"Far better an approximate answer to the right question, which is often vague, than the exact answer to the wrong question, which can always be made precise"
About this Quote
Tukey is taking a scalpel to a habit scientists and institutions quietly reward: mistaking precision for truth. His line lands because it reframes “vagueness” not as intellectual sloppiness, but as an honest property of the world. Real problems arrive messy, underspecified, entangled with human goals and constraints. You can’t always formalize them cleanly at the start, and pretending you can is often the first error.
The jab is aimed at a certain kind of technical virtue-signaling: the exact answer that flatters the method more than it serves the question. “Which can always be made precise” is the tell. Tukey implies that the wrong question is seductive precisely because it’s easy to operationalize. You can define it, measure it, optimize it, publish it. Precision becomes a bureaucratic shield: if the math is airtight, who can argue? But airtight math doesn’t rescue a mis-aimed inquiry; it just makes the mis-aim harder to notice.
Context matters: Tukey helped invent modern data analysis at a moment when computation and statistical machinery were accelerating. His ethos wasn’t anti-math; it was anti-ritual. He’s defending exploratory work, iteration, and pragmatic approximation as intellectually serious, because they keep you tethered to what you actually want to know. The subtext is almost moral: the duty of a scientist isn’t to produce immaculate answers, but to prevent elegant irrelevance. In today’s metrics-driven culture, it reads like a warning label for dashboards, A/B tests, and model scores that optimize the measurable while quietly abandoning the meaningful.
The jab is aimed at a certain kind of technical virtue-signaling: the exact answer that flatters the method more than it serves the question. “Which can always be made precise” is the tell. Tukey implies that the wrong question is seductive precisely because it’s easy to operationalize. You can define it, measure it, optimize it, publish it. Precision becomes a bureaucratic shield: if the math is airtight, who can argue? But airtight math doesn’t rescue a mis-aimed inquiry; it just makes the mis-aim harder to notice.
Context matters: Tukey helped invent modern data analysis at a moment when computation and statistical machinery were accelerating. His ethos wasn’t anti-math; it was anti-ritual. He’s defending exploratory work, iteration, and pragmatic approximation as intellectually serious, because they keep you tethered to what you actually want to know. The subtext is almost moral: the duty of a scientist isn’t to produce immaculate answers, but to prevent elegant irrelevance. In today’s metrics-driven culture, it reads like a warning label for dashboards, A/B tests, and model scores that optimize the measurable while quietly abandoning the meaningful.
Quote Details
| Topic | Reason & Logic |
|---|---|
| Source | Unverified source: The Future of Data Analysis (John Tukey, 1962)
Evidence: Page 13 (printed page number); also appears as p. 62 in JSTOR/PDF pagination, in section '11. Facing uncertainty.'. Primary source is John W. Tukey’s own article: 'The Future of Data Analysis,' The Annals of Mathematical Statistics, Vol. 33, No. 1 (March 1962), pp. 1–67. The quote appears in sect... Other candidates (2) Deep Learning to See (Alessandro Betti, Marco Gori, Stefano..., 2022) compilation98.2% ... John Tukey , who made outstanding contributions on the Fourier series ... Far better an approximate answer to the... John Tukey (John Tukey) compilation95.2% an statistician sourced far better an approximate answer to the right question which is often vague than an exact ans... |
More Quotes by John
Add to List






