There can be little doubt that music theorists value consistency as much as any epistemic value. The reasons for this are clear enough. Claiming that something and its opposite are both true creates difficulties in making predictions; though prediction may not be the sole purpose of scientific inquiry, it is always the bottom line. To quote Quine: “[Prediction] is what gives science its empirical content, its link with nature. It is what makes the difference between science, however high flown and imaginative, and sheer fancy.” When we confirm a theory, we do so through “the verification of its predictions.” The more predictions we verify, the more confident we will be about using our theory. The same can be said about building and testing music theories, as Schenker made perfectly clear near the start of Kontrapunkt I. According to him:

Several pages later, Schenker reinforced the point by noting that he was primarily interested in describing the abstract effects that a particular tone might have on the motion of a voice and not the psychological effects it might have on a listener: “Tones mean nothing but themselves; they are as living beings with their own social laws.”

Although Schenker went to great lengths to ensure the consistency of his theory, he repeatedly ran into problems in one particular area: the treatment of parallel perfect octaves and fifths. To understand the source of these contradictions, it is important to remember that tonal voice leading is founded on the notion that contrapuntal lines tend to move in contrary motion or in parallel thirds and sixths.