zanzibar: Not sure who submitted this game's PGN (though we could round up the usual suspects, I suspect), but it has the marks of a good historian:<[Source "Minneapolis Sunday Tribune, 1909.02.21, p.39"]>
Still, the topic of <compound sources> came up once upon a time, and this is a good moment for a recapitulation via a hypothetical...
Assuming the historian found the game on Winter's site, and used it to transcribe the score. Then the source is Winter. Of course Winter gave the primary source.
A *really* good historian will list both sources, like this:
<[Source "Winter (CN 10922) / Minneapolis Sunday Tribune, 1909.02.21, p.39"]>
Rationale:
First of all, the transcription was likely cut-and-pasted from Winter. Thus, if an error was discovered in the movelist, we should know where to point the finger (and even do the courtesy of posting a heads-up to Winter).
Secondly, the primary source given by Winter (or rather, Winter's correspondent) might be mistaken. Again, the record should be clear where the CG contributor got the source (it's conceivable they cut-out Winter by going to the newspaper archives, but we're assuming otherwise).
Ah, you might complain that Winter provided the actual scan. Still, there is no guarantee the primary source citation is absolutely correct. The previous point still stands.
Finally, giving Winter a little nod now and again is just good manners. Of course, some people might like to pull the tiger's tail, but I consider that unnecessarily risky.
A bit long-winded, but I think the points worth considering.
Kibitzing
