> > ---------- Forwarded message ---------- > From: Kieren MacMillan <[email protected]> > To: "Hans Åberg" <[email protected]> > Cc: Thomas Morley <[email protected]>, Dan Eble <[email protected]>, > lilypond-devel <[email protected]> > Bcc: > Date: Tue, 6 Jan 2026 05:42:52 -0500 > Subject: Re: "Compound" meter > Hi Hans, >
Just chiming in as an English speaker with my perspective on these. > 1. For 7/8 and sim., I like “irregular”. > I am unconvinced 7/8 represents any kind of category of time signatures that we need to label. Is there a use case that lilypond needs a word or distinction for this? I suppose that if we did want to identify signatures that were neither simple (meaning more specifically something like "even dyadic") nor compound, the term more commonly but informally used would be "odd time signature". But what are the cases where we would need to make a distinction? > > 2. For (3+4)/8 and sim., I like “additive”. > Downvote. I would reserve "additive" for the ones like 3/8 + 2/4, where we are adding together actually different signatures. If the concept here is that we have the same denominator and are just putting subdivisions in the numerator, then I would suggest a new term like "expanded" or "subdivided" time signature. Musically, I would categorize (3+4)/8 the same as 7/8, which is what I call an odd time signature, but with the subdivision spelled out in the numerator. But since this form of signature can be used in time signatures that are not odd, like 3+2+3/8, that is obviously not a good option, and avoids answering the question of how to distinguish them, since at least in terms of printing the signature, lilypond needs to treat them differently. > > 3. For 3/8+2/4 and sim., I like “mixed”. > Downvote. I would use the term "additive" for this type, where a single bar is comprised of two or more signatures that do not share the same denominator. I agree with Hans that "mixed meter" in English generally refers to anything that contains more than one time signature, at any point. In fact, I feel like it usually refers to something that does NOT have a strictly repeating pattern at the measure level, but more likely at the phrase level. Something with a strictly repeating pattern would instead fall into the categories of either odd, additive, or alternating, depending on how it was notated (and typically, someone just hearing the piece will not know how it is notated.) > 4. For 6/8 (3/4) and sim., I like “alternating”. > Ultimately, I agree with this. However, I am unclear on how common this notation is used ONLY to mean strictly alternating measures, or is it ever used to mean that each measure may be one or the other? If it is not strictly alternating, then the engraver will need to do a bunch of manual beaming (or hidden time signaure changes). I only wonder, might this have options for both strictly alternating beaming, and always defaulting to the primary signature beaming? > 5. For 1/3 and sim, I like “non-dyadic” (and dislike “irrational”). > I support "non-dyadic" but recognize that to the extent that these things are championed, the people who champion them unfortunately call them "irrational". I actually like the term "irrational" for these. Not because it is an accurate description of the time signature--which it clearly isn't. But because most advocates of these signatures typically add an irrational requirement: insisting a change in time signature also implies a metric modulation, but without notating the modulation. Since I think that is a fight worth fighting, I don't mind the shady implication of "irrational" for this practice, and therefore for this category of signatures. > 6. I would like to save “irrational” for 3/π and other truly irrational > signatures. > Since this category does not really exist, It is certainly not a good enough reason to reject "irrational" for the non-dyadic category. The reason this is a non-category or theoretical category is that the notation of musical rhythm is explicitly rational, as we choose to define duration only with respect to other durations. There is no way to use irrationals in a time signature and have the measures work out unless you are also collapsing them by using tuplets with the same irrational base. In which case, you could have chosen a more useful base to begin with. > > 7. “Compound” can be reserved for cases in which [to quote that Wikipedia > page] “the note values specified by the bottom number are grouped into > threes, and the upper number is a multiple of 3, such as 6, 9, or 12”, and > “simple” can be reserved for the rest. > Upvote. This is the only accurate way to use the term "compound meter" in English. -Elaine Alt 415.341.4954
