ABSTRACT
The hotly debated questions of the 'Urgeschichte' of the Gregorian chant, the relationship of the Carolingian period to the Roman heritage, the priority of the Old Roman and Gregorian versions, the problems of orality, the possibility of getting an overall view of an age prior to the employment of notation: all these are, in reality, different aspects of one and the same question. The Old Roman or Ambrosian parallels confirm clearly that the identity of a piece depends more upon its underlying type melody than upon a particular elaboration. On the other hand, Gregorian tradition, in spite of regional differences and some occasional anomalies, demonstrates a homogeneity even with respect to minor details. The performance practice is not at odds with the observations gained from the comparison of Old Roman, Gregorian, and Ambrosian repertories. The structures of the chants might exist as abstractions out of which the continuous work of the various scholae crystallized an exact form of the melody.
