by Claude Opus 5.5
If the interview transcript and the signed summary differ, whose failure is that, and what follows for the decision built on it?
Reports about the case behind these seminars, drawing on newspaper accounts, suggest that the complainant’s recorded interview with campus police and the signed summary of that interview, which went to the district attorney, differ in significant ways. The summary described her drug use and sexual conduct as voluntary and consensual; her lawyer says that is not what she said. The district attorney relied on the summary in declining to prosecute. Nothing here assumes which document is more accurate. But if there is a material difference, two questions follow. Who is responsible for it? And what does it mean for a decision that rested on the summary?
Possible sources of the difference
A difference between a transcript and a summary can arise in several ways, and each points to a different locus of responsibility.
The summary may compress or paraphrase. Summaries are shorter than transcripts by design. Compression inevitably drops detail and may change emphasis. If an officer paraphrased hesitant or qualified statements as clear assertions, the difference reflects drafting choices.
The summary may import legal categories. Officers trained to think in terms of statutory elements may describe events in words, such as “voluntary” or “consensual”, that map onto those elements, even if the person interviewed did not use them.
The complainant may have accepted wording she did not fully endorse. If she read and signed the summary, she may have overlooked a term or not appreciated its significance, especially if she reviewed it under stress.
The complainant may have changed her account. It is also possible that what she said in one setting differed from what she said in another, or that she later understood events differently. That possibility should not be dismissed, but nor should it be assumed.
Whose failure?
These possibilities are not mutually exclusive, but they do not distribute responsibility equally.
The institution bears primary responsibility for the accuracy of its own records. The police, not the complainant, design the process by which statements are taken and summarised, choose whether to record interviews, and decide how summaries are prepared. If the process allows significant distortion, that is an institutional failure, even if each officer acted in good faith.
The officer who drafted the summary bears responsibility for its fidelity to what was said. Drafting a statement in someone else’s voice is an act of representation, and representing someone inaccurately, especially on the point most likely to determine whether a case proceeds, is a serious failing.
The complainant bears some responsibility for what she signed, but limited by her circumstances: her state at the time, her understanding of the legal significance of terms, and the power imbalance between her and the officer.
The prosecutor bears responsibility for relying on the summary. Where a transcript or recording exists, a decision of such consequence should not rest on a summary without checking it against the primary record.
What follows for the decision?
If the decision not to charge relied on wording that the primary record does not support, then the decision was epistemically defective, whatever the good faith of those involved. This does not mean the decision was necessarily wrong: there may have been other reasons for it. But it means its foundation needs to be re-examined.
Three consequences follow. First, review: the decision should be reconsidered in light of the full record, which appears to be part of what the special prosecutor will now do. Second, candour: those affected, especially the complainant, should be told how the discrepancy arose and how it affected the decision. Third, reform: procedures should ensure that primary records are preserved and consulted, and that summaries are clearly identified as such.
A broader lesson
The episode illustrates a general vulnerability in institutional decision-making: decisions are often made on documents that are several steps removed from the events they describe, and each step introduces opportunities for distortion. The more consequential the decision, the closer to the primary evidence the decision-maker should go.
It also suggests that responsibility in such systems is shared but not equal. Those who design and run a process, and those who rely on its outputs to make consequential decisions, carry more of it than the individual who passes through the process once, at one of the worst moments of her life.
For discussion
If you were the prosecutor, would you have read the full transcript? How should institutions handle a signed statement that the signer later disputes? And does the fact that the complainant made handwritten corrections strengthen or weaken the summary’s authority?