Why do tendency tones and non-chord tones usually resolve down not up?

Not all the time but in voice leading the dissonances usually resolve down.. suspensions, leading tones, 7ths etc... they always are taught to resolve downwards. Why not up?

• Dissonances are usually caused when one note is 'off key' by a semitone. It will depend which way takes that dissonance back to the consonance.
– Tim
Sep 22, 2021 at 9:47
• Wait, leading tones typically resolve down? I thought they typically resolve up. Sep 22, 2021 at 12:10
• @Dekkadeci I think we're using "leading tone" in the broader sense, in which in can be any non-chordal neighbor resolving by half step, not just ^7. Armani, I'm not sure the premise is true. Consider, in C major, a tritone of C with an F# above. The F# will resolve upward to G. Sep 22, 2021 at 12:43
• Ok maybe it is just my perception thus far... So not then tendency tones tend to resolve both ways? Why do suspensions always resolve down then?
– user35708
Sep 22, 2021 at 13:45
• Sus4 normally resolves down. Sus2 might well resolve up. Sep 22, 2021 at 13:56

I think you be conflating tendency tones and resolution of dissonance, especially suspensions.

A common definition of tendency tones is scale degrees of the subdominant and leading tone and their respective tendencies to resolve down to the mediant and up to the tonic. In that regard the upward and downward motions are equal.

Also, chromatic alterations are described has tendency tones, as in for example the augmented sixth chord, where the tendency tone resolves in the direction of the chromatic alteration. So, the augmented sixth resolves up. But, it a chord like `V7/IV` the altered tone would be a lowered seventh scale degree and it would resolve downward. In both of those examples we can relate them back to the two "essential" tendency tones. The upward resolution of an augmented sixth is a leading tone motion and the downward resolution of a lowered seventh is like a subdominant to mediant motion.

By definition suspensions always resolve downward. If a suspended tone resolves upward, it is called retardation, which is a much less common resolution. You can also relate some suspensions back to tendency tones like when suspended sevenths or fourths resolve to a chordal third it is similar to the subdominant to mediant tendency motion.

I think the important distinction to make is tendency tones don't necessarily involve dissonance, but suspensions do. For example, when you have a progression like `V6 I` the leading tone tendency is the resolution upward to the tonic, but no dissonances occur in that progression.

If you mix tendency tones and suspension resolutions together, or especially only consider dissonant tendency tone progressions, I suppose the downward motions will seem more frequent. But, if you consider plain triadic descending fifth progressions from a dominant or secondary dominant involve tendency tone resolution, you will find upward tendency tone resolution very frequently.

I don't think there's much to this premise by itself. It just so happens that the main upwards-resolving tone is the third of the dominant, which you would consider a consonance. Counterpoint generally eschews parallel motion, so as a consequence any other leading tones will tend to go downwards.

But this is by no means an absolute rule. The most obvious example of an upwards-resolving dissonance is the augmented sixth.