When musical historiography returns to the beginnings of music theory, a remarkable fact presents itself: in the first essays aiming to organize this language, it is not auditory perception that is taken as the fundamental basis, but rather a kind of logical thinking. This statement leads us, however, to two complementary problems. They are the subjects of this work.
Firstly, if the above statement is true, it is reasonable to suppose that the concept of music was understood in a completely different way than came to be considered later, that is, it was viewed as the organization of perceptive sound phenomena. The second problem comes from the justifications that the historiography itself presents for this statement. In our point of view, the statement is indeed true, but not for the reasons usually given, which suffer from a methodological problem: to take ancient music theory out of the historical context, restricting thereby to technical aspects, hindering the comprehension of the concept of music.
Accordingly, we recover the musical concepts of Antiquity -- Pythagoras and his School -- as primary sources of information. They are analyzed in parallel with texts by other philosophers from the same period.
It follows that the Greek concept of mousike must be considered to be distinct from that of "music" as understood by music theorists since the 18th century.