Perotin and John Cage

Or, rather, the problems of defining classical music, the subject of my last book riff, and many comments. Including some very passionate ones from a man I greatly respect (hi, Michael), who longs with all his heart for a definition of classical music that’s based on our current classical music culture.

Other commenters have correctly noted ambiguities in the definition I proposed (I specified that classical music is the music of the European tradition, and that it’s written out in advance by a composer). Broadway musicals, for instance. Wouldn’t they be classical music, because they’re written out in advance?

Probably not, I think, because often the composer just writes the songs (melody and harmony), while someone else writes the details of the accompaniment, does the orchestration, and composes the overture and dance music. That doesn’t sound like classical music to me.

Still. Definitions are tricky. And, as I keep saying in response to the comments, leaky. There will always be exceptions.

What’s helpful here, I think, is to look at existing dictionary definitions of other things. Definitions of “dog,” for instance. These vary, as various dictionaries enumerate various physical characteristics of dogs. But no “dog” definition I’ve seen talks about dog culture, how adorable dogs are, how smart, how loyal, how deeply loved, or whatever. (I love dogs myself.)

Or “mammal.” Here every definition I’ve seen says that mammals are viviparous, but notes the exception every child can name. But the definitions also say that mammals have hair or fur, without stopping to note that hairless mammals exist. (There are hairless cats, for instance.)

So here we see the difference between a dictionary definition and an encyclopedia entry. Dictionary definitions (which are what I had in mind when I sketched my own version) are typically concise. They can note major exceptions to what they say, but not rare or subtle ones. And they may not go into what people feel about the thing being defined.

An encyclopedia article works differently. One on classical music might start with an objective definition of what classical music is (see below for my revised attempt). And then it might note exceptions. Some classical works — some contemporary pieces — aren’t fully notated in advance. Broadway musicals live in a limbo zone, resembling classical music in some ways, not resembling it in others.

Much classical music performance today, the entry could go on to say, takes place inside the aura of classical music culture, which could be sympathetically described. But this culture is relatively new historically, the entry could then add, wasn’t in play during the age when many of the great composers lived, and isn’t found in all parts of the classical music world today. (New music, for instance.)

In an encyclopedia entry, we’d have time to stretch, consider nuances, state exceptions, talk about how people feel about classical music, all to our heart’s content. A good entry, I’d love to think, might even note how much people like me dislike the traditional classical music culture, even while we work to save classical music.

But back to the dictionary definition. This, by necessity, has to be concise. Michael, would you care to try writing one that includes classical music culture? “Classical music: Music in the European tradition, adored by its listeners, and typically listened to in rapt silence.” I think many of us might feel embarrassed to read that, knowing that it tells only one part of the story. Michael, maybe you can do better. I’d love to see that.

I want a definition that encompasses all forms of classical music I know. John Cage may not be central to Michael’s classical music experience, but he’s central to mine. And since mine isn’t exactly the view of a vanishingly small minority (even though we surely are a minority), I don’t think a dictionary definition can ignore us, as if we were hairless cats.

More broadly, I want a definition that includes (a) whatever we know about what happened when Perotin’s works were played, more or less 800 years ago (b) a performance of Brahms by the Chicago Symphony today (surrounded by classical music culture) (c) a performance of Cage’s 4’33”, his famous silent piece (no classical music culture in sight), and (d) the world premiere of the Barber of Seville (also with none of today’s classical music culture in evidence).

I might try to write that definition like this. I’ve modeled it on the Shorter Oxford English Dictionary definition of “mammal,” which also packs several things into one long sentence. And I’ll note that dictionaries finesse well-known exceptions with judicious use of words like “typically.”

Classical music:

Music in a long, and originally European, tradition, starting with Gregorian chant and continuing into the present day; typically composed and notated in advance of performance, and taking either traditional forms (such as symphony, sonata, opera, art song, and string quartet) or new ones (including pieces for mixed ensembles, minimal music, serial music, electronic music, music for composer/performers, and pieces in which a composer specifies only a process of performance, instead of notating all every musical detail). 

Not claiming that’s perfect. But at least it’s reasonably thorough, finesses a few subtleties, and avoids the value judgments and neglect of new music typical of the existing dictionary definitions I quoted in my book riff.

And maybe Michael could add a clause about classical music culture. But I think that would add some contradictions that can’t be easily finessed, starting with this one: The classical music culture we know now just didn’t exist throughout most of the long tradition this definition refers to.

Share on FacebookTweet about this on TwitterShare on RedditEmail this to someone


  1. says

    One of the big challenges of a definition of “Classical Music” is that the cultural lineage is almost always able to trump the acoustic properties of a work. Kyle Gann presented an interesting example on his blog a few years back, the details of which I don’t recall but which serves as inspiration for the following:

    Program a drum machine to play a standard rock pattern–kick drum anchoring the beat, a snare backbeat, and a hihat pattern. Play that loop for 20 minutes. Is it rock or is it classical? It depends on the intention of the composer and the reception of the piece. If it’s written by a composer who intends to situate it within the classical tradition, it’s classical. If it’s produced by a techno DJ who intends to situate it within a popular music tradition, it’s pop or rock. If it’s written by a DJ but Carnegie hall programs it along side Reich and Glass, it will function as classical music in that context. Similarly, if it’s written by a classical composer but a DJ plays it in a club between DJ Spooky and Squarepusher, it functions as pop music in that context.

    There are a few lessons here:

    1. Modeling our definition after biology only gets us so far, and that’s largely because Horizontal Gene Transfer is exceedingly rare in eukaryotes, and even more rare in multi-celled eukaryotes. That results in relatively cleanly differentiated branches in the tree of life, and makes categorization relatively manageable. Horizontal Meme Transfer, however, happens all the time, and we have to have ways of dealing with it.

    2. Our definition has to accommodate perception of artifacts in addition to characteristics of the artifacts themselves.

    3. Our definition has to accommodate vagueness, in the Sorites Paradox sense. Some artifacts are pretty clearly classical, and some are pretty clearly popular, but some are legitimately somewhere in between.

    4. However difficult it is to make the definition explicit, the fact that I was able to go through that example and come up with plausible solutions for the different scenarios means that there is in fact some underlying heuristic by which we make these judgments. It should be discoverable–although at the same time, different people have different heuristics, and that’s not necessarily because some people are right and some people are wrong.

  2. says

    A quite good definition.

    But why do we have to include medieval music and modern experimental music in the definition of classical music? Maybe because today it is regarded as part of one culture? Personally I don´t like the term «classical music» because it comes with certain connotations (not necessarily for me but for society as a whole) like «old», «conservative», even «calm» and «noble». Like you say in your «common-sense definition», classical music is for most people first and foremost the European tradition from 1600 untill sometime in the 20th century, which constitutes a unified tradition and certain musical values, and which is performed on classical instruments. A composition by Cage would have nothing in common with this, not to mention Alvin Lucier.

    Why do we need one definition to include all this diverse music? I think for example that a group performing medieval music shouldn´t market themselves as a classical music ensemble, but rather as a medieval music ensemble. A composer of experimental electronic music wouldn´t call himself a classical composer, but something else. Maybe also classical ensembles could define themselves more to the point, as period ensembles do like for example The Orchestra of the Age of Enlightenment.

    Modern composers can be said to compose in the tradition of classical music, by composing for the same instruments and even using more or less the same idiom. But even then I wouldn´t call them classical composers, because I feel that classical music is a defined canon of works from a past era.

    The composers of today who write music most similar to the classical tradition are maybe the typical film music composers, like John Williams. The paradox is that somehow this music is not regarded as classical music, I think? While modern music that have musically little in common with the classical tradition other than being written for the same instruments and meant to be performed in a concert hall, is said to be classical.

  3. says

    Your definition is pretty good, certainly an improvement on the dictionary definitions listed in the riff.

    Just one clarification: your definition mentions performance or performers three times. Is it fair to infer that you are requiring that classical music is music that is performed by musicians as opposed to music that is realized electronically?

    You include “electronic music” as a form, but does that assume that a musician is playing/operating an electronic instrument?

    Hate to be picky, but all my PCs want to know… :-)

  4. mclaren says

    Well, to start with, wouldn’t we have to ask which European tradition?

    Not to be needlessly difficult, withal, we can identify at least five major incompatible European traditions, can’t we?

    1) The classical Hellenic tradition of the rhapsode. The text to be sung was written down but the music itself appears to have been improvised, albeit within the limits of the pitches suggested by the fact that Hellenic Greek was a tonal language.

    2) Medieval plainchant. Largely memorized but not improvised. After the introduction of neumes, the music was written out but modified by musica ficta.

    3) Renaissance through late Romantic periods. Fits your description best, with a minor exception for improvised cadenzas and figured bass. Ars Subtilis stretches the boundaries.

    4) Early 20th century to mid-20th-century period (ca. late 1940s). Similar to number 3 but with additions for prepared piano instructions, microtonal accidentals, etc. In other words, number 3 with a few expansions. A handful of exceptions such as the scores for Russolo’s intonarumori stand out.

    5) Post-1950. At this point the European tradition explodes into a kaleidoscopic variety which includes Pierre Schaeffer’s written scores for musique concrete, Music V scores done at Bell Labs and, later, csor format Csound computer music scores. As early as the 1970s, David Behrman and the members of the League of Automatic Composers used code fragments for early KIM-1 microcomputer boards to generate music interactively in real time. The “scores” for these compositions partly involves assembly or FORTH computer code running in real time. The Bell Labs GROOVE system uses scores written in FORTRAN V or C (see, for example, Laurie Spiegel’s “Expanding Universe” from 1975). Contemporary computer scores typically involve MAX or PD visual programming language patches, but more recent computer music involves live programming in which the composer (re)writes the codes that generates the music while the music is playing. The era of graphic scores (examples include Earl Brown’s, the graphic score for Ligeti’s “Artikulation,” Penderecki’s graphic score for “Threnody for the Victims of Hiroshima,” Feldman’s graphic score for various of his pieces) tends to blend into the entirely verbal scores (examples include some of Pauline Oliveros’ Sonic Meditations, certain entirely verbal text scores by Fluxus members, James Tenney’s “postcard” pieces, and some of LaMonte Young’s early one-liner text scores).

    However, the term “score” also expanded ca. 1950 to include fixed systems for generating a particular reproducible sonic result in which the “score” was to a large extent embodied in electronic or computer hardware together with verbal instructions. Such “scores” occur in the Alvin Lucier composition “I am sitting in a room,” David Tudor’s Rain Forest installation series, Trimpin’s various electromechanical installations, and specific circuit topologies such as Ron Kuivila’s practice of passing specific spoken audio phrases through a phase-locked loop circuit. Each of these combos of circuit topology + specified sonic input generates a reproducible musical end result that falls within a fairly narrow range, and musically qualifies as roughly as distinctive and “composed”-sounding as most tape music or computer music. We should also recognize an expansion of the circuit-as-score paradigm in which the circuits generate initial raw material which then gets reordered and layered outside of real time. Louis and Bebe Barron along with David Tudor designed specific analog circuit topologies to embody certain musical gestures which then formed the basis of chronological sections of the resulting music. Combining these predefined circuit topologies with spliced tape tends to blow apart your definition, since at this point the pre-specified musical material has become a soundspace of gestures specified by a particular circuit diagram with a characteristic (but broad) sonic behavior. In the Barrons’ case, the musical section ended when their vacuum tubes overdrove to the point of thermal runaway and burnout. That made the length of individual musical sections dependent on the electronic characteristics of the vacuum tube circuits, including the specific resistances and capacitances and inductances used, together with the mu of the particular audio tubes. Phill Niblock’s use of Pro Tools to layer performed produced by verbal + notated musical scores represents the inverse of this tradition.

    Permit me to suggest that the systematic variation introduced by active computer programs running in real time together with specific musical input largely recaps the European classical tradition specified in Number 2, and is quite different from either the Number 3 or Number 4 European classical traditions. Adroit application of musica ficta to memorized but malleable plainchant melodies modified by troping approaches closely the results of real-time modification of a pre-specified input by an interactive computer program. In each case the musical result involves variants of a unique input within a specified range of variations.

    Electromechanical installations and generative computer programs do not appear to have recognizable precedents in European music history.

    Conlon Nancarrow’s player piano rolls and the MIDI files common today also tend to burst the confines of conventional notation. A MIDI file exists as a set of binary values which specify MIDI note start times and MIDI note lengths, but the MIDI notes do not necessarily correspond to any pitches in the conventional Western tuning, and the MIDI note numbers can trigger sufficiently complex sounds that they may not qualify as recognizable conventional notes in the sense of European musical traditions 1 through 4.

    So your definition holds pretty well for the latter half of 2) (late medieval plainchant notated with neumes) and all of 3) and 4). However, it’s doubtful that your definition works for 1) or much of 5).

    Your definition also gets dicey with much minimalist music (Terry Riley’s “In C” is written out only in the sense that individual melodic fragments get specified together with transition rules, likewise the remarkably sparse score for Dennis Johnson’s “November” from 1959, in which less than a page of chords or dyads get systematically expanded into 3 hours + of music) and with a fair amount of jazz. The notated form of much jazz offers only the starting point for significantly more elaborate live variations. It gets hard to draw the line in these instances between the end result heard in concert and the original notation, which amounts to more of a suggestion than a full score.

    Live coding as defined in the TOPLAP Manifesto and as performed in the LOSS Festival from 2007 would appear to present the greatest challenge to your definition of classical music.

    An alternative definition involves musical structure. Pop music typically consists of an A-B-A-B-A-B structure in which the musical material in A and B repeat more or less exactly, only the lyrics changing; pop music also typically limits itself to pieces under 5 or 6 minutes total. Pop music usually doesn’t conclude with a finale and a cadence, but instead fades out during one of the repetitions. Classical music then would typically encompass pieces longer than 5 minutes, pieces in which there exist more than 2 exactly repeating blocks of identical material in alternation, and pieces in which the structure extends beyond an A-B-A-B…etc. type binary verse-chorus form.

    The extent to which we can talk about pre-Hellenic music in the European musical tradition remains limited by the fragmentary nature of the extant musical scores from the Babylonians, i.e., the Hurrian Hymn, but this apears to embrace an entirely different and sixth European classical tradition.

    God damn, Galen Brown’s comment about should get published as its own article. Got to be the most insightful discussion I’ve ever encountered in a comment on any blog.

    One minor addendum to Brown’s remarks — lateral gene transfer from bacteria and viruses into multicelluar organisms may be much more than common than previously believed. See the article “Widespread Lateral Gene Transfer from Intracellular Bacteria to Multicellular Eukaryotes”, Science, Vol. 317, no. 5845, September 2007, pp. 1753 – 1756.