The Taxonomy of Expert-Report Failure

The Document You Were Told to Accept

An expert report — forensic, financial, clinical, technical, actuarial — carries a particular kind of authority. It is written by someone qualified, in a register you are not, about a question you were not equipped to answer yourself, which is precisely why you were given the report instead of the underlying material. That structure makes the report almost impossible to challenge on its own terms. You lack the specialism. You were meant to.

But expert reports do not fail in mysterious, specialism-specific ways. They fail in a small number of recurring structural ways, and those failures have signatures that are visible from outside the expertise — because they are failures of reasoning and reporting, not of technical knowledge. You cannot check whether the science is right. You can check whether the document behaves the way a sound one would. The ten modes below are that checklist. None requires the qualification. Each requires only the willingness to read the report as a structure rather than a verdict.

The Ten

1. The foregone conclusion. The report's conclusion is consistent with everything in it because everything in it was selected to be consistent with the conclusion. The tell: no finding anywhere in the document cuts against the result. Real inquiry produces tension; its absence is not cleanliness, it is curation.

2. Overinterpretation. A thin observation is made to carry a thick conclusion. A brief, narrow, or single instance becomes the basis for a broad, stable, general claim. The tell: the strength of the conclusion exceeds the weight of the thing it rests on, and the gap is bridged by the author's authority rather than by evidence.

3. Pathologising the ordinary. Normal-range variation is recoded as deficit or abnormality because the framing requires a finding. Composure becomes coldness; caution becomes evasion; an ordinary response to an extraordinary situation becomes a trait. The tell: the same behaviour would be unremarkable, or read as a strength, in a subject the report was not built to characterise a particular way.

4. The missing alternative. A single explanation is presented for a phenomenon that has several plausible ones, and the others are not mentioned, let alone excluded. The tell: you can generate an innocent or competing explanation in thirty seconds that the report, written by an expert over weeks, does not address at all.

5. The negative emphasis. Unfavourable material is detailed and foregrounded; favourable or neutral material is compressed, deferred, or omitted. The tell: asymmetry of treatment — the damaging point gets a paragraph and a quotation, the exculpatory point gets a clause, if it appears.

6. Charged language. The findings are conveyed partly through vocabulary rather than evidence — words that pre-load the reader's judgment before the basis for it is established. The tell: removing the adjectives materially weakens the impression the report leaves, which means the adjectives were doing evidentiary work they are not entitled to do.

7. Selective reporting. Source material is quoted accurately but partially, in a way that omits the qualifying or reversing context. The tell: the cited fragment supports the report's reading; the surrounding material, when you can obtain it, complicates or contradicts it.

8. The absent context. A fact is presented stripped of the circumstances that change its meaning. An act without its situation, a statement without its prompt, a number without its base. The tell: the fact is technically true and its meaning is unrecoverable from the report alone, which is the condition the omission was — wittingly or not — designed to produce.

9. Overgeneralisation. A finding valid within a narrow domain is applied beyond it. A result about one context, one moment, or one measure becomes a statement about the subject in general. The tell: the conclusion's scope is wider than the evidence's scope, and the widening is silent.

10. Narrative alignment. Every ambiguity in the report resolves in the same direction — the direction of the account the report was commissioned within. The tell: ambiguity is supposed to resolve randomly; when it resolves consistently toward the commissioning narrative, the consistency is the finding, not the conclusions it produced.

Why the Taxonomy Travels

The power of an expert report is that it relocates the question into a space where you cannot follow. The power of the taxonomy is that it does not try to follow. It does not ask you to evaluate the specialism. It asks you to evaluate whether the document reasons and reports the way a sound one would — whether its conclusions exceed its evidence, whether its ambiguities all fall the same way, whether removing its adjectives removes its case. Those are not specialist questions. They are structural ones, and the structure is on the page, available to anyone told to accept what it concluded.

That is the whole reason to write it down. An instrument used to audit reports is only honest if it can be used on this one too. Read it against any expert document you are handed — and against the analyses in this publication. A piece here that exhibits a mode on this list has failed by its own standard, and a standard that exempts the hand that wrote it was never a standard at all.