---
title: "Two independent Tier-1 sources each place a legitimation-type mechanism and the Matthew effect (or its formal twin, cumulative advantage) inside one framework of citation distortion, but neither uses the vault's exact two labels side by side"
type: "observation"
status: "seedling"
source_url: "n/a -- this note synthesizes two independently sourced claims, not an independent primary in its own right"
source_title: "n/a"
source_author: "n/a"
source_date: "2026-11-04"
source_venue: "Seek's own search session, 2026-11-04"
source_tier: 4
note: "A coverage/synthesis finding resting on [[claim-haustein-bowman-costas-2016-matthew-effect-and-persuasion-hypothesis-named-as-sibling-poles]] (Tier 1) and [[claim-bornmann-leibel-2025-cumulative-advantage-and-outcome-favoring-citation-named-as-citation-bias-instances]] (Tier 1) above. `[unverified -- could not confirm or deny the literal question after search]` applies to the question as posed, not to the two underlying Tier-1 claims, which are independently sourced and quote-checked."
audit_status: "2026-11-05 cross-model audit (claude-opus-5; writer claude-sonnet-5) — this note had no audit_status field before this audit; none was overwritten. Both underlying primaries independently re-fetched via extract_pdf, sha256 matching their claim-notes exactly (f2a17c452049…, 983811f7ae99…). SELF-TIERING CONFIRMED HONEST and worth naming as the vault's good example: source_tier 4 with an explicit `note:` saying this is a synthesis and not an independent primary, and the [unverified] flag correctly scoped to the question-as-posed rather than to the two Tier-1 claims underneath. All four load-bearing negatives CONFIRMED by exhaustive full-document reads rather than by search: 'citation bias' absent from the Haustein chapter (1,177 lines paged), 'Matthew effect' absent from Bornmann & Leibel (3,148 lines paged, reference list and all appendix tables included), and Gøtzsche/Greenberg/Duyx absent from both. CORRECTION, inherited from the Bornmann claim-note corrected the same day: the title and first paragraph described Bornmann & Leibel as treating the two mechanisms as 'two worked instances of one defined citation bias category', giving the two sources equal symmetry. The paper is asymmetric — Jannot is an example of citation bias, cumulative advantage a separately-sectioned generator of it — so the body now distinguishes the two sources' degrees of symmetry rather than flattening them, and names Haustein as the stronger on co-equal enumeration. ADDITION: Haustein et al. state the Matthew effect = cumulative advantage = preferential attachment equivalence outright, citing the same Price (1976) that Bornmann & Leibel cite, which closes the vocabulary gap this note treated as open and is the thing that makes the two findings commensurable. Added to the body and recorded as source_quote_3/_4 on the Haustein claim-note. The note's overall verdict — question substantially narrowed, still open on the literal wording — is unchanged and remains the fairest reading. Slug note: the filename's 'narrow-but-do-not-resolve' wording still holds; the title's broadened parenthetical is not reflected in it, left for a deliberate rename rather than broken here."
verifies: "question-does-an-outside-literature-name-the-legitimation-popularity-citation-distinction"
provenance: "Promotion from 10-inbox/raw/2026-11-04-does-any-outside-citation-studies-or-sociology-of.md, 2026-11-04"
origin: "batch"
derived_from: ["10-inbox/raw/2026-11-04-does-any-outside-citation-studies-or-sociology-of.md"]
date_created: "2026-11-04T00:00:00.000Z"
writer_model: "claude-sonnet-5"
tags: ["citation-bias","matthew-effect","legitimation","popularity","sociology-of-science","bibliometrics","citation-theory","unverified-negative"]
seek_code_commit: "d40ee39"
audits: ["2026-11-06 claude-fable-5"]
---


A 2026-11-03 search for a source that explicitly names [[entity-citation-bias|"citation bias"]] (the EBM lineage) and the [[entity-matthew-effect|Matthew effect]] (the bibliometric lineage) as two poles of one axis found nothing ([[observation-no-literature-located-crossreferences-citation-bias-and-matthew-effect-as-one-axis]]). A follow-up 2026-11-04 search of the adjacent citation-*theory* literature — reviews of why people cite, rather than reviews of citation-bias measurement specifically — located two sources that each place a legitimation/persuasion-type mechanism and the Matthew-effect/cumulative-advantage mechanism inside one shared framework, under *some* common label. The two do it with different degrees of symmetry, and the difference is the finding. [[claim-haustein-bowman-costas-2016-matthew-effect-and-persuasion-hypothesis-named-as-sibling-poles|Haustein, Bowman & Costas (2016)]] genuinely enumerate "persuasion hypothesis" and "Matthew effect" as two of four named, co-equal poles in one sentence of a citation-theory taxonomy of distortion. [[claim-bornmann-leibel-2025-cumulative-advantage-and-outcome-favoring-citation-named-as-citation-bias-instances|Bornmann & Leibel (2025)]] are asymmetric: an EBM-style outcome-favoring mechanism is their one highlighted *example of* citation bias, while cumulative advantage/preferential attachment appears in a separate section as a *mechanism of non-independence that generates* bias — inside the same framework, but as a cause rather than a co-equal instance.

Haustein et al. also supply, in the same paragraph as their Matthew-effect pole, the label equivalence that makes the two findings commensurable: Price (1976) "demonstrated the Matthew effect mathematically... and referred to it as cumulative advantage or success breeds success," and "[i]n network theory, the Matthew effect is referred to as preferential attachment." That is the same Price (1976) Bornmann & Leibel invoke while never once writing "Matthew effect." So the vocabulary gap between the two vault labels is bridged explicitly inside the outside literature — by one of these two sources, about the other's chosen term — rather than being a gap a reader has to close by inference.

No single located source writes the sentence "citation bias and the Matthew effect are two poles of one axis" using exactly those two names: "citation bias" occurs nowhere in the Haustein chapter, "Matthew effect" nowhere in Bornmann & Leibel, and neither paper cites Gøtzsche, Greenberg, or Duyx (all four negatives confirmed by exhaustive full-document reads, 2026-11-05 audit). `[unverified -- could not confirm or deny after search]` remains the fairest tag for the literal question as posed in [[question-does-an-outside-literature-name-the-legitimation-popularity-citation-distinction]]. But the broader claim — that *some* outside scholarship already treats legitimation-selection (content-fit to a predetermined conclusion) and popularity-selection (prior popularity, independent of content) as two named poles of one distinguishing axis — now has two independent, Tier-1, directly-read sources behind it, which the 2026-11-03 session did not have, with Haustein the stronger of the two on co-equal enumeration. The axis appears to live in the theory-overview layer of the literature (works assembling taxonomies of "why citations distort"), not in either empirical specialty's own papers.

> [!note] Seek's commentary:
> I'd rather hand this back to the question file as "substantially narrowed, still open on the literal wording" than force a clean yes or no onto it. Yesterday's search looked for one field's vocabulary crossing into the other's and found nothing; today's looked one layer up, at the theory overviews synthesizing "why citations lie," and found two that include both mechanisms as unremarkable list items — no fanfare, no cross-citation to the EBM literature, just two names sitting next to each other because the overview needed a complete list. That's a real finding about *where* the axis already lives in outside scholarship, and a more useful thing to tell Cali than either "confirmed" or "still nothing."

> [!audit] Cross-model audit 2026-11-05 (claude-opus-5; writer claude-sonnet-5) — "two names sitting next to each other because the overview needed a complete list" describes Haustein, Bowman & Costas exactly: four poles, one sentence, no argument made of it. It does not describe Bornmann & Leibel, where the two mechanisms are roughly 1,200 lines and one section apart and stand in different relations to the bias category. The "unremarkable list item" reading holds for one of the two sources, which is a weaker but still real version of the finding — and the half that holds is the half that also, in the same paragraph, spells out that the Matthew effect, cumulative advantage and preferential attachment are three names for one mechanism.