Key signature with neume notation

[ahankinson]

From @BaMikusi 's Word Doc feedback:

“For neume notation, the key signature MUST be omitted.”
Are you really sure there are no plainchant sources with key signature? (I have a vague memory that a single b occasionally occurs, and would ask an expert if that might make a difference.)

My take on this is that it's better to simply mark the notes as flat inline, rather than allow all key signatures with neume notation. We could specify that only a single type of key signature is allowed with neume notation (one flat) but then someone will come with their two-flat source, etc. etc.

So I'm inclined to continue to disallow this for neume notation, but I'm putting it here for further discussion.

[BaMikusi]

I see what you mean but, just to have the full picture, shall nonetheless ask a real expert as to the role of key signatures in the plainchant repertory. Perhaps she says something decisive that spares us any further disussion here.

[BaMikusi]

I have received feedback from a Gregorian expert and she confirmed my vague memory that key signature as a concept is not unknown in plainchant sources, either, and it does occur even in early sources, not merely in late retrospective anthologies already published for a tonal-ear public. That said, it is indeed exclusively a single 'b-rotundum' that can appear in the sources in such a function, and it is also true that its notation is initially not as consistent as with the later key signatures of tonal music -- whereby it is nonetheless quite typical to have a bB key signature for pieces in the 6th mode, and later on (in the "new style") for pieces with a clear major-mode character (in a sense a merger of modes 5 and 6) its use is already quite consistent, similarly to how it would count as 'the rule' in an F-major piece in later times.

With all of that in mind, it think we should allow for `bB´ as the only possible key signature with neume notation. This is the only variant that is an inherent part of the style -- which is to say that we cannot exclude that someone (as @ahankinson suggests) might dig out a source with some other key signature, but that should certainly be viewed as a white crow that PAE is not supposed to account for.