In all of these cases, I encountered witnesses who struggled to explain what happened when the only people present wore uniforms. That struggle was not, in my observation, a sign of dishonesty. It was a sign of the conditions under which the event occurred. A witness who has been stopped, searched, or detained does not observe the event the way a bystander observes a traffic accident. She observes it while frightened, while being told what to do, while trying to decide whether compliance will protect her or endanger her. Later, in a deposition or on a witness stand, she is asked to reconstruct that event in chronological order, with distances and times, as if fear had not compressed or distorted her perception. When her account wavers on a detail — the exact words used, the order of two events that occurred seconds apart — opposing counsel treats the wavering as a reason to disregard the whole account. I learned to listen for the core of the testimony, the part that did not change across retellings, and to separate it from the peripheral details that memory naturally revises.
I also learned that credibility is not distributed equally. An officer testifying in uniform, speaking in the measured cadence of training, is often received as inherently reliable. A civilian witness, nervous, unfamiliar with courtroom language, speaking in fragments and correcting herself, is often received as uncertain, even when she is telling the truth. That disparity is not a rule of evidence. It is a habit of perception, and a judge who does not recognize it will reproduce it without intending to. I made it a practice, as counsel and later as a judge, to ask myself a specific question when evaluating testimony: am I crediting this witness because of what she said, or because of how she said it, and who she is? If I could not answer that question honestly, I had not yet finished evaluating the testimony.
The lesson that stayed with me from those years is not complicated, though it is difficult to apply. Truth is not always difficult to find because nobody knows it. Sometimes it is difficult to establish because the people who know it are afraid to speak. I do not mean afraid in the abstract sense. I mean afraid in the concrete, practical sense: afraid of retaliation at work, afraid of being stopped again, afraid of being disbelieved in public, afraid of the cost of telling the truth in a forum where the other side has an institution behind it and the witness has only herself. Fear does not make a witness unreliable. It makes her reluctant, and reluctance is often misread as evasiveness, and evasiveness is often misread as falsehood.
I saw that misreading happen in cases where the objective record, when it finally surfaced, confirmed the reluctant witness. A recording, a log, a photograph, a document filed by someone who did not understand its significance at the time — these would arrive late, sometimes after a witness had already been discounted. The lesson was not that witnesses should be believed without testing. Testing is essential. The lesson was that testing should be directed at the substance of what is claimed, not at the witness's composure while claiming it, and that a court should be slow to conclude that silence means absence. Sometimes silence means that the person who could speak has calculated, reasonably, that speaking will cost more than it will gain.
That calculation is familiar to me for reasons that have nothing to do with my legal training. It is a calculation I watched being made, carefully and daily, by someone who had no legal training at all, and who understood the cost of speaking long before I learned to name it in a brief.
My public defender years gave that lesson its first form. I remember, in general terms rather than in the particulars of any one file, the clients who arrived with a version of events that did not match the police report, and who were advised — sometimes by well-meaning people, sometimes by the structure of the process itself — to accept the report's version because contesting it would be too risky. A plea offer would be presented as a certainty, and trial as a gamble. The certainty was immediate: a known sentence, a known outcome, a case closed. The gamble was everything else: time in custody while awaiting trial, the possibility of a higher sentence if convicted, the exposure of family members to testimony, the risk that a jury would credit the uniform over the defendant regardless of the evidence. Many clients chose certainty. I understood why. I also understood what that choice cost the record. When a plea is entered, the facts are rarely tested. The report becomes the history, not because it was proven, but because it was never challenged.
I did not blame clients for that choice. I blamed a system that made the choice so uneven. My job, as I understood it, was to make the choice an informed one. That meant telling a client plainly what the evidence showed, what it did not show, what a motion to suppress might accomplish, and what it might not. It meant explaining that a search could be unlawful even if contraband was found, and that the law's remedy for an unlawful search was exclusion, not approval. It meant explaining that an officer's testimony could be impeached with his own prior statements, but that impeachment required obtaining those statements, and that obtaining them required time and motions and persistence. Some clients, after hearing all of that, still chose to plead. Some chose to fight. My respect for both decisions grew over time, because both were made under pressure that I, as counsel, did not personally bear.