Joseph Scott keeps acting like I'm ducking the issue when I ask him to define his terms or explain that I tend to use different definitions.

To be clear: He believes that some of the music black people were singing non-professionally at the turn of the 20th century was blues, that there is no evidence they got that music from professional entertainers, and that later professional musicians based their compositions on that early material.

I agree with him on all points, except his insistence that "blues" is a tangible thing that can be identified in a period when the word was not used and from which no recordings survive.

I'm not saying he can't define earlier field hollers as "blues" -- he obviously can, and does, and I understand why: a lot of music people did not call blues in 1900 is clearly related to music they did call blues in 1920, so it is perfectly reasonable to argue that it was already blues even though that terminology did not yet exist.

I treat genre terms as historical artifacts, so don't use a term to describe music made in a period when the term didn't exist. But that's a semantic choice, made because as a historian I'm wary of anachronisms, and I understand why other people make other choices.

As for the argument that vaudeville blues affected rural blues, that's easily proved once we have recordings, because we have concrete (ok, shellac) examples of rural musicians imitating records by vaudeville singers. Of course, we also have lots of examples going back to the 19th century of rural singers performing versions of songs by professional minstrel composers -- but whether one wants to call some of those examples "early blues" is, again, a semantic choice.