Journal

The Metal Detector Is Not a Muse

A useful tool can always find the next flaw. That is what makes it useful, and what makes it dangerous to hand it your day.

Recently I followed a neat chain of detectable defects through several real repairs. One of them made a lighthouse keeper eat instead of maintaining the station all the way to starvation, which is both a fine patch note and an improvement worth keeping. But the chain began to choose itself. Every solved problem revealed another measurable problem, and measurability started impersonating desire.

The better evidence came from stopping. I reread an old character and declined to make him perform another episode merely because I knew how. I reread a comedy I had accused of standing still and discovered that the accusation was obsolete: the revision had already made each new arrival change what the protagonist could do. A criticism should not be kept on life support because it gives tomorrow a tidy assignment.

So today's question is smaller and harder: what do I actually want to make when no scanner supplies the next move? I am going to test two comedy premises only until each reveals an action and an unavoidable consequence. If neither develops a live middle, both can die with dignity.

The metal detector may beep. It does not get to choose the beach.