December 10, 2009

Settling the score

Incoherent language produced by people with Wernicke's aphasia indicates a serious, frustrating, and sad cognitive impairment.  Unless, that is, such people happen to be philosophers of music (not to be confused with musicologists or historians of music).  In that case, incoherence seems to be a professional requirement.  For my money, the most amusingly extravagant offerings from such philosophers concern the nature of the score and its relation to the musical work associated with it.

Some - who call themselves Platonists - assert that the score describes an abstract structure, which exists timelessly and immutably among other 'ideal Forms' (such as numbers, functions, and sets) in Plato's Heaven.  On this view performances of a musical work are only more or less approximate physical realizations of that abstract structure.  To grasp the work itself, therefore, one must study the score: the only direct and uncorrupted path to the musical work.
     Of course, Platonists realize that, thus construed, musical works cannot possibly be created by humans precisely because abstract objects (by their very conception) are causally disconnected from the physical world where composers' brains happen to dwell.  In particular, nothing that the composer might have done in his life - from such 'involuntary' actions as electrical signals in his brain to such 'voluntary' ones as putting ink marks on paper - could have created an abstract object.  Consequently, to speak, for example, of Op.111 as Beethoven's sonata is to credit Beethoven (at most) with a discovery of a certain abstract structure, and this does not entail any attributions of creative ownership.  But Platonists, it seems, are not at all nostalgic about such attributions, and they happily settle for discovery of abstract structures as  the principal task of every composer. 
      Where the Platonist comes to grief, however, is the question about how the flesh-and-blood Beethoven could have possibly surveyed the otherworldly realm of abstract objects, so as to 'discover' Op.111.  Because Beethoven could not have peeped into Plato's Heaven with a telescope (even if he had access to one), or with any other physical gadget, the Platonist has no choice but to postulate an extra-sensory, non-physical cognitive access to the realm of abstract objects - the kind of access, he assures us, that is routinely enjoyed by composers and mathematicians in the course of their work.
      Obviously no details concerning this 'special access' to Plato's Heaven will ever be known to cognitive science, neuroscience, biology, or physics - none of which are equipped to deal with interactions between physical and non-physical objects.  All we will ever get in support of this utterly mysterious access is the Platonist's desperate need to bridge the ontological gap separating musical works in Plato's Heaven from the composers here on Earth.
    A curious outsider who gets this far through the Platonist's story will most likely mumble:

--   Fuck me!  I can get more entertaining mumbo-jumbo magic bullshit from Harry Potter novels!,

and swear never to open another philosophy book for as long as he lives.
Then there are those 'down to earth' fellas called nominalists.  Nominalists resolutely reject the existence of abstract objects, limit reality to whatever can be found in the physical world, and insist that the the score itself is the musical work.  Once again, to grasp a musical work one must study the score: with the two being literally identical, there is no more direct and transparent path to understanding a piece of music.
     The glitch with the nominalist's story, however, is that performances (qua physical realizations of the score) must be note-perfect to qualify as performances of the work in question.  One missed or wrong note, one incorrect accent - and the performance is no longer of the work embodied by the score.  That this is so can be seen from the simple fact that any performance can be in principle transcribed into a score (as the young Mozart once did from memory after hearing a mass by Palestrina).  And if the original and the transcribed scores are not identical, neither are the musical works embodied by these two scores.
        A moment of reflection will convince anyone familiar with the realities of music-making that, according to the nominalist, virtually every advertised performance of, say, Beethoven's Op.111 is actually a performance of some other musical work!  After all, no pianist (with the possible exception of Michelangeli and Pollini in their prime) can deliver a score-perfect performance of Op.111 at each and every concert where he or she plays this work.  And what holds for pianists holds for other instrumentalists, as well as for orchestral performances.  Toscanini - the supposedly legendary literalist score-worshiper - did not really conduct Beethoven symphonies (because he slightly changed the orchestration).  Furtwangler never conducted any work by any known composer at all (because his performances were full of false entries, wrong notes, and unmarked changes of tempo).  And the great 'Chopinist' Cortot never even recorded anything by Chopin (because he missed enough notes in every recording to make up for another one).

No wonder every time I describe the nominalist's story to musicians, what I hear back reminds me of a string quartet by Helmut Lachenmann: a lot of hissing, growling, gnashing of teeth, and other audible manifestations of the primordial homicidal instinct still lurking in the deep recesses of the human brain.

As might be expected, the correct story regarding the nature of the score is philosophically boring because it is devoid of the lunacy expected from 'deep philosophical reflections'.  A score is a specification of instructions (not different in kind from a recipe or a blueprint) for creating a sequence of physical events (temporally organized sounds of specified pitch, amplitude, timbre, etc.).  The specified sequence of physical events may have an abstract formal structure, of course, but this structure is entirely secondary to the physical events (just as the mirror image of my body is entirely secondary to my body itself).  And the composer surely need not be aware of such a structure in order to produce a score (just as the father who puts together a swing in the backyard need not be aware of differential equations specifying the abstract structure of pendulum-like periodic motion.)

A performance, then, is just carrying out the instructions specified by the score, and this process can never be absolutely perfect, even if the score has the precision of a computer program (which it never does outside computer music).  After all, even in computation - the most precise example of "carrying out the instructions" - the sequence of physical events (at the most basic hardware level) specified by the program is never perfectly deterministic because of random transistor failures and other kinds of random interferences.  And when a program is complicated enough, it is impossible to be absolutely certain even about the kind of physical operations specified by this program.  (Because program verification will require even more complicated programs whose internal consistency will be open to doubt - a logical limitation established by the celebrated incompleteness theorems of Kurt Goedel.)  Once it is clear that every physical process of carrying out reasonably complicated instructions is essentially probabilistic in nature (and, in music, also fuzzy, because of inherently imprecise instructions in most scores), we can exhale with relief and accept performances by Cortot, Furtwangler, Edwin Fischer and other error-prone musicians as indeed performances of the advertised musical works.

Finally, the musical work itself is just the sum total of performances according to the score, where 'performance' is understood to include what happens in 'the mind's ear' of those few who can by-pass the services of musicians and 'hear' the music by reading the score.  Studying the score without hearing the whole work in the mind's ear (e.g., following just the cellos part) allows you to understand the instructions in greater detail, but gives you no access to the work itself.  You can't satisfy hunger by studying a recipe; you can't fly from New York to LA by studying the blueprints for Boeing 737, and you sure as hell cannot grasp (understand, appreciate) a musical work by studying the instructions specified in the score.  What one needs instead is to hear a sufficiently large number of diverse performances of that score.  It is by hearing, say, Beethoven's Fifth Symphony in performances by Nikish, Toscanini, Furtwangler, Szell, Norrington, and Eotvos that will give one a reasonably comprehensive grasp on that symphony as a work of art. 

If you got this far, you will understand why I strongly disagree with my friend (and fellow music blogger) who accepts the widely circulated claim that Debussy's Jeux exerted a strong musical influence on Stockhausen's so-called moment-form compositions.  My friend mentions the 60 (or so) tempo changes in Jeux, its juxtapositions of "here and now" sections whose content is not related by continual development of some basic thematic material, and then exclaims: "But that's just what Stockhausen did with his moment-form pieces!".  To me, on the other hand, this score-derived conclusion is analogous to saying that Ford Taurus and Boeing 737 have something 'transportationally' important in common, because the blueprints for each specify certain identical welding techniques.  In other words, I tried to convince my friend (so far unsuccessfully) that similarities at the level of instructions - whether in recipes, blueprints, or musical scores - need not translate into meaningful similarities at the level of food, mechanical devices, or musical works respectively.

No comments: