Lynne Kiesling
I am a big fan of English renaissance choral music, particularly sacred polyphony from Tallis and Byrd (and stretching back to Taverner, but he’s not as distinctively polyphonic). One of the best ensembles performing such music is Stile Antico, a group of 13 British singers who do an outstanding job with this music, and whose recordings I have recommended here before. Especially at this time of year, their music really resonates and adds joy and beauty to life.
A couple of weeks ago we got to hear Stile Antico perform live in Milwaukee: Thomas Tallis’ Puer Natus Est mass interspersed with pieces from Byrd, White, and Taverner. The music was gorgeous, the voices delightful, and the artists charming and gracious.
But what really struck me was their method of decentralized coordination. Typically when we think of musical performance beyond, say, a chamber quintet, coordination involves hierarchy in the form of a conductor, to “keep everyone on the same page”. The larger the number of performers doing different things, the harder to coordinate, and therefore the greater need for a conductor … right?
Not so in this case. 13 singers, each with a particular part, bringing a distinctive element to the work. But in some ways the music is simultaneously so lush and yet so spare that if their timing is off, the beauty of the result is diminished. 13 singers with no conductor, and they coordinate by taking their visual and verbal cues from each other in a dynamic and evolutionary manner. This is a vivid example of decentralized coordination.
Of course the goal is harmony (in the general sense). If each individual acts and reacts to the actions of the other individuals in a way that produces a harmonious outcome, that’s beauty. And it’s an emergent outcome; each has his or her own score and acts accordingly, adapting to the actions of the others in a way that creates emergent harmony.
The music metaphor illustrates achieving emergent order through decentralized coordination, and it’s a metaphor for social cooperation too. Adam Smith employs the harmony metaphor for social cooperation in The Theory of Moral Sentiments, in which he invokes harmony as a desirable outcome of social interaction repeatedly (and refers to the music metaphor directly in the last reference). Note the emphasis on harmony as distinct from uniformity — each individual brings personal, private, heterogeneous features to social interaction (whether musical or economic), and they are not the same, not uniform. Each has an incentive, a desire to coordinate, to harmonize; in music it’s finding the complementary notes, in social systems it’s grounded in our innate desire for sympathy and mutual sympathy, according to Smith. Each individual brings something different to the party/performance/market. The most beautiful and sublime outcomes emerge when each acts on its individual traits with a view toward creating harmony and sympathy. And it does not necessarily require the top-down imposition of control or system-wide hierarchy, but can be achieved through decentralized coordination.
Of course there are limits to applying the music metaphor to institutional design and social cooperation, such as the scale/number of actors. But it reminds us of the possibility of cooperation and harmony through decentralized coordination, without the need for imposed system-level control.
Terrific point, but I think jazz is an even better analogy, especially where there is simultanous improvisation among all the members of the group, individual but integrated–and marvelous!
Exactly Lynne. Let’s take two other examples:
1. Jazz improvisation. Not quite harmony there, but there is the same decentralized coordination, but what’s cool about this one is that there are formal and informal norms that guide the process. The players know the chord progressions and the basic language of jazz, from which they can create both order and uniqueness. Often they produce something unintentionally beautiful. Markets, no?
2. One of our favorite mutual interests: hockey. Hockey players are jazz quintets on ice. There are very few set plays in hockey (power plays being one). Most of what happens is improvised, but again, the players know each other and they understand the their roles and the rules of the game in a way that produces beautiful harmony without centralized coordination. They respond to each other’s actions through mutual adjustment, just as in the market.
Thanks to you both for mentioning jazz, and I completely agree. Mark, sorry your comment got hung up in moderation, but now that WP has your email address it shouldn’t happen again.
You both might be interested in this TED talk from my favorite vibraphonist, Stefon Harris, in which he argues that there is no such thing as a mistake in jazz:
http://www.ted.com/talks/stefon_harris_there_are_no_mistakes_on_the_bandstand.html
His argument may be particularly interesting from an Austrian epistemological perspective.
Steve, I would add soccer to hockey, and the fluidity, adaptation, and emergent order nature of the game is why it’s called “the beautiful game”. Basketball fans might argue for including it with hockey and soccer, but I am not one, so I won’t :-).
Pingback: Holiday music encore « Knowledge Problem
Very nice. “Stile Antico” provides a nice Pandora stream, too.
“The larger the number of performers doing different things, the harder to coordinate, and therefore the greater need for a conductor … right?”
Not so much “right” or “wrong” as a generalization that tends to be true but with the threshold being soft and depending on the circumstances and skill level of, in this case, the singers. Please do observe that the Stile Antico singers in the followup post are standing in a semicircle, and are occupying a tight semicircular space. This makes it (relatively) easy to see each other, and to hear each other with little delay in the sound. In addition, the texture of the Palestrina makes it a little simpler to coordinate than might be the case for other pieces. I’ve done a fair bit of this sort of thing in groups from about eight to about twelve, and it can certainly be done under the right conditions. In fact, “chamber choirs” (which by a common technical definition do not use a conductor) often number somewhere around twelve.
Also observe that in the other videos in the followup post, with larger groups occupying more open spaces, a conductor is indeed used. As the group increases in size, and the space becomes more open, and accompaniment is added, one may reach a point where centralized coordination is helpful because everyone is no longer hearing/seeing everyone else adequately enough. But it’s a soft threshold – no bright fiery lights will come on in the sky to say that *now*, conducting will help.
A bottom line might be that decentralized coordination can work well *under the right conditions*, which include appropriate feedback and communication – plus some willingness to yield to a group consensus. Then there’s the time factor – it may not be so at Stile Antico’s very considerable skill level, but on occasion it may be more expedient in terms of rehearsal time (for which read expense) to use a conductor. It may also be expedient as a function of skill; a conductor may be able to help a group who wouldn’t coordinate well enough on their own. And with some groups, a conductor may save great gouts of time that would otherwise be expended on unsettlable arguments about, let’s say, performance style.
As usual in this life, there is “diversity” and there are often tradeoffs… one’s mileage may vary…
Great post. Although I am a long-time choral singer and equally long time familiar with hierarchical coordination vs. spontaneous order (featured in Ch 1 of my textbook) I had never thought about the two together. What I would add to this discussion is that conductor vs. spontaneous order in choral singing is definitely not either or. Yes, the conductor does have a central control role. Examples of places where the conductor is important: Setting the overall interpretation as the chorus develops the piece, setting the tempo during the performance, making on-the-spot modifications to tempo and dynamics to suit a change of acoustics as the group moves from one hall to another (esp. from practice room to performance hall). With the groups of 30-100 or so with which I have mostly sung, those things do take central control.
However, other crucial things require spontaneous order to get right. The most important is tuning and pitch. The conductor can do little there. The singer have to listen to one another and adjust so that the sound of the group is a harmonious whole. That is one of the biggest differences between the sound of a professional group and amateurs. That is why a good group often mixes singers around in small groups rather than placing them in bloc sections. Also why a good conductor will “voice match” placement of singers so that they can best hear one another and interact.
Then there is the interesting point of interaction with audience. A really good performance requires a two-way communication between the singers and the audience, with the singers reading the feedback they get from the audience. Interestingly, the conductor is almost useless there because she has her back to the audience!
Good examples of the subtleties of the coordination problem, which in the real world–things like a factory floor or university classroom–are always a complex mix of hierarchical and spontaneous coordination methods.