# Cherry-Picking The rhetorical move of selectively presenting only the evidence, data, examples, or quotations that support a desired conclusion while ignoring (or actively suppressing) the contradicting evidence — producing a presentation that is *locally accurate* (every individual citation is real and verifiable) but *globally misleading* (the overall picture conveyed is distorted, sometimes inverted, relative to the underlying record). The fallacy's distinctive analytical feature is that it cannot usually be exposed by checking individual claims: every cherry-picked point holds up on its own. The deception is in the *omission*. ## Etymology and lineage The image is cherry-orchard harvesting: picking only the ripe red cherries, ignoring the unripe ones, and presenting a basket that looks more uniformly excellent than the orchard actually produces. The phrase enters general American English by the mid-20th-c.; the formal-logic version of the concept is older and known under several names: *suppressed evidence* (Aristotle, [[John Stuart Mill]], contemporary informal-logic textbooks), *fallacy of incomplete evidence*, *one-sided assessment*, *card stacking* (Institute for Propaganda Analysis's 1937 *seven devices of propaganda* — see [[The Big Lie]] for the related IPA framework). The contemporary technical sense covers a family of related operations: - **Quotation cherry-picking** — citing only the part of a source that supports the writer's claim, omitting the surrounding qualifications that change the meaning - **Data cherry-picking** — selecting a subset of observations, time periods, or measurement points that produce a desired pattern while excluding the rest - **Source cherry-picking** — citing only the studies, experts, or outlets that agree with the desired conclusion while ignoring contrary published work - **Example cherry-picking** — presenting one or a handful of vivid cases as if they were representative when the broader pattern is otherwise ## Why cherry-picking is the disinformation-vocabulary heavyweight In any given week of contemporary American political and media discourse, cherry-picking is statistically the *most common* disinformation technique deployed — substantially more so than outright lying, fabrication, or [[The Big Lie]]-style audacious falsehoods. The reasons are structural: 1. **It is harder to refute than fabrication** — refuting a lie requires showing the lie. Refuting a cherry-pick requires assembling the *omitted* evidence, presenting it, and persuading the audience that what was omitted matters 2. **It carries plausible deniability** — the speaker can always say "I cited X, X is real, you can verify it." The audience that does verify finds X confirmed and may not investigate further 3. **It survives fact-checking** — most fact-checking infrastructure is built to evaluate *claims*, not the *completeness* of an argument. PolitiFact and FactCheck.org both publish methodology notes acknowledging this limitation 4. **It is asymmetrically expensive to detect** — producing a cherry-picked argument is cheap (you ignore inconvenient evidence); producing the complete picture for comparison is expensive (you have to do the underlying research) The asymmetry is what makes cherry-picking so structurally favorable to bad-faith argument in high-volume political-discourse environments. ## Canonical case categories A non-exhaustive partial catalog: - **Climate-denial cherry-picks** — picking a single starting year (1998's El Niño peak) to claim "no warming since [date]"; citing a handful of skeptical scientists from a literature in which 97%+ of climate publications agree on anthropogenic warming; quoting selected paleo-temperature records to argue "it was warmer in the medieval period." [[Naomi Oreskes]] and Erik Conway's *Merchants of Doubt* (2010) documents the playbook in detail - **Crime-statistics cherry-picks** — picking the one demographic-and-jurisdiction combination that supports a desired narrative while ignoring national patterns; choosing the year that produces the most dramatic increase or decrease - **Quotation cherry-picks** — the famous Lincoln quote-mining ("you can fool some of the people all of the time"), the Jefferson "tree of liberty" quote stripped of its surrounding context, the Founders' Federalist-Papers cherry-picks deployed in contemporary policy debate - **Scriptural cherry-picks** — selecting verses that support a political position while ignoring contradicting passages from the same text; the [[Biblical Worldview]] tradition's selective-emphasis pattern - **Economic-statistics cherry-picks** — picking the start date for a stock-market or wage-growth comparison that produces the desired political read; quoting one BLS series while ignoring the matched corrective - **Polling cherry-picks** — citing the one outlier poll that supports the claim while ignoring the average; choosing the question wording or subgroup that produces the most favorable result - **Studies cherry-picks** — vaccine-denial "but this one study said" citations of single discredited papers; ivermectin and hydroxychloroquine COVID-era arguments built around individual small-trial results in defiance of meta-analytic consensus ## The detection-and-rebuttal asymmetry The empirical regularity of high-volume political-discourse environments is that cherry-picking is *easier to deploy than to refute* — which is why fact-checking organizations have evolved a distinct response category for cherry-picked claims (PolitiFact uses *Mostly False* for claims that are "technically accurate but [...] takes things out of context" — that's the cherry-picking judgment). The category sits between *True* and *False* because the underlying citations are real; what is misleading is the omission. Newer audience-side responses to the asymmetry include: - **Pre-bunking** ([[Prebunking]]) — teach audiences to ask *what is being omitted?* rather than only *is what is being said true?* - **Network-context tools** — Wikipedia's *Citation Needed* / *Better Citation* tagging, journalistic *full context* boxes alongside political claims, X / Bluesky community-notes (Ratio) - **Aggregator + comparison tools** — climate-data dashboards that let users see the full data series rather than the chosen-window slice; election-data aggregators that show the polling-average rather than the cherry-picked outlier - **Metaliteracy education** — formal media-literacy curricula that explicitly teach the cherry-picking pattern alongside other classic fallacies These help at the margin; the structural asymmetry has not been eliminated. ## Cherry-picking vs. emphasis: where the line is Every communicator has to choose what to include and what to leave out — selective emphasis is intrinsic to communication. Three rough markers distinguish ordinary editorial selection from analytical cherry-picking: 1. **Magnitude of the omission** — was the omitted evidence *more important* than what was cited? An honest summary may omit secondary details; cherry-picking omits the central rebutting evidence 2. **Direction of the omission** — were the omissions systematically on one side? Honest selection often leaves out items on both sides; cherry-picking shows directional bias 3. **Acknowledgment of the omission** — did the speaker note that other evidence exists and explain the choice to focus on the cited subset? Honest emphasis is transparent; cherry-picking presents the selection as if it were the full picture The markers are heuristic, not bright-line; in practice judgment is required. The analytical interest of *cherry-picking* as a term is that it provides a vocabulary for naming the systematic-and-undeclared form of the move. ## See Also - [[The Big Lie]] — the audacity-and-repetition cousin (different mechanism, same delegitimizing goal) - [[Alternative Facts]] — the perspectival-cover companion - [[False Equivalence]] — the bothsidesism kin - [[Climate Change Denialism]] — canonical operating environment for cherry-picking - [[Confirmation Bias]] — the cognitive substrate that makes cherry-picked arguments persuasive - [[Prebunking]] — the audience-side inoculation response - [[Disinformation Hub]] — Subject hub - [[Disinformation Lexicon]] — alphabetical MOC