LogicalMANIPULATIVE
Argument from Silence
What it is
Concluding that something did not happen, or that someone does not hold a view, because a source that might have mentioned it does not — treating the absence of a record as a record of absence without showing that the record would exist if the thing were true.
How it works
Real-world examples
- •Wood (1995) argued that Marco Polo never reached China because his account omits the Great Wall, tea, and foot-binding; critics replied that the Wall as we know it is largely later Ming construction and that a Venetian merchant might not record what his audience would not value — the silence was never shown to be expected.
- •In “Silver Blaze” (Doyle 1892), Holmes reasons from the dog that did not bark to an intruder the dog knew: the legitimate form, because the expected-record condition holds.
- •In 2012 Donald Trump publicly offered a donation if Barack Obama released his college transcripts, implying that the unreleased records contained something damaging; from 2016 opponents of Trump drew the same inference from his unreleased tax returns. Each treats a withheld document as evidence of its contents.
- •A small clinical study that finds no effect is cited as proof that there is none; Altman and Bland (1995) pointed out that absence of evidence is not evidence of absence when the study lacked the power to detect the effect.
- •A reference letter that omits any mention of teamwork is read as a coded warning; it is one only if the writer would have mentioned teamwork had it been good, a condition that varies by writer, country, and template.
Ethical guidelines
- ●Before arguing from silence, state why the source would have recorded the thing if it were so; if you cannot, the silence is not evidence.
- ●Do not treat “has not denied it” as an admission; people decline to answer for many reasons, and the burden of proof does not move because a question was asked.
- ●When you cite a negative result, report the power of the search: how hard, where, and with what sensitivity someone looked.
- ●Prefer the honest phrasing — “we found no record” — over “there was none.”
How to defend against it
- ►Ask the expected-record question: “If this had happened, why would this particular source have mentioned it?” Silence is evidence in exact proportion to how confidently that question can be answered.
- ►Ask how hard anyone looked; a search, study, or archive that was never designed to capture X cannot testify to its absence.
- ►Separate absence of evidence from evidence of absence out loud, and ask for a positive reason to believe the negative claim — a better-placed source that does record things of this kind and omits this one.
- ►When “he has never denied it” is offered, name the move: an unanswered accusation is still an accusation, and the burden of proof stays where it started.
- ►Look for the barking-dog conditions — a source with the motive, the position, and the habit of recording such things. If all three are present, take the silence seriously; if not, do not.
From the Defense Playbook
Every playbook entry states how strong its evidence is and when not to use it. Browse the full playbook.
References
- Fischer, D. H. (1970). Historians' Fallacies: Toward a Logic of Historical Thought. Harper & RowThe fallacy of negative proof in historical argument.
- Walton, D. (1996). Arguments from Ignorance. Pennsylvania State University PressThe treatment of argument from silence as a species of argument from ignorance and the expected-record condition for its legitimate use.
- Altman, D. G., & Bland, J. M. (1995). Absence of evidence is not evidence of absence. BMJ, 311(7003), 485The statistical version: an underpowered null result is not proof of no effect.
- Doyle, A. C. (1892). The Adventure of Silver Blaze. The Strand Magazine; collected in The Memoirs of Sherlock Holmes (1894)The “dog that did not bark” as the legitimate form of the inference.
- Wood, F. (1995). Did Marco Polo Go to China?. Secker & WarburgThe Marco Polo argument from silence used as the historical example.
Last reviewed
Suggest a correction