Re: Chords in LilyPond

[David Kastrup]

Ivan Kuznetsov <address@hidden> writes:

> Kieren MacMillan <address@hidden> wrote:
>>
>> How is it "wrong" for the chord <c e g a> to [additionally] include the
>> information 'root = a'?
>
> In some instances the root could be C and the A would be a passing tone.
> In other instances, calling any of those four tones a root would
> have no meaning.  It would depend on the context.
>
> As I read through this thread though, the sense I get is that
> "chord semantics" are to be additional information optional
> added by the user, so that my fear of lilypond doing ad-hoc
> amateur musical analysis will not be happening (I hope not).

The ChordNames context is already doing ad-hoc amateur musical
analysis.  When the music has been entered in \chordmode, additional
information is placed into the notes that makes the ad-hoc analysis much
more likely to come up with something approximating the sense of the
original input.

This project basically is about being able to preserve more of the
original sense and thus cause ChordNames and its ilk to do _less_ rather
than more ad-hoc amateur musical analysis.

-- 
David Kastrup