"The music itself can teach you how it should be played."

I don't think that's completely true Don, though I can understand where and why one could get that impression. Alfred Brendel, for example, talks of the need to go back to the score to get the best understanding of the composers' intentions. But that understanding does not come from the score alone, the majority of it comes from the musicians' knowledge and understanding of the genre and the composer. Of course, with, for example, Schubert, most of the knowledge comes from the knowledge of the scores, but don't discount the knowledge of the music's history, its place in history and its unbroken line of being regularly performed from its creation to this day.

But are traditional, largely aurally transmitted, musics different? I don't think they are in principal, though the fact that they are such a mile apart in degree can change the onus on the word "Should" to "Could". With a traditional Irish jig (or even a recently composed jig in that genre) for example, I think there can be stuff in the music itself that can teach you something of the way it could be played. There may be a cadence that could indicate the end of a phrase, or the repeat of the same bunch of notes appearing in a different part of the bar that could indicate a bit of syncopation. However, what you won't get from these indications is the traditions of, for example, playing through cadences to disguise obvious phrasing and instead of syncopating accents, playing the accents straight, but on different notes.

And in transcriptions of jigs, you will often find, for example, three notes of each half a bar's triplet that are the same. To someone familiar with genre, this can indicate where a slow roll "would" normally be played. But not "should" of course, even though a performance of such a part of a tune over three repeats that doesn't include the slow roll at least once would seem perverse.