On 2014-07-11, 18:33 , Andrew Smith wrote:
> I think there's something that could be said about the language used to describe an algorithm that is used to make music. The programming languages used to construct an algorithmic piece make certain assumptions about the resulting music (and the structure it should take). Just as certain programming paradigms lend themselves better to certain applications, certain musical languages lend themselves better to certain musical ideas and structures.
I emphatically agree - and this is rather independent of the underlying
topic (music algorithms).
Whether you use a low-level imperative paradigm, some high-level OO
ideas (like Squeak, not imperative-on-steroids like Java), functional
(Haskell) or logical (Prolog), your mental model of how the processing
of data changes drastically. Or at least it does once you become
sufficiently adept at the new paradigm.
It seems that what one should really be exploring is the question of
"What are you really trying to say" when making 'algorithmic music'.
What I mean is, when describing what you have done to colleagues, what
vocabulary do you use? The human-computer vocabulary is boring, what is
interesting is the human-human vocabulary. That is the one which should
ultimately be captured in 'code'.
The really fun part is that, just like 'raw' programming, there is a lot
of room for style: there is no one all-encompassing programming
paradigm, and there is unlikely to be a single "language of music" which
works well for all situations.
> I'm not thinking in the sense of a particular language, but more like a comparative study of the notion of "instrument" or "voice" across algorithmic music languages. SuperCollider has SynthDef, Max/MSP has the poly~ situation, CSound has an orchestra, and there's more out there I don't know about. The language contains an abstraction and an assumption about music, and it's interesting to see how different languages interpret these common structural ideas.
Exactly. The level of abstractions can be radically different, which
leads to different kinds of music being easy/hard to 'say'. There is a
definite tension between being able to define every single note
precisely, and being able to do orchestra-level changes. But that is
obvious from offline music: the conductor's "language" [which is mostly
gestural] has nothing in common with 'musical notation' used by each
player for their part.
The point is that a piece of music has different levels, and these
levels have different natural vocabularies. The division into 'levels'
is not unique, and even within a level, one can use radically different
languages that focus on different aspects.
To me, it all comes down to Kolmogorov Complexity or, more precisely,
Minimum Description Length: you pick the language which eases the
description of the piece you have in mind.
Jacques
--
Read the whole topic here: livecode:
http://lurk.org/r/topic/1JCry3p86k1msq2lwNCAm5
To leave livecode, email livecode_at_group.lurk.org with the following email subject: unsubscribe
Received on Sun Jul 13 2014 - 16:36:05 BST