fabricated
--  start is a date this system invented and would then cite.
--
--  So the claim is kept here with its words and its own reason. Nothing false is stored, the rate is
--  countable, and the sentences are the corpus a later decision would be made against.
--
--  ── WHAT THIS DELIBERATELY DOES NOT DO ───────────────────────────────────────────────────────
--
--  Using the past tense to CLOSE a fact already held. If memory holds an open `lives_in(user,
--  Amman)`, this sentence is testimony that it ended and the honest response is to close the interval
--  rather than to refuse. That needs an evidence model for an ending — the receipt comes from a
--  different message than the one that opened the fact — and a read that survives a second evidence
--  row, which the current inner join does not. It is held separately rather than assumed here.

ALTER TABLE {schema}.rejected_claim
    DROP CONSTRAINT rejected_claim_reason_chk;

ALTER TABLE {schema}.rejected_claim
    ADD CONSTRAINT rejected_claim_reason_chk
    CHECK (reason = ANY (ARRAY['unmapped_relation', 'unlocatable_quote',
                               'duplicate_claim', 'not_asserted', 'unresolvable_subject',
                               'not_current']));

COMMENT ON CONSTRAINT rejected_claim_reason_chk ON {schema}.rejected_claim IS
    'Six reasons a proposal did not become a fact, each meaning something different: the vocabulary '
    'cannot represent it, the message cannot cite it, this message already said it, the message did '
    'not assert it, it is about nothing the message establishes, or the message puts it in the past. '
    'The last two are the ones whose alternative was a false fact rather than a missing one.';