For Western classical music, this is often claimed and is probably only a little true these days (e.g. Schenker is truly post facto, though illuminating). Howe

For Western classical music, this is often claimed and is probably only a little true these days (e.g. Schenker is truly post facto, though illuminating). However, the theory really was deeply tied to the practice of composition and performance. One of the first exercises Mozart set his composition students was to write out various cadences. Likewise, Bach started his pupils with voice-leading. Theory was also needed for various forms of performance -- look at any manual on continuo playing (this is where the figures on roman numeral notation come from). Handel wrote a manual on continuo playing in which he touches on how to improvise a fugue!