Thursday, January 14, 2010

The two leading tones in Schenkerian analysis

This is a continuation of the topic from the last two posts, where (two days ago) a conflict between melodic parsimony and a melodic descent from ^3 to ^1 was observed to be unresolvable in four-part writing; and (yesterday) basic sets of oppositions with unequal (marked and unmarked) terms were seen operating in Schachter's reading of D779n13.

One can preserve Agawu's syntactically necessary "^2-^1 over V-I" only in three-voice counterpoint and then only by accepting incomplete tonic chords -- see below. In other words, three-part writing is no better than four-part writing at solving the conflict, but it is certainly better at revealing the basic elements of the archetypal progression that Agawu invokes: I-V-I in the bass with a progression including the descending line from ^3 and both leading tones resolving correctly.

In two-voice counterpoint, of course, the ^2-^1 is possible but not the V-I (at least, not in the Fuxian species), which is why the structural levels of a Schenkerian graph can be understand as replicating the order of the Fuxian species only conceptually, not literally (we just saw something nearer the literal in the examples above). Schenker was interested in marrying the melodic principle in the old cadence, with its 6-8 or 3-1 interval sequences, to a nineteenth-century, abstract conception of harmony, and so he resorted to a Hegelian dialectic of basic musical forces: melody vs. harmony in synthesis become the counterpoint of the Ursatz -- the rational perfection of first species, not its duplicate or imitation.

In the old clausula vera or Tenor-Klausel, there are two equal "leading tones," the equivalents of ^2 and ^7 in the major-minor scale system. Here is an example of the latter as it is expressed in the final two bars of J. S. Bach, Eb-major Cello Suite, Prelude.

As I suggested in yesterday's post, however, Schenkerian analysis has devices that remove the ^7 to later levels than the ^2. Two of these we have already seen:(1) the bias toward descending lines from ^3, and (2) the attendant willingness to give priority to inner voices if they carry such lines. The other two are, in a way, more pernicious: (3) the leading-tone third line, which provides "automatic" explanations for ^7 as subordinate to ^2; and (4) the implied ^2, which supplies ^2s even if not present in the score.

This erasure of the ^7 has some far-reaching consequences for interpretation.