# A Summary of a Hedged Source Strips the Hedge ## Claim When a primary source states a claim and then immediately qualifies it, a downstream summary of that source tends to keep the claim and drop the qualification, converting a hedged observation into a clean assertion. <!-- atomicity: 1 idea ✓ --> ## Scope & conditions - Applies to any compression step — a recap, an abstract, a paraphrase, a second-hand citation — that has to select what to keep from a longer primary source. - Does not require bad faith; ordinary summarization pressure (word count, narrative cleanliness, marketing tone) is sufficient to produce the drop without any intent to mislead. - The failure is specifically asymmetric: hedges are more likely to be cut than the claims they qualify, because a hedge reads as a weakening of the point the summary exists to make. - Detecting this requires access to the primary source — the pattern is invisible from the summary alone, which is exactly what makes it durable. ## Evidence A webinar speaker states that deals with earlier buying-committee engagement closed in 30 days versus 60, then immediately adds, in the same breath, that the 30-day deals were also not competing against other vendors — explicitly calling the two deal types "totally different." A blog recap of the same webinar, published by the same organization, reports the 30-vs-60-day statistic as a clean payoff of early engagement and omits the competitive-context caveat entirely. Both documents exist in the same source library, making the drop directly verifiable by comparison rather than an inference about the recap author's intent. ## Links - [[3. Resources/Permanent Notes/A Trusted Intermediary Makes a False Claim Durable]] — extends: that note is about authority substituting for verification once a claim already travels through an intermediary; this note is about the specific compression step that produces the unverified claim in the first place. - [[3. Resources/Permanent Notes/An Absolute Claim Signals an Unexamined Failure Mode]] — contrasts: that note is about a source's own overclaiming; this note is about what a *later* summary does to a source that was actually careful. - [[3. Resources/MOCs/DevRel MOC]] — applies-in: when citing a stat that matters, locate and read the primary source rather than the recap, abstract, or press release derived from it. - [[3. Resources/Permanent Notes/Check a Shakeup Headline's Verb Against the Governance Action]] — the same compression measured across a thirteen-outlet cluster: hedges go first, then the verb itself hardens from "placed on leave" to "ousted". ## Tests / falsifiers - Across a sample of primary sources with recap or press coverage, does the recap's treatment of stated caveats predict how favorable the underlying framing is to the recap publisher? If caveats survive equally regardless of favorability, the mechanism is ordinary compression rather than motivated compression. - Would providing the recap author with a hard word-count-neutral incentive to preserve caveats change the outcome? If caveats still drop under no space pressure, the cause is narrative/marketing pull rather than compression itself. ## Maintenance - Last tended: 2026-08-16 - Next review: 2026-11-16 ## Source - [Accounts Don't Buy. People Do.](3.%20Resources/Readwise/Articles/Accounts%20Don't%20Buy.%20People%20Do..md) — the primary transcript containing the confound. - [[3. Resources/Readwise/Articles/How to Find the Real Buying Committee Behind Your Hot Accounts Common Room]] — the recap that omits it. - [[3. Resources/Literature Notes/Accounts Don't Buy. People Do. (full webinar)]] — Literature Note documenting the comparison.