The Interval Is Not the Tradition
09 November 2026
Comparative music-modeling essay

Abstract
Praxymuse v0.6 separates mathematical pitch relations from the contextual grammar of musical traditions. Its tuning equation and cents formula describe intervals precisely, but its mode representation also includes gravity, phrase grammar, ornament and intonation. This essay asks what gets lost when a musical system is reduced to a pitch collection. It proposes a disciplined comparison rather than claiming new ethnomusicological findings. [1, original pp. 4–5, M03 and M06]

A ratio makes one statement
The source writes equal divisions of an octave using a reference frequency and an exponential pitch relation. It also records rational frequency ratios and the cents measure, 1200 log₂ of a frequency ratio. These are quantitative descriptions of interval size. They do not specify which note should feel like an arrival, how an ornament unfolds or what a phrase means in a particular performance setting. [1, p. 4, M03]

For an octave ratio of two, the formula gives 1200 cents. For a unison ratio of one, it gives zero. Those identities follow directly from the logarithm. They do not establish that listeners treat all equal-sized intervals identically across registers, timbres or contexts. Mathematical equality of a descriptor is narrower than equivalence of musical function.

A mode has a direction
Praxymuse defines a mode with intervals, gravity, phrase grammar, ornament and intonation. It states that twelve-tone equal temperament is optional and that raga, maqam and patet require their own phrase grammars. This is a warning against treating a named tradition as a preset scale attached to an otherwise unchanged generator. It is not a completed formal model of those traditions. [1, p. 4]

A pitch inventory tells a system which materials are available. A grammar can constrain when and how they are used. A phrase may establish a center through repetition, approach it indirectly or withhold it. Continuous inflection and ornament may matter as much as nominal note labels. Reducing these behaviors to a list can preserve membership while erasing the relations that make the phrase recognizable.

Optimization carries a style
The source offers a voice-leading cost containing movement, range, crossing and parallel-motion terms, while explicitly labeling the penalties tradition-dependent. A penalty against a particular motion is a modeling choice rooted in a target practice. It is not a universal law that all music must obey. [1, p. 4, M03]

Suppose an optimizer minimizes pitch movement alone. It may favor a very smooth succession even when a dramatic register shift is structurally important. Adding a range penalty or a phrase-level objective changes the problem. Such choices should be described as decisions about the intended idiom and function, not as corrections toward an absolute optimum called musical quality. Dissonance may resolve, persist or be reframed; the source leaves those possibilities open.

An atlas needs context
The historical and cultural atlas records region, era, community, occasion, tuning, meter, rhythm, form, grammar, ensemble, technique, production, affect and provenance. It distinguishes a tradition from an instrument class, influence from license and chronological order from quality order. Its inventories are finite references, not a claim to universal coverage. [1, p. 5, M06]

This design has a practical consequence for dataset construction. A recording should not be labeled solely by a broad geographic tag when its occasion, community and performance practice are known. At the same time, missing information should remain missing rather than being inferred confidently from sound. An instrument name is insufficient evidence for the social or historical identity of the music around it.

A proposed comparison
Choose one narrowly defined, rights-cleared source collection with reliable contextual documentation. Ask qualified reviewers to identify a small set of phrase-level features relevant to that collection. Compare a pitch-set-only generator with a generator that additionally represents those features, keeping the available recordings, evaluation excerpts and resource budget matched. This is a proposed research design; no source collection or listening panel was assembled here.

Evaluation should separate interval accuracy, phrase plausibility and contextual appropriateness. A system could improve one without improving the others. Reviewers should be able to mark uncertainty or disagreement, and their expertise should be described without presenting any one panel as the sole authority over a living tradition. Outputs must not imply endorsement by the source community merely because the experiment used its materials.

Conclusion
Interval mathematics is valuable precisely because it makes a limited claim clearly. It becomes reductive only when its scope is expanded to stand for an entire musical practice. A stronger model connects ratios to phrase behavior and provenance while remaining explicit about what it has not learned.

Reference
[1] Praxymuse Theorem v0.6, Source-preserving mathematical integration, 6 November 2026, original PDF pp. 4–5, modules M03 and M06. The supplied scientific excerpt excludes embedded source attachments. The proposed comparison is not a completed experiment or a claim of cultural expertise.
