May I float the idea that sometimes people get the definition problem base over apex.

I think Steve (who I know has supported the more 'factual' approach of Roud) might find something to sympathise with in the idea that you have a look at what happened (which isn't always possible) and then find words to describe it afterwards.

Whereas (and here I risk getting on my hobby horse) sometimes, it seems to me that within 'folk', people adopt definitions for ideological grounds and then seek to select the data to fit them.

It isn't even as if I dont' sympathise fully with a lot of leftie stuff (though often it is just so out of date viz a viz climate crises/identity/race etc), just that the analytic/historical side of me sees it so clearly whereas people inside what I sometimes call 'the bubble' don't.

Sorry if I appear to be trolling, let me know if this is how you see it and I shall go away (once bitten twice shy)