The argument about whether gymnastics is a sport has been running for as long as I have been watching it. It never moves. One side points out that nobody crosses a line first in a gymnastics final and that the result is a number somebody has decided on. The other side points at the training and the injuries and asks what more anybody wants. Both sides are arguing in good faith and both keep stepping over the point the other is making. A gymnast trains harder than most sprinters, and anybody who has watched a vault landing knows the risk is not decorative. A sprint result is also settled by a clock, and not by nine people with clipboards, which is a genuine difference and not a snobbish one. The trouble is that the word sport is being asked to do two jobs at once, and only one of those jobs is a question about the athlete.

The first job is a claim about effort and skill. On that measure the case closed years ago. The second job is a claim about how a result is produced, and the two sides never notice they have swapped from one to the other. A race produces its result by measurement. The finishing order exists before anyone announces it, so an announcement that contradicts the order can be shown to be wrong by going back to the recording. A judged event produces its result by opinion. Nine trained people watch the same ninety seconds, each writes down a number, and the extremes are thrown away before the rest are averaged. There is no fact underneath that score which the score is trying to match. This is not a complaint about gymnastics. It is a description of what judges are for. We use them precisely where the thing we care about, the quality of a performance, has no instrument that reads it, and the panel at a world final sits behind a row of monitors on the far side of the floor for exactly that reason. The mistake is to treat the judged events as failed measured ones, waiting for a better stopwatch.

What the definitional argument is really about, I suspect, is embarrassment. A measured result can be wrong in a way everybody can see, in which case the wrongness belongs to a machine. A judged result can be wrong in a way everybody can see, and then it belongs to nine named people who have to walk out of the arena afterwards. That is uncomfortable. The discomfort is worth more than the definition ever was, because somebody signed that score. You can find out who they are and watch them do it again next week. So my answer to the question is that the question is the wrong one to be asking. Asking whether gymnastics is a sport gets you a ruling about a word. Asking who is answerable for the number gets you something you can act on, and the judged events already answer it better than the measured ones do: they publish a panel, publish its scores, and let anyone set one judge against the other eight. The measured events have built no such machinery because they have never needed it, and a championship whose timing system fails discovers within the hour that it has no practised way of arguing about a result at all. Sport is a word we hand out. Answerability is a thing we can inspect. I would rather argue about the second.