Welcome to
      Home    Listen    Site Index    BLOG  

027 < >

Taking a Chance with the Music: Mozart's Dice Game


In 1787, the same year he wrote his opera "Don Giovanni" and a couple dozen other masterpieces, Mozart took some time out for a little parlor game. He wrote nearly 300 measures of music for a little minuet. But this was only the raw material: In order to figure out which measures would actually make the cut, and in which order, the composer had to roll dice, and for each number rolled, there was a corresponding measure of music. Anybody could "write" music under this scheme.

Mozart didn't surrender his will entirely. There is a specific, set number of measures for the entire composition (16 in the minuet, for example, and 16 in the accompanying trio, a contrasting section which was customary before returning to the minuet itself). Each measure has to be "rolled" in sequence, and has its own column of possibilities. This way, the "grammar" of the piece is still preserved. For example, each sentence in English needs a verb in order to be complete. If Mozart's game were to allow you to compose sentences rather than minuets, the first column would feature a bunch of nouns; then when it was time for the verb, you would only be able to choose from a list of verbs. The next roll would feature a list of direct objects. That way, the sentence would always make some kind of grammatical sense, and none of the necessary elements would be missing. In Mozart's case, the important thing for a minuet to do is to move from the tonic to the dominant and then back again, and that is carefully controlled by making every possible measure do the same thing musically at each of the important structural points, even though the surface possibilities for how you do it are quite varied. There may be 300 measures, but, at any point in time, you are only allowed 11 choices, and they exist only for a particular measure (say, measure 5 has these 11, and measure 6 has these other 11, and so on).

But, one imagines, part of the fun will be in coming up with silly combinations, much like those children's books where each picture of an animal is divided into three independent parts and by flipping those parts of pages individually you can wind up with the head of an alligator on the body of a giraffe.

When I taught this game one year to my piano students, before composing minutes, we had fun rolling the dice to create silly sentences from my list of articles, nouns, verbs, and so on, such as "the banana went to the store" or "the toothbrush had a picnic." On a structural level, these make sense. They employ the appropriate parts of speech in the correct places. But they are plainly nuts. This is the kind of silly Mozart enjoyed, and his game is a useful way to understand musical architecture while being absurd at the same time.

It was probably also a nice respite for Mozart from having to make all of those important compositional decisions. In the 20th century, some composers made a philosophy out of divorcing their will from the compositional process. John Cage often composed by throwing dice and notating the results (with far less control over the output). Mozart would never have considered such a possibility. But he loved to have a little fun.

There are scads of pages devoted to this game on the web. I've listed only a few. Some of them enable you to actually play the game yourself with a few clicks of your mouse.



When I first wrote this article 500 years ago, you couldn't get the score of this piece free on the International Score Library Project. Now you can. Here.

http://donaldsauter.com/mozart-dice-game.htm   This is just a fun read. And if you are into (ancient) computer programming, you can help yourself to code that will allow you to recreate the game without the need to roll actual dice.

Mozart's Musikalisches W�rfelspiel John Chuang's site allows you to listen to all 272 measures of Mozart's original (although there is no musical notation involved) and has links to other sites.

Mozart Dice Game Most of the sites I've found are interested in the math involved, or the programming. This site shows you the computer programming code involved in creating the kind of games seen above.

That none of these sites were created by (historically aware) musicians is evident from the musical errors throughout. The last site in particular insists on referring to Mozart's minuet as a waltz. Apparently, this error can be traced back to the original publisher, who probably figured the average non-Royal customer wouldn't want something known as a "Minuet," that old-fashioned dance the fancy-pants royalty used to do. None of the above sites takes the customary repeat of each half of the minuet before going on, which changes the overall effect as well.

Mozart Interactive The most visually appealing version of the game although the translated instructions do not make much sense. You can see the die being cast and the actual measures being assembled from the available choices. You'll have to wait until all the measures for the minuet section (no trio here, either) have been assembled in order to download the results, something I haven't yet been able to do successfully due to "internal configuration errors," which is a bit disappointing.

Math Trek: Mozart's Melody Machine, Science News Online, Sept. 1, 2001 a great resource with many links for those wanting information from a variety of angles.

Mozart's Musical Dice Game from Carousel Publications Ltd  Musicians and educators...Wondering how to get a hold of the actual game? For 20 bucks you can own the score, the instructions, and a pair of dice. Not surprisingly, the publisher's site is quite certain that Mozart really composed this game, despite the doubts of many others. Most people waffle, suggesting that there is no real evidence that Mozart actually authored the game, but then stating that "no real objections" have come forth from musicologists.  In an accompanying essay, A Dr. Willis Wager states with certainty a number of things that he doesn't trouble himself to back up with source material, including this gem: "Mozart was still fairly early in his career and not quite so widely known as he was to be later." In 1787? He died in 1791! How early are we talking about here?  And he claims it is listed in the Koechel catalogue of Mozart's works as 516f, which DOES date it to 1787, but so far (and I've only seen an online copy of the catalogue) I haven't been able to find any entry for a 516f. hmmmm.....


comments powered by Disqus