Draft the integration-response section under the reactivation reframing

Part of #15

Blocked by #17.

## Question

Draft (a cheap artifact to react to) the results-section narrative + claim for the integration response under the **context-gated perceptual reactivation** reframing. Cover:

1. **Reactivation of the perceptual representation** (not acoustic) — established by the effect emerging on fixed-acoustics within-completion trials.
2. **Word-end specificity ⇒ interaction**, not simple delayed reactivation — the re-expression is gated by the specific subsequent sounds (perceptual state × completion).
3. **Non-exhaustive** — a complementary subset conditions on acoustic context instead.
4. Retain the **negative claims** already supported (not lexicality / mismatch / surprisal; not extended acoustic-or-perceptual response to the word-initial sound).
5. Reconcile with CLAUDE.md's "reactivation tentatively disfavored / transfer bimodality" — this is a *positive* late-reactivation finding, of the perceptual (not acoustic) code.

**Corrected 2026-07-21:** this is NOT takeable in parallel with #16/#17. The effort is currently in exploratory data analysis (waiting on #16's prod run); there is no writing to draft yet, and pre-drafting risks anchoring the narrative on a framing #17 might reject. This ticket only opens once #17 lands a GO.

Do not draft yet: this issue is blocked by #17 and awaits #16's production run, with work opening only after a GO from #17. Once unblocked, read CLAUDE.md and the resulting exploratory analysis, then draft the results-section narrative and claim covering the listed reactivation, interaction, non-exhaustive, negative, and perceptual-code points.

Written by the indexing model from the issue text.