For the uninitiated, let me quickly explain what a “6-4” chord is:
- Triads have three positions, or “inversions.” These are: root position (root in the bass), 1st inversion (third in the bass), and 2nd inversion (fifth in the bass).
- A “6-4” chord is a triad in 2nd inversion, since the intervals between the bass and the other two chord members are a 6th, and a 4th. (The designation “6-4” comes from figured bass notation.)
6-4 chords are famously characteristic; they are an extremely recognizable, high-profile harmony. They are one of the sonorities that most demand something – i. e. some kind of treatment or resolution. So much so that in common practice theory, the 6-4 is often not regarded as a second inversion of a triad, but rather a bass over which two suspensions (6-5 and 4-3) occur. The 4th above the bass is probably the reason for all this. The interval of the 4th is a curious interval in that it is one of the “perfect” (doesn’t exist in major or minor forms) intervals, yet it is a dissonance. 4ths (measured from the bass) require some deliberation in order to be employed successfully.
Therefore, the use of the 6-4 has been a point of particular consideration in tonal theory. It is permissible when it is used in one of these three ways:
- Passing 6-4: A 6-4 struck as the bass ascends or descends stepwise, and preferably not struck on a strong beat. The upper voices should also preferably move by step. This helps the listener to understand the 6-4’s transitory, non-functional character in these instances.
- Cadential 6-4: A 6-4 struck as the beginning of a cadential formula, i. e. I 6-4, V, I. It is this formulation that leads many (perhaps most) theorists to think of the 6-4 as a contrapuntal construct in which the bass (on the dominant) is accompanied by a 6-5 suspension and a 4-3 suspension.
- Pedal 6-4: A 6-4 struck as a result of the movement of the upper voices over a pedal tone.
I’m not so sure we should recognize the pedal 6-4. As Schenker noted, the definitive characteristic of a pedal point is not simply length. Rather, for a pedal point to be legitimate, the upper parts must proceed through harmonies (or at least one harmony) that do not contain the pedal tone. The upper parts relate only to themselves, while the pedal tone is a completely separate phenomenon, operating independently to create a certain sonic atmosphere. Why should we suddenly recognize, during the course of a pedal point, a harmony in the upper parts that relates itself to the pedal tone?
The only way, it seems to me, to retain the idea of a pedal 6-4, is if the pedal point begins as a 6-4, which actually happens quite a lot. But in this case, one can think of it as an extended cadential 6-4.