The Chord Ontology (CO) meeting at City University has shown some of the problems in defining a CO. Christopher Sutton listed them in http://motools.sourceforge.net/chord_draft_1/chord.html , so here’s my discussion. Some of them are beyond our reach, namely those to do with the ambiguity of music itself:

Allow absolute pitch specification - by MIDI note number?

Explicitly specify whether chords are descriptive or prescriptive (ie. transcription vs. score)

Specify voicing

• Fully ? (eg. intervals are ordered and include octave information)
• Just whether a chord is played in open, closed or mixed position ?

Representation of metrical time. Currently the TimeLine^? ontology doesn't cover this case, but it would be useful (eg. for translating from MMA chord files or scores)

Fix the requirement to specify a named root note. This would be partly solved by allowing Notes to be specified by absolute pitch.

Chord shorthands (eg. 'maj', 'min7')

• Do they attach semantics or simply act as a syntactic aid
• Should 'jazz' shorthands be used instead of those from [Harte05] ? (can they co-exist without confusion ?)

-- MatthiasMauch - 26 Oct 2007