I’ve been wrangling with software this morning – trying to get Finale to do tuplets across barlines. I’ve trawled the web for other examples where people have come to a similar problem and found solutions, yet nothing seems to really fit. The problem is twofold: I want to have very complex tuplets crossing barlines, and I want them to be accurately produced in when exported to midi, since this is for the midi piano part in my new piece Si el clima.
I’ve decided that at this stage it is unfeasible to try what I was proposing. Neither Finale nor Sibelius will cut it. The ad hoc solutions all fail for my purposes.
The idea was, within a shifting metric landscape, to create ‘hypermeasures’ (groups of a couple bars) which are then divided into even groups of say three or four, to which then tuplets can be added. So, for example: I have three bars 2/4, 5/4, 6/4. This adds up to 13 beats. I divide these 13 beats into 4 equal parts of 13 semiquavers. I apply a rhythmic process over the top, say a gradual slowing down, comprising of tuplets: 15/13, 14/13, 13/13, 12/13. Two out of the four of these will cross barlines. I also want to have the freedom to add nested tuplets, which may or may not also cross the barlines.
Ok. So that’s pretty much majorly complex, and I probably wouldn’t do it in quite that way when writing for a human performer. But one could think of simpler variants that would also be equally unrealisable in the current software (things like you often see in Finnissy scores).
For this piece – since I need to compose it in Finale for the sake of exporting the midi file to Logic, and since my RSI still hasn’t fully cleared up – I’ve decided I’ll take on other strategies, and abandon this idea of tuplets across bars. The solutions will involve finding other rhythmic elements, processes, structures, which can provide unity across the ‘hypermeasures’ and obscure individual bar lines (unless it’s desired that they be clearly presented). These will mostly rely on creating grouping patterns and/or processes that cut across barlines even if the basic subdivision rests on the length of the bar, or on the basic beat. I think it should still be sufficient for the purposes of the piece. In future pieces I’ll come back to the problem, but it might be that I hand-write the score, or come up with a workable ‘fake’ that looks right, but may not produce an accurate midi (since that won’t be necessary).
But why all of this? Why try for tuplets across barlines – why hypermeasures – in the first place?
To me the answer is counterpoint, and the role that the measure plays, what it represents, in complex, stratified counterpoint.
To me the measure represents some kind of formal collective structure, tendentially uniting all the lines of a counterpoint. I say ‘tendencially’, since there are many strategies by which an individual might thwart the collective structure. And in fact, in a world as alienated and unreconciled as ours, the measure is not really in a position to represent the ‘form of the content’, but instead an external form imposed upon the content and operated separately to it (in this sense it’s something like the ‘State’ in politics, from which we have been alienated). This is all the more the case when you have many voices that wish to have relative freedom from one another. The political economy of it all is discussion for another time, but for now let’s take that statement as a given, and go on to the musical-technical stuff.
So, for me, it’s not a matter of creating a notation and conception of collective rhythm that is totally metreless, where each line pursues its own processes without reference to a mediating collective structure. Nor is it a matter of having a single metric structure (even if composed of multiple processes, as in Ferneyhough) and using this as the basic ‘frame’ for all the lines in the texture (as is common in Ferneyhough), and slavishly following it. The former to my thinking has a tendency towards sheer polyphony; the latter towards a constriction of polyphony to a sequence of gestures.
So, my strategy is to unite the two. Firstly, construct a grid of shifting metrics. Sufficiently tendential to give a section a global process and direction (an obvious example would be a progressive shortening of metre, which is what I do in the third section of Si el clima). Next is to come up with a series of strategies for relating (or not relating) to this metric framework.
I mentioned this in an earlier post on a new day in the desert, and outlined a couple of strategies. Here’s a quick recap and refinement. Of course there’s infinite nuance between these, and probably a good many that could be invented beyond this list…
- Metre defining: A line that tends to outline both the downbeat of the measure, as well as the essential grouping of the basic beats of the bar. At this stage I’m pretty keen to always have a line more or less taking up this job in an ensemble texture, since then everyone else has a clearer idea of where they are in their part! (Particularly important if you don’t have a conductor). Of course the danger is that it can become a bit claustrophobic sticking to this, but there are ways around that…
- Metrically defined: The length of the bar is taken as the starting point, and as a given. It is then subdivided in ways that don’t outline the basic groupings therein, but instead present an overlapping rhythmic structure. Proceeding from bar to bar of course there can be processes of speeding up, slowing down, and so on.
- Hyper-metrically defined: A number of bars (probably maximum only a handful, before you start to flip over into the next category) are conceived of as a ‘hyper-metre’ and are subdivided in such a way that they gain a certain unity. This can be done via tuplets that cross barlines (fuck you Finale), or by groupings that cross barlines. Likewise there can be a variety of rhythmic processes both within the hyper-measure, and from one hyper-measure to another.
- Non-metric (or counter-metric): The rhythmic logic of the line is devised with disregard for the written metre. So it could just be an additive/subtractive rhythmic process that unfolds across many, many measures, and perhaps incidentally lines up with barlines here and there, but not in any rigorously controlled way. Or it could be a metric identity of its own (a ‘counter-metre’). This is of course easiest to write if it sticks to the basic subdivisions (quavers, semiquavers, etc), since you won’t get tuplets crossing barlines then. But there’s no reason (apart from software) why this metric function couldn’t itself take up a complex structure of tuplets.
(To any and all of these other rhythmic filters can be of course added, including layers of deletions, grouping, accenting (including via ‘extended techniques’), and tuplets/nested tuplets). (Likewise, other parameters will have an impact on the rhythmic identity of the line, for instance, repeating a pitch at a certain register in a certain way will have a perceptible metric impact: this could either support or undermine the metre or metric function outlined. Dynamics also can play a huge role).
So then, as a composer, in terms of counterpoint – which is a matter of relations over time, not just of mere super-position – you have a series of local parameters to deal with (i.e. how to morph one kind of metric function into another), and a global parameter to deal with (i.e. to what extent are all the voices ‘governed’ by the metre, and to what extent are they ‘independent’ of it).
For me this is the vision of the ability to move from a totally stratified ‘Ivesian’ vision, to a total gestural unity across lines, and everything in-between, including various metric alliances and disputes between parts in the overall texture.
For me this is but one way of creating relations between lines that takes us from ‘polyphony’ to ‘counterpoint’ in the absence of thematism.