I take your point. The semantics to me aren't really that important. Good doctors who have made a lot of observations and read the results of many medical exper

I take your point. The semantics to me aren't really that important. Good doctors who have made a lot of observations and read the results of many medical experiments and research are probably pretty qualified people. In a sense when they do "pattern matching" they are performing micro-experiments. They take in synptoms as data, draw out a hypothesis and then treat as an experiment. If the treatment is a success it proves their illness hypothesis.