This music cannot be profitably analyzed in terms of eighteenth-century tonal harmony; likewise, searching for forms common to recent Western music, for example, sonata and rondo or even simple binary and ternary forms, would be relatively futile. Certainly harmonic and formal analogies do exist, but they are clearly of minor importance. Therefore, the Western musicologist must work out a new methodology for analysis that considers the music's own particular logic and organization. He is fortunate if a music has been studied by theorists native to its country of origin and if a utilizatile framework and a terminology already exist. For the past thousand years, Persian music has rarely been at a loss for theoretical explanation. Indeed, this music has been analyzed in inordinate detail by some of the greatest thinkers of the Middle Ages. Moreover, in contemporary Persia, theoretical speculation continues.
Yet, from the viewpoint of a non-Persian musicologist, the work of Persian theorists is too greatly isolated from practice. Native systems have not come to grips with the music's primary methodological problem, which is this: how can one analyze a music that changes with each performer and even with the same performer on different occasions? Because Persian music is improvised, the form of any given performance, that is, that order or organization of sound, is far less certain than the order of a given piece of Western music, which, with certain reservations stemming from interpretation and slight differences in the sources, is the same each time it is played. But whereas improvisation is such a natural and almost intuitive procedure for the Persian theorist that he does not feel the need to explain it, for the foreign musicologist, an analysis that omits the subject of improvisation is quite an incomplete one.
One solution to the problem of analyzing an improvised music is to set aside the complete performance and study the material used as a basis for improvisation. By isolating the relatively stable and unchangeable part of the performance, that is, the dastgah, or melody type, one can deal with a body of music stable enough to analyze. Then, after the investigator becomes familiar with this basic material, it can be studied within the context of the total performance. In other words, once the model for improvisation has been clarified, the way the model is used in performance can be investigated.