Wednesday, September 10, 2008

Posthumous Casting

A pair of recent articles in the Frankfurter Allgemeine Zeitung (which is one of those odd German institutions, with conservative news and editorial pages, which can be ignored, and a liberal Feuilleton section, which is often well worth reading) dealt with the tricky issues of what to do with the creative remains of an artist.

The first article celebrates the completion of D.E. Sattler's Frankfurt edition of the complete works of the poet Hölderlin, a project of some 33 years, and one which was initially greeted with great skepticism and critique because it took the notion of "complete" so literally, decyphering every possible alteration found in or alternative reading of the manuscripts (in producing the edition, Sattler so mastered Hölderlin penmanship that he was able to perform as a live calligrapher in Walter Zimmermann's opera Hyperion). Advocates of traditional editions found the project absurd, faulting Stattler for failing to execute proper editorial judgement and to define, as strongly as possible, the most "authentic" or "authoritative" text. Instead, Sattler offered not only his best judgement on the texts, but also supplied every possible objection to his judgement, thus ultimately refusing to cast any poem in a canonical form. Now, however, the Frankfurt edition has established itself as a valid mode of edition-making and similar editions of the works of Kleist, Trakl, Kafka and Walser are in progress.

The second article (not yet online) reports from a symposium Arp-Museum in Remagen-Rolandseck on "posthumous casting", the practice through which new bronzes are cast from the original forms of deceased artists, and frequently a rather lucrative business for the estate and the foundry. The question of the "authority" of a bronze cast without the express consent of the artist is a controversial one, both legally and culturally.

In music, the question of the opus posthumous is rather less heated, probably because music, without much in the way of marketable physical objects, is playing for pennies rather than the pounds which chance visual artworks. Nevertheless, there are similar controversies surrounding the estates of composers. I have recently been close to a few cases in which it has been difficult to locate an heir or the control over the works was unclear, one case in which a composer insisted that a large number of pieces, mostly juvenalia, were not to be played, another case in which a composer insisted that every manuscript be destroyed, and several instances in which composers left behind unfinished works. Attempts to complete unfinished works seem to have a special attraction for specialists and the public alike, from Mozart, Beethoven and Schubert, to Debussy, Puccini, Skryabin, Ives, Mahler, Berg, and Stravinsky, even John Cage.

But those example are limited to a handful of major works left unfinished at the composer's death. With more recent works, however, composed with the use of explicit algorithms, the possibilities begin to multiply. A number of my own works, for example, were written according to a plan that could be described in a rather brief algorithm and with the substitution of a few variables into the algorithm, one could rapidly generate a huge number of new, variant — and sometimes wildly so — works, and the composer need no longer be present for this to happen. This situation bothers me enough that I have made a point of not keeping my sketches for many pieces, making my steps much harder to trace. I particularly admire the composer Andrew Culver, who assisted John Cage in much of his late work, often by writing computer programs for the realization of chance operations. In principle, Culver had access to all the resources which would have been required to, in effect, to cast an indefinite number of "new works by John Cage", but he refrained from doing so, keeping with Cage's own ethical and aesthetic principle of not indiscriminately creating new pieces by repeating a compositional procedure with new chance-derived results.


Samuel Vriezen said...

If it would involve new values for some variables in the algorithm, that would IMO never constitute a new piece by the same composer. Often in such pieces, the choice of variables is the hardest part: the algorithm could be merely the idea of the piece, but the variables give it its dimensions.

OTOH if you could just run the program again to get a new version, and if this is exactly what the original composer did, I would not have a problem of principles with that. I'd just wonder, why?

You could make a website where a performer would just have to enter the ensemble combination he wants a piece for and a duration and the piece comes out. Like a Xenakis ST-site or something. Do sites like these perhaps already exist?

Paul H. Muller said...

Interesting. This also takes copyright issues to a whole new level. It is difficult enough to manage the rights to the performable end product - a score, and now it may be necessary to protect the process as well!

I guess the jazz musicians are way ahead of the curve - they keep the algorithm in their head and only execute it when paid.

It may be that written notation itself was the first musical algorithm - our orchestra director is always saying that we have to stop worrying about all that stuff on the page at some point and start playing music.

As music becomes more and more realized electronically and performed less by real people, the concept of what constitutes composing will have to change.