The Sound of Many Hands Clapping: Group Intentionality

I originally posted this in March 2014. I'm bumping it to the top of the queue as this topic is much on my mind these days.

* * *

In my earlier post on the busy bee brain I quoted some passages from Beethoven’s Anvil in which I discussed synchronized flashing among fireflies. Now I’d like to quote a somewhat longer passage, one about synchronized clapping (pp. 67-68):

You may be familiar with the synchronized clapping that routinely rewards a successful performances—music, drama, circus, etc.—in eastern European communities, but which is less common in western Europe and North America. Z. Néda and colleagues have investigated this phenomenon, recording applause for a number of performances in Romania and Hungary. The applause would start out randomly and then quickly become strongly synchronized. Synchronized clapping would continue for a short while and then disintegrate into random clapping, from which synchronized clapping would reemerge, and so forth.

Analysis of the recordings revealed two things:

1. The average noise level was greater during the random clapping than during the synchronized clapping.
2. During random clapping individuals clapped at roughly twice the frequency they used during synchronized clapping.

Clearly the greater volume during random clapping came because individuals were clapping faster. But during this phase, the time between individual claps varied more than when people clapped at the lower rate. That variability made it impossible for the group to synchronize at the higher rate—a result that has emerged in a number of studies of groups of globally coupled oscillators.

Néda concluded that audience members were caught in a conflict. On the one hand, they can express one value clapping as rapidly as possible, thereby making the loudest noise. If, however, they wish to express another value by synchronizing their clapping, then they have to clap more slowly, thereby lowering the volume. It is impossible simultaneously to maximize these two aims. The group deals with this conflict by switching back and forth between two different expressive regimes.

We would, of course, like to know what these two values are. The investigators assume that the loudness of the clapping reflects the audience members’ enthusiasm for the performance, while synchronous clapping expresses group solidarity. This seems reasonable enough. For our purposes, however, what is significant is the mechanism by which these two values, whatever they are, were expressed by the group. That mechanism is clearly self-organizing. No one leads audiences in this behavior. It just happens.

Walter Freeman has techniques for identifying and studying intentionality in the brains of individual animals. Néda and colleagues have demonstrated a method for studying group intentionality in this one very simple case—a simple case, however, that is grounded in coupled behavior. If we are to understand musicking we need techniques that work in more complicated cases, … , or the bell magic, or the jam session we turn to next. The important point is simply that group intentionality is amenable to empirical study.

What’s particularly interesting about this example is that we have two modes of behavior, synchronized clapping (expressing group solidarity) and loud random clapping (expressing enthusiasm for the performance), and they alternate with one another at quasi-periodic intervals. No one is directing this behavior. It just happens.

How? What’s the neural mechanism?

As far as I know, we no one’s looked into it. But I did a little thinking about it and here are some of my old notes:

Let us begin with the observation that the peripheral motor system has two major sets of fiber tracts, the so-called pyramidal and extrapyramidal systems. These tracts, in turn, have different connections with brain regions governing the skeletal muscles. The extrapyramidal system is phylogenetically older and is associated with general muscle tone, stereotypical behavior, and emotional expression. The new pyramidal system is associated with so-called voluntary behavior and fine motor control. Different though these systems are, they drive (and sense) the same muscles.

Let me repeat and emphasize that last statement. Our muscles are controlled through two different sets of nerves. One set originates in the neocortex and mediates voluntary behavior, actions that you can will, such as pounding a nail, threading a needle, and so forth. The other set originates in subcortical structures – the so-called lizard brain – and governs behaviors which is not under the direct control of our will. We can’t will the facial expression of emotion, for example, or laughter. These two systems sometimes come in conflict, for example, when we try to suppress (voluntary) the facial expression of anger (involuntary). Let’s continue with the notes:

I’m going to associate both the periodic aspect of clapping and synchronization with the pyramidal system. Why? Because there is an area of the motor cortex known as the supplementary motor area (SMA) and it is associated with repetitive and sequential actions in monkeys and humans. The SMA affects muscles through the pyramidal system. I note that, among the primates, the sort of interpersonal synchrony we see in clapping seems to be unique to humans.

In contrast, I’m going to associate enthusiasm with the extrapyramidal system and its central correlates (such as the basal ganglia). Clapping is a form of social signaling, and primates do quite a bit of that. I thus speculate that this behavior is neurally grounded in some form of primate signaling that is most likely mediated by the limbic system.

This suggests an neural interpretation of the alternation between loud random clapping and synchronized clapping. Rather than thinking of this message as being about the performance (loud) and the group (synchronized) we could think of it as being about the limbic system (loud) and the neocortical system (synchronized). How do we decide between these two alternatives, and any others that might be suggested? I suspect that this question is ill-posed. All we’ve got in the signal is an alternation between two modes. That alternation results from the overall properties of the behaving system rather than being explicitly controlled by some switching mechanism. We can examine the system and, on the basis of that examination, derive whatever labels we wish for the two modes. But those labels are just conveniences for us, the analysts and modelers. From the point of view of the individuals in the behaving system all we’ve got are two different modes. No more, but also, no less.

I suggest finally that the entire behavioral complex displays a form of social interaction that is uniquely human. This social interaction is not mediated by some specific “mental module” but by a pattern of functional interconnections that extends throughout the brain. It is the entire pattern that is the locus of specifically-human sociality, not any one neural center implicated in that pattern. The interesting feature of that pattern is the relationship between the cortical and subcortical structures, between the phylogenetically new and the phylogenetically old.

I speculate that all of the relevant neural structures predate the final emergence of modern human behavior patterns. The emergence of this form of social interaction thus has no particularly strenuous genetic requirements. It would seem to be something that just emerged when, somehow, some individuals first “captured” the possibility of synchronizing their repetitive moves, thus bringing about a new pattern of neuro-muscular interaction.

I note that this activity is intrinsically a group activity. An isolated individual can be enthusiastic about a performance they’re watching on TV, for example, and express that enthusiasm but jumping up and down and hooting and hollering. But there’s no group around with which they can express solidarity through synchronized clapping.

Thus, this example takes us into the realm of memes, the building blocks of collective culture, which I’ve been discussing in another series of posts. What about the oscillation between these two modes of clapping: Is that memetic?