Borrowed Certainty

What it is

The category exists because two nearer ones did not fit. Source-class inflation describes what class of thing a citation is. Citation asymmetry describes which side of a live dispute a citation sits on. Both need citations to be wrong about, and in the pure case of borrowed certainty there are none. The defect lives in the prose rather than in the notes.

The second half of the definition is what makes it a borrowing rather than merely a gap. A claim with no source and no owner is a folk observation, and a reader meets it as one. A claim with no source whose owner is identifiable and published is a different object: the finding has arrived with its provenance stripped off. That is one of the ways a result becomes a thing everyone knows and stops being checkable.

Say plainly what it is not. A practitioner writing from experience is doing something legitimate and often more useful than the research, and a book that declared itself as such would be doing nothing wrong. The defect is experience and other people's research wearing each other's clothes, with nothing in the typography or the voice marking the switch.

In effect

The defining instance in this library is Hasard Lee's The Art of Clear Thinking. Six independent readers took the whole book and found no psychology researcher named in 252 pages: no psychologist, no neuroscientist, no decision scientist, no study, no journal, no experiment. The people named are colonels, generals, pilots, businessmen, a physicist and a missile commander.

The ideas, meanwhile, overwhelmingly belong to somebody. Emotional hijack is Daniel Goleman's. The neocortex against the limbic system is Paul MacLean's, and obsolete, stated twice as current brain science. Stress helping simple tasks and hurting complex ones is Robert Yerkes and John Dodson's, given as several studies by the Air Force. Structured repetition until a skill is automatic is Anders Ericsson's. The first number anchoring a discussion is Amos Tversky and Daniel Kahneman's, mechanism described and word withheld. The outfielder running on a constant gaze angle is Gerd Gigerenzer's. Accidents as chains bypassing redundancies is James Reason's model without the model or the name. The clearest case is the miles-per-gallon illusion at pp. 48 to 49, which is Richard Larrick and Jack Soll's finding published in Science in 2008 and arrives as the author's own worked example, with no Larrick, no Soll and no Science entry anywhere in the 43-item bibliography.

The second instance is Adrian Furnham's The New Psychology, and it is instructive because the author is a research psychologist writing inside his own discipline, so nothing in it is a borrowing from a stranger. What the two books share is the mechanism: the register of settled science applied to claims with nobody named, in type identical to the claims that are sourced. Research repeatedly demonstrates, with no citation. Ten happiness myths drawn up by leading researchers, nobody named. A now-established paradox, with no evidence offered in the chapter. Longitudinal studies have demonstrated the causal pattern, with no study named. And a series of constructs whose owners go unnamed in both the prose and the chapter bibliography, including Elaine Aron, Pauline Clance and Suzanne Imes, Amy Edmondson, Joseph Henrich with Steven Heine and Ara Norenzayan, and Melvin Lerner.

Two honest differences matter. Furnham has an apparatus of sorts, three to eight references per chapter, and he tells the reader at p. xiv that it is an idiosyncratic taster rather than a source list. Declaring it is to his credit and does not repair it, because the declaration sits in the front matter and the register sits on every page after. He was also working to a publisher's fixed length of 1,350 words an entry, which he says out loud.

The usable part of the concept is the test, because neither counting citations nor grading their class will find this defect. Take any confident psychological claim in a book and ask who said it. If the book cannot tell you, and its own back matter cannot tell you either, that is the defect. Three quick proxies: search the text for studies show, researchers found and research shows, and see whether any of them resolves to a name; check whether any named person in the book is a researcher rather than a soldier, a founder or a friend of the author; and check whether the apparatus contains anybody who could possibly disagree.

What it does not say

It does not rest on research. This is a category this publication defines and applies, not a proposition anyone has tested.

It does not say trade authors owe a literature review. Superscripts genuinely interrupt prose, ideas do enter general circulation and stop needing a citation, and nobody credits Pavlov for the word conditioning. The claim here is narrower: the constructions above are specific invocations of research authority, and an author who reaches for that authority has taken on the obligation that comes with it. The cost of discharging it is one clause.

It does not say the reader is misled about the source. The reader is misled about there being one. An unsourced claim that sounds unsourced does no damage, because in my experience tells a reader exactly what they have been given.

It does not follow from an absent apparatus. Great Minds Don't Think Alike has no apparatus at all and roughly seven claims in ten are unsourced anecdote, and it is not an instance, because it never adopts the register. A reader is not misled about where a claim comes from when no claim looks sourced.

It is not the only way a book can strand a reader. Naming the researcher is necessary and not sufficient: The Molecule of More names its researchers constantly in the prose and its back matter barely overlaps with them, and The Modern Mind names thousands of people and footnotes between seventy-eight and eighty-eight per cent of attributed ideas while leaving all forty-two of its chapter-closing theses bare. Those are separate defects with separate names in this vault.

It is not a judgement on the authors. The criticism here is of a register rather than of persons, and it is recorded as a criticism of how two books handle evidence. Lee's operational material is the best-specified thing in his book, and Furnham's map of his own field is unusually broad.

Two generalisations follow, and they are the most useful things on this page. Confidence and evidence are independent variables: The Elephant in the Brain is about forty-four per cent peer-reviewed and argues against itself in its own notes, Deep Work is about eight per cent and states two contested constructs as fact, and The Art of Clear Thinking is effectively nothing and never once sounds uncertain. And the usual proxy, whether an author seems careful, does not work here: Lee hedges his own testimony better than most writers in this library, giving his own unsuccessful rescue mission an unflattering post-mortem and saying when a probability was a guess, and the hedges vanish the instant he is relaying something he did not witness.

---