Home | Site Map | Watch | FAQ | History | Store | Contact

Conversation between Alex and Stephen (prev top next)

2007feb23

Alex, responses to some of the particulars:

In the process of generating a visual analogy of music you can initially focus on the formal elements that underpin musical structure, intervals, rhythm, chords, keys, etc. Ideally these would be informed by established culturally defined systems of notation to make them more accessible (and this style of visual representation is what I refer to whenever I use the term music visualisation loosely).

The further I get into this stuff, the more I mistrust the idea of "formal elements." At least, the ones that have historically been recognized as part of music theory. Take intervals. You and I agree about what intervals are, and when somebody plays, say, a major sixth, we can both say "that is a major sixth." But people who haven't been trained as musicians not only can't do that, but if you play the sonority of a major sixth (that is, both notes at once) by itself and say "this is what a major sixth sounds like," and then ask them whether some other given sonority has a major sixth in it, they often won't be able to tell you. To be able to do this reliably requires specialized training --- training that's not required to appreciate music. So, I wonder whether intervals are the right thing to consider "elements." If you expose somebody to a melody or a motivic theme, most people will be able to tell you whether it is present in a given passage. Do they do this by recognizing intervals? If you play Happy Birthday with the intervals inverted, most people won't recognize it, but if you play it with all the intervals changed by a half-step, most people will recognize it. This suggests that it's not the intervals per se, but something more general, like the shape of the melody.

This is fine for offering a dry anatomical analogy of musical structure ...

This leads to the question: if a given "formal element" can't be recognized, can it contribute meaningfully to the structure of a piece of music? My sense is that other things need to be recognized as elements.

Emotion is an emergent phenomenon from musical structure.

I'm not enough of a philosopher to know exactly what an emergent phenomenon is, but I don't think I agree that emotion emerges from musical structure ... not in the sense that the emotion couldn't be predicted or foreseen from the elements that come together to produce it.

The simple visual ratio of 2:1 versus a ratio of 16:15 can be easily exploited as a basis for a contrasting visual aesthetic.

When I first read that, I made a mistake; I thought you were referring to an octave (2:1) and a major seventh (15:8), and I wrote this:

I'm not sure I know what you have in mind. One of these rectangles

has sides in the ratio of 2:1 and the other has sides in the ratio of either 15:8 or 17:8. For me, it's not easy to tell which is which. Here they are again

and again

For me, the ratios of these lengths doesn't seem very obvious, and depends on context. The only thing that's obvious is that for each of them, the long side is about twice as long as the short side. Not close enough to distinguish an octave from a minor ninth. So, I don't think this is the visual expression of 2:1 that you had in mind. Now, there is certainly an obvious 2:1 ratio here

but this is a 2:1 ratio of count, not of size ...

... at which point I realized my mistake. Take #2 ...

Okay, there it's very obvious which is 2:1 and which is 16:15. But it's not so obvious that the blue one is16:15 and not 1:1 or 17:16 (or that the red one is 2:1 and not 17:8 or 15:8). So I'm still not getting it. For this kind of analogy to work, it seems like you need at a minimum for the visual domain to be analogous to the aural domain in two ways: (1) ratios that are close in size in the aural domain are close in some attribute (could be size but that's not required) in the visual domain, and (2) some attribute in the visual domain corresponds to consonance, such that an octave is seen as being more like a perfect fifth with respect to that attribute than it is to a major seventh or a minor ninth (which it is more similar to in the first attribute). There's a third thing I'd want in addition: for the visual expression of pitched sounds to contain something that would allow you to predict, ahead of time, whether a given pair of them would combine consonantly or dissonantly. For example, if the second harmonic of a note was shown, it would be easy to tell that an octave would be consonant

but a minor ninth would not

I'm not saying that this particular implementation is practice, but you get the idea.

S.

(prev top next)