Ternary scale theorems: Difference between revisions
No edit summary Tags: Mobile edit Mobile web edit |
Tags: Manual revert Mobile edit Mobile web edit |
||
| (14 intermediate revisions by the same user not shown) | |||
| Line 343: | Line 343: | ||
== Theorem 8 (Classification of bicatenary ternary scales) == | == Theorem 8 (Classification of bicatenary ternary scales) == | ||
A ternary scale is ''bicatenary'' if | A ternary scale word ''s'' is ''bicatenary'' if ''s'' has a torsion-free equave-equivalent pitch-class lattice and has an interval vector '''g''' such that the pitch class lattice consists of two parallel chains of '''g''', differing in length by 0 or 1. | ||
If a bicatenary | If a ternary scale word is bicatenary, then it is (a) odd-regular, (b) even-regular, or (c) a 2-[[interleaving]] of a primitive MOS, of [[MOS substitution]] type {{nowrap|''a'''''X'''(''b'''''Y'''[''a'' − ''b'']'''Z''')}} where 0 < ''b'' < ''a'' and gcd(''a'', ''b'') = 1. | ||
=== Proof === | === Proof === | ||
If |''s''| is odd, then we may assume that |'''g'''| = ''k'' is even. Now it is routine to get a splitting of the two chains of '''g''' into an AGS of ''k''/2-steps. | |||
If |''s''| is even, then the chains are of the same length, and thus ''s'' is a parallelogram scale with dimensions 2 × |''s''|/2. Since ternary parallelogram scales are MOS substitution, one letter, say '''X''', is a slot letter, parallel to one of the parallelogram's sides. We have two cases: | |||
Case 1: '''X''' is parallel to the short (2-long) side of the parallelogram. In this case, |''s''|<sub>'''X'''</sub> = |''s''|/2, and the scale is a 2-interleaving of a MOS with |''s''|/2 letters. Said MOS must be primitive because the pitch-class lattice of ''s'' is torsion-free. | |||
Case 2: '''X''' is parallel to the |''s''|/2-long side of the parallelogram. In this case, '''X''' occurs an even number of times, namely {{nowrap|2({{!}}''s''{{!}}/2 − ''m'')}} times where ''m'' is the number of axial steps '''X''' makes in the parallelogram lattice. Now, following our argument (in [[Ternary parallelogram scales are MOS substitution]]) showing that the filling word is a MOS, when you kill the dimension corresponding to the axial step, the resulting non-'''X''' steps are forced to be one of "move left" and "move right", and they must alternate. Thus we have an even-regular scale. {{Qed}} | |||
== Open problems == | == Open problems == | ||