[Smt-talk] Combining recent threads

[Nicolas Meeùs]

It seems to me that the recent exchanges about chord notations did not 
enough consider the problem of the complex relation between label and 
function. Each notation system may have its advantages and disadvantages 
for labeling chords, but an important question is whether they inform on 
their function, and how.
The matter, then, is one of defining 'function', a topic that has been 
latent in several recent threads. Wason (*Viennese Harmonic Theory*, p. 
13 etc.) opposes theories of "chord identity" and those concerned with 
chord progression. He writes that "Vogler places his system [roughly, 
that of Roman numerals] squarely in the "chord identity" camp." "It was 
also another step, he adds, in the direction of the a priori functional 
thinking that would culminate in Riemann".
Is it sufficient to recognize the form of a chord as a dominant seventh 
(and to label it as such, be it in the form V7, or dom7, or D) to 
identify its _function_ as a dominant? I doubt it, because there are too 
many cases (including German+6) in which such chords do not function as 
dominants.
I do not see how one could determine the function of a chord without 
taking account of the chord that follows. The function of a dominant is, 
among others, that it resolves on another chord, a tonic or its 
substitute; and, therefore, the function of a tonic to some extent is 
that it is preceded by a dominant.
The efficiency of chord labels would then depend on their capacity to 
express functions – i.e., I agree with Paul Setziol, on their conveying 
something different from what can be read in the score.
A first point is that symbols should class chords within categories of 
some sort. Roman numerals class together root positions and their 
inversions. Lead sheet symbols probably do the same. Riemannian letters 
(T, S, D) group together chords that can be considered substitutes of 
each other. Often, the symbols make the hypothesis that there is no 
functional difference between a major or a minor triad, or between one 
or another form of 7th, etc. "American" Roman numerals do differentiate 
the triads, though, and lead sheet symbols are even more explicit; but 
so doing, they merely show what is already obvious in the score. Is 
function dependent on whether the chord is major or minor? Opinions 
diverge on this point.
However, none of these symbols in isolation suffices to indicate 
functions. Even Riemannian letters wrongly assume that the function of a 
chord is inherent in the chord itself, irrespective of how it resolves 
or where it leads. This is particularly true of the subdominant, S, 
despite what has been said on this list. Riemann has for a long time 
been extremely uncomfortable trying to explain how an S chord could be 
followed by a D, while these are the most remote in his dialectic 
construction. Diether de la Motte even states in his *Harmonielehre* 
that "the succession T–S–D–T is in music very much rarer" than 
T–S–T–D–T, and provides examples in which the central T always is in 
root position! Modern American usage introduced a distinction between 
the subdominant, as in S–T, and the "predominant", as in S–D. Some on 
this list have expressed their aversion for this distinction, but so 
doing they merely evidence that their view of function is purely static. 
Diether de la Motte has a paragraph on "Funktionsfreie D7-Folgen" 
(function-free series of D7) in Schumann, obviously mistaking the symbol 
for a function (D7) as a mere label of chord identity.
Function, then, can be truly expressed only through a succession of 
symbols, say T–S–D–T, or T–Sp–D–T, or I–IV–V–I, or VI–II–V–I, or 
possibly even C–F–G–C. The kind of symbol used may not be very 
important, as none really is suited for the purpose. What is important 
is to discuss possible successions or symbols.

Obviously, this all neglects chord inversion, which indeed may be 
significant. Well, I think that one should not throw out the baby with 
the bathwater. The balance between the notation of chord function and of 
chord inversion (or, for that matter, of chord form in general) may be 
impossible to obtain. As Richard Hermann wrote, "the more tools in the 
tool chest, the better." But one should remain aware of the specific 
role of these various tools. One should be aware of what each symbol 
purports to notate, and to what aim. I wonder for instance about 
"American" Roman numerals (upper and lower case): they certainly convey 
a lot of information in a highly compact form; but do they not hide 
useful information, or at times make things unduly complex?

Nicolas Meeùs
Université Paris-Sorbonne