The Errata File: How the Sturgess Inquest Handled Evidence That Changed

A record under revision

In a hearing of this length and complexity, corrections are not an aberration. They are an expected part of testing evidence. The more important question is not whether a witness, expert or document was revised, but how the revision was handled once it reached the record. At the Sturgess inquest, the answer has depended on a quiet but significant procedural discipline: the original evidence is not simply erased. It is explained, tested and placed alongside the correction so that the jury can see the movement in the account.

For a public-interest hearing, that discipline matters. The Sturgess inquest is not merely a dispute between lawyers. It is an inquiry into how a woman died after exposure to a nerve agent in Wiltshire in 2018, and into the decisions and failures that surrounded that death. If the record is to carry public confidence, it must be able to show its own workings. A corrected statement, a clarified expert report or a witness who returns to the box to alter an answer are all part of that showing.

The routes by which evidence changes

Corrections at an inquest of this kind arrive through several routes. A witness may provide a supplementary statement before giving oral evidence. A counsel to the inquiry may put a previous account to a witness and invite clarification. An expert may file an addendum after seeing another report. A document may be re-dated, re-read or found to contain a transcription error. Occasionally, a witness in the box will simply say that an earlier answer was wrong.

The procedural response is rarely dramatic. The coroner or counsel will usually ask the witness to explain the change, whether it is a slip of memory, a misunderstanding of the question, or a substantive revision. Where the change is material, the witness may be asked to provide a short written note. That note can then be read into the record or treated as fresh evidence, with core participants given an opportunity to respond. The aim is not to trap the witness. It is to ensure that the jury is not left with two incompatible accounts and no way of understanding the distance between them.

This is particularly important where the correction touches on matters that are central rather than peripheral. A revised date or a corrected job title may be a minor administrative matter. A change in recollection about a warning, a visit, a handover or a decision is different. The procedure must be proportionate, but it must also be visible.

Reading a correction into the record

When a correction is made, the original answer normally remains in the transcript. That is as it should be. A hearing record is not a clean manuscript prepared for publication; it is a sequence of questions, answers and interventions, some of which will later be qualified. The jury is entitled to see both the first account and the revision, not merely the tidied version.

In practice, the coroner may direct the jury to note that a witness has corrected an earlier statement, and may explain why the correction was made. If the correction arises from a document that the witness had not seen when first interviewed, that context may be relevant. If it arises from a lapse of memory, the jury may be invited to consider the passage of time since 2018. If it arises from an expert's reconsideration of the evidence, the jury may need to understand the difference between a change of opinion and a clarification of language.

The key point is that a correction is not automatically a concession. Nor is it automatically a sign of unreliability. It is evidence about evidence. The jury is asked to weigh it in the ordinary way.

The coroner's direction on reliability

The coroner's role is to keep the jury focused on the difference between reliability and certainty. A witness who corrects a detail may still be reliable on the main substance. A witness who never corrects a detail may still be mistaken. The jury is not required to choose between accepting everything a witness says or rejecting everything. It can accept part and set aside part.

In directing the jury on reliability, the coroner is likely to emphasise several considerations: whether the correction was made promptly, whether it was explained, whether it is consistent with contemporaneous documents, and whether it touches on a matter that the jury must decide. The jury may also be invited to consider the pressure of giving evidence in a public hearing, the length of time since the events in question, and the difficulty of recalling precise conversations years later.

That direction is not a softening of scrutiny. It is a recognition that fact-finding in a public inquiry is not a laboratory experiment. It depends on human memory, documentary records and expert judgment, each of which has limits.

What corrections reveal about care and limits

The correction column is therefore a useful measure of institutional care. A hearing that never corrects anything may look tidy, but it may also be a hearing that has not tested its evidence hard enough. A hearing that corrects too often, or too late, may raise different questions about preparation and disclosure. The Sturgess inquest sits somewhere between those poles. Its record contains moments of clarification that show witnesses and lawyers engaging with the detail, and moments that expose the difficulty of reconstructing events that took place years earlier.

Those difficulties are not an argument for treating the record with suspicion. They are an argument for reading it with care. The public-interest case for the Sturgess inquest does not rest on the idea that its record is flawless. It rests on the idea that the process by which the record is made, tested and corrected is open to inspection. The errata file is not a side issue. It is part of the evidence of how the inquiry has gone about its work.