Changelog · how the belief was versioned

Recency Illusion

The recency illusion is the belief that a usage, word, or phenomenon you have only recently noticed is therefore itself recent, when in fact it is long-established.

# README · how it works

Zwicky attributes the recency illusion to selective attention: once a person first consciously notices a form, they have no synoptic record of its prior distribution across time, place, and social group, so they mistake the moment of their own noticing for the moment of the form's emergence. Educated speakers who trust 'engines of correctness' (usage guides, school grammar) are described as especially prone to it, because they treat their own subjective impressions as fair samples and fail to consult historical sources. Popular accounts of the closely related frequency illusion add confirmation bias as a reinforcing mechanism, but Zwicky's original framing for the recency illusion is specifically selective attention.

v0.1Arnold M. Zwicky · 2005initial release

+First shipped: Just between Dr. Language and I (Language Log). Coined by Stanford linguist Arnold Zwicky in an August 7, 2005 Language Log post, the recency illusion is "the belief that things YOU have noticed only recently are in fact recent," which Zwicky attributed to a selective-attention effect. It is a folk/coined concept from linguistics rather than an experimentally established psychological effect: Zwicky illustrated it with usages (e.g. "between you and I," singular they, the "going-to-go" construction) that long predate the dates critics assign them. The only quantitative test to date, van der Meulen's 2022 corpus study of Dutch prescriptive writing, found that the great majority of checkable recency claims in usage guides were indeed inaccurate ("Recency Illusions"), though recency statements themselves were relatively rare.

v1.1Marten van der Meulen · 202287% · +5678

+van der Meulen (2022) — Dutch prescriptivism corpus study. 238 entries contained at least one recency statement, targeting 341 usage items (67.6% lexis, 28.8% grammar); recency statements were comparatively rare and declined over the 20th century. Of the checkable lexical recency statements examined, the large majority were Recency Illusions (17 of 23 lexical statements were illusions; reported as 20 of 23 ≈ 87% in the lexis analysis), and checkable grammatical recency statements were 'almost all' illusions. For 17 statements the earliest attestation predated the recency claim by more than 50 years.added support · Normative Database of Dutch: 5,678 prescriptive entries (1910–2018); 238 entries / 341 usage items carried recency statements; 23 lexical + 13 grammatical recency statements were checkable against usage data.

deprecated · community-coined

Release stability — folk-concept. The recency illusion is a concept coined in descriptive linguistics, not a quantitatively established psychological effect with a replication record. No controlled or pre-registered psychology experiments establish or test it as an isolated cognitive bias. The single quantitative investigation (van der Meulen 2022, an observational corpus audit of Dutch usage guides) supports the claim that recency statements in prescriptive writing are usually inaccurate, but also qualifies it: such statements are relatively rare and partly reflect genre/rhetorical conventions rather than pure false belief. Grounding therefore rests on documented historical attestations plus one corpus study, not on experimental psychology.

Issues filed in production

#001 open · 2005
"Between you and I" branded a ~20-year-old error
The case that prompted Zwicky to name the illusion: a yourdictionary.com 'Dr. Language' column (echoing James Cochrane's book 'Between You and I') claimed the construction had emerged only in the last ~20 years and could be 'nipped in the bud.' Zwicky showed via the OED and Merriam-Webster's Dictionary of English Usage that 'between you and I' goes back at least 150–160 years in Late Modern English, with earlier uses ~400 years old.
#002 open · 2005
Hockett's "recent colloquial pattern" GoToGo construction
Zwicky notes that even a professional linguist fell for it: Charles Hockett, in 'A Course in Modern Linguistics' (1958, p. 428), described 'I'm going home and eat' (the GoToGo construction) as a 'recent colloquial pattern,' yet David Denison had collected examples from roughly 30 years earlier, showing the construction was not new in 1958.
#003 open · 2022
Dutch usage guides: 'Amsterdam en haar grachten' deemed a modern trend
In van der Meulen's corpus, journalist Heldring complained in 1993 that neuter place-names were increasingly referred to with feminine possessives (e.g. 'Amsterdam en haar grachten' — 'Amsterdam and her canals'), treating it as a recent development; the identical complaint was made by Van Hoogstraten in 1700, and the practice is documented in prescriptive works back to at least the 1930s — a textbook Recency Illusion spanning centuries.

Recommended migration

Replace impression with attestation: before asserting that something is recent, check a dated historical source (dictionary with first-citation dates, a time-stamped usage corpus, archives). Zwicky's own remedy is blunt — 'you have to check the facts. Again and again ... the phenomena turn out to have been around ... for very much longer than you think.' Treat your memory of first noticing a thing as evidence about your attention, not about the world's timeline.

In van der Meulen's 2022 audit, roughly 87% of checkable lexical recency claims in Dutch usage guides were false once dated against corpora, and for 17 statements the form was attested more than 50 years before the 'it's recent' claim — i.e. the fact-check almost always overturned the impression.

Catch it in the act

Watch for any claim of the form "this word/usage/trend is new" or "people only started doing this in the last X years," especially when it is based on personal impression and aimed at something the speaker dislikes ("kids these days"). The tell is an unchecked dating: the speaker reports when THEY first noticed it, not when it is actually attested. Reach for a historical source (OED, a usage corpus, Google Books) and you will usually find the form predates the claim by decades or centuries.

Dependencies

Frequency IllusionsiblingIntroduced by Zwicky in the very same 2005 post: once you notice something you think it happens 'all the time.' Popularly called the Baader–Meinhof phenomenon. Recency and frequency illusions often co-occur and share a selective-attention basis; the frequency illusion (not the recency illusion) entered the OED in March 2019.
Baader–Meinhof phenomenoneasily-confusedA popular name (coined 1994 by Terry Mullen in a St. Paul Pioneer Press letter) for the frequency illusion, i.e. seeing a newly-noticed thing everywhere. Frequently confused with the recency illusion, but it concerns perceived frequency, not perceived newness.
Adolescent (Age) IllusionsiblingThe third member of Zwicky's 'trifecta of illusions': the belief that an innovation belongs to the young / current adolescents. Like the recency illusion, Zwicky roots it in taking one's own experience as the measure of all things. Note: the Adolescent Illusion was introduced in a separate post dated August 17, 2005 ('More illusions,' http://itre.cis.upenn.edu/myl/languagelog/archives/002407.html), ten days after the August 7 post that named the Recency and Frequency Illusions.
Confirmation biasmechanistically-linkedPopular explanations of the related frequency illusion invoke confirmation bias (preferentially registering confirming instances). It is a candidate reinforcing mechanism, though Zwicky's original account of the recency illusion specifies selective attention rather than confirmation bias.
Declinism / 'golden-age' (kids-these-days) biassiblingThe recency illusion typically serves a 'kids these days' narrative of language decay; Zwicky explicitly counters that the maligned forms are not new ('It's not just Kids These Days').

Open the same case on your own draft.