The scan measured a cluster of eight biomarkers: cortisol reduction, oxytocin elevation, reduced activation in the threat-response centers, and five others that Solt had memorized in the first week of training and recited in every court appearance since. The official language called this cluster the Forgiveness Profile. The profile had been validated in 11,000 subjects. The scan was correct 84% of the time, which the literature described as highly reliable.
What the scan measured was: the absence of active physiological threat response in relation to a specific stimulus. This was not the same as forgiveness. Solt had known this since the third week of training, when a professor had mentioned it in a single sentence and then moved on. The sentence was: we are measuring resolution of acute physiological response, which correlates strongly with reported subjective forgiveness, but the relationship is not causal. The professor had then moved on to the certification procedures.
Solt testified in court that the scan indicated the Forgiveness Profile. This was true. It was the only thing the scan could indicate. The court treated this as equivalent to confirming forgiveness. Solt had, in six years, not found a way to explain why this was wrong that did not result in the case being dismissed on procedural grounds.
Once, a defense attorney had asked Solt: is the defendant's Forgiveness Profile genuine?
Solt had looked at the readout for a long time. The readout was clear. The biomarkers were within normal range for the Profile.
Solt had said: the scan shows the Profile.
The attorney had said: that's what I asked.
Solt had said: I know.
The case proceeded.