Cherry-Picking Meaning

A sufficiently large body of writing can usually be made to support whatever point we had already decided to make.

A single sentence from a respected source — a scientist, a famous author, a founding document — is lifted out and set beside a modern concern, letting the authority of the source flow conveniently into the argument.

Sometimes this is thoughtful interpretation. Sometimes it is simply decoration.

The difficulty is that quotation creates an impression of fidelity. The words are real. The source is respected. The connection may feel profound. And none of that guarantees it is honest.

The selected sentence may represent only a small fragment of the original. Elsewhere, the same text may carry a different emphasis, an apparent contradiction or an inconvenient story that complicates the tidy conclusion.

The pattern appears everywhere.

Politicians quote economists. Managers quote company values. Lawyers quote clauses. Engineers quote documentation. Religious speakers quote scripture. Journalists quote interviewees. All can preserve the literal words while quietly mangling the meaning.

A quotation is a crop.

Cropping is necessary. We cannot reproduce every surrounding page each time we refer to an idea. The question is whether the crop preserves the scene or conceals it.

There are several ways context disappears.

The first is historical context. Words written for a particular audience, conflict or institution are treated as timeless instructions, as though nothing has changed since.

The second is textual context. A sentence is prised away from the argument around it. A warning becomes a promise. A question becomes a statement. An exception becomes a rule.

The third is authorial context. A person’s provisional thought is presented as their settled position — or an early opinion is quoted long after they publicly abandoned it.

The fourth is comparative context. One supporting source is displayed while equally authoritative sources pointing the other way are left discreetly in the drawer.

This last form is especially powerful because it is so hard to detect. The audience sees what was included, never what the speaker quietly stepped over.

A large corpus offers abundant opportunity. Search tools make it easier still. Begin with a conclusion, type a few keywords, and suitable authority can usually be produced before the kettle boils.

Artificial intelligence accelerates this considerably. It can locate relevant passages, generate thematic connections and produce a thoroughly convincing explanation of why the quotation matters.

That can support serious study. It can also industrialise confirmation bias.

At its worst, this is not interpretation but projection — reading one’s own conclusion into the text and then affecting surprise at finding it there.

The test is not merely whether the quote is accurate. It is whether the selection is fair.

Would a knowledgeable reader of the source recognise the interpretation? Does the surrounding material strengthen or weaken the point? Could an equally persuasive opposite article be assembled from the same body of work?

If so, perhaps the text does not settle the matter quite as decisively as the speaker implies.

There is also a standing temptation to borrow authority instead of making an argument. A quotation from scripture, science or a famous thinker can close a discussion without anyone doing the actual work of explanation.

But authority is not a substitute for reasoning. Even a revered source must still be interpreted, applied and reconciled with a reality that stubbornly refuses to read.

The fairer approach is usually to dial down the certainty of the claim.

Instead of “This text proves…”, perhaps:

• This passage suggests…
• One interpretation is…
• This theme appears here, though other passages complicate it.
• I find this a useful lens, rather than a final verdict.

That language sounds less powerful. It is also more honest.

The purpose of context is not to stop us drawing meaning from old words. It is to stop us turning those words into ventriloquist’s dummies for conclusions that were ours all along.

A quotation should illuminate an argument, not impersonate one.

• #thinking
• #context
• #interpretation

@online{byatt2026-cherrypickingmeaning,
  author  = {Stephen Byatt},
  title   = {Cherry-Picking Meaning},
  year    = {2026},
  date    = {2026-02-24},
  url     = {https://learningoutloud.stephenbyatt.com/blog/cherry-picking-meaning/},
  urldate = {2026-02-24},
  note    = {Learning Out Loud}
}