Well, it all depends really.

No, really, it does, not only from composer to composer but from piece to piece. I remember playing Time and Motion Study I to Brian for the first time and him gently but firmly beating out the quavers to make sure I was playing the 10s - 11s - 12s - 13s - 14s accurately. On the other hand he's famously resistant to the idea of having his scores printed in 'time-space' as people like Richard do so that the rhythms are represented not just in the ratios but in the spacing of the notes on the page - which is a notational tool one migt expect a composer only concerned with exactitude to jump at.

I don't generally often find such things 'unrealisable' as such - indeed they might not be playable so that a listener might think (to take a reasonably famous example) 'ah, yes, that's 70:67', but neither would even something as straightforward as 6:5 be playable so that a listener could precisely identify it numerically under most circumstances. I've very rarely met such examples in conducted music where it's completely unfeasible simply to play it in time so that one arrives at the next downbeat (or wherever) with the conductor, for example.

Some composers (Chris Dench for example) use such things to point out structural allegiances across a work; there I suppose it's a matter of leaving the scaffolding visible so that the performer isn't deprived of potentially useful information.

Of course the issue could fill several books. But don't worry, not from me it won't.

Very different things in either case (and different again in the work of other composers). In Lemma-Icon-Epigram, the rhythms do arise from some deep-seated structural working, and sometimes linear or other rhythmic progressions are apparent on the surface. In this sense they have a certain 'integrative' role, whereas in the later Opus Contra Naturam, which has significantly more complex tuplet patterns (four levels of nested tuplets in the first bar), they seem more to serve a certain 'defamiliarisation' role, steering the performer away from more regular or familiar patterns as far as possible (of course there are some more meaningfully audible rhythmic relationships in the later piece, and some more defamiliarising effects in the earlier one, it's a question of degree).

English Country-Tunes: the extant sketches for that work aren't that comprehensive in terms of illuminating the compositional processes, so I can't say exactly why they emerged the way they did (if that is necessarily important). Whilst Finnissy does sometimes uses audible rhythmic processes in his more 'a-metric' works (for example in the work Fast Dances, Slow Dances), overall they more fulfill the role of a type of 'localised rhythmic coloration', to be worked on carefully but then played quite freely in concert (certainly he plays them extremely freely himself, though he wouldn't necessarily advocate his own approach to others). In those rhythmic processes I've identified from sketches, there is usually a level of randomness involved, as a means of producing a certain generalised effect. Nothing like as extensively as in, say, Cage's Music of Changes, though, which has rhythms every bit as complex as the aforementioned works, if not more so.

Thank you Ian and Oliver-sort of as I thought, but I still can't get my head around the notion of extreme precision of notation being the means to guarantee a free approach in performance. I think I do the same as Oliver does, which seems to satisfy others but I hate being in a position where the notion of 'getting it right' seems to be a movable feast.

Well, to paraphrase Lachenmann (in a different context), maybe 'complex notation produces denial of habit'?

I very much like it myself. After all it's a situation which applies in most of the repertoire, I would say; there's a certain way rhythms are treated which is responsible for the 'feel' of so much music from Baroque overtures and notes inégales to Strauss waltzes, Mahlerian rubato and swing. (None of which I am any good at.)

I'm told it's rare for notes which are equal on the page and which take part in any kind of accent scheme to be really metronomically equal - the urge to give the accentuation just a little bit of presence in the note spacing is apparently hard to resist. Anyone who plays in ensemble will probably have had some kind of discussion of whether to play on the 'front' or the 'back' of the beat even when there's no conductor... most such things can be as complex (or as complicated) as you want to make them.

The examples you give, however, rely upon the inalienable sense of pulse which animates not only music but life itself. I'm sometimes at a loss with really complex music how one retains this without effectively rewriting the music, which certainly seems to be done by many of its most prominent exponents-admirably not in Ian's case!

Thank goodness you actually said it Ian! Has anybody ever admitted to trying to produce a complete meltdown upon sightreading?

(I ask, purely as a hammer-chewer)