Iain Couzin's Collective Behavior

Couzin's work is aimed at understanding how collective behaviour in animals emerges, using a combination of fieldwork, computer simulations, lab experiments, and mathematical models. wikipedia

By developing an integrated experimental and theoretical research program we aim to explore functional properties of groups in a context that can reveal how, and why, social behavior has evolved. site

Our research brings us to the interface of a variety of fields, from physics, to computer science, to politics, to psychology. Many of these interdisciplinary collaborations are fostered in the Center for the Advanced Study of Collective Behaviour at the University of Konstanz, a DFG cluster of excellence which Iain Couzin co-directs. site

Excerpts from the "Joy of Why" interview. transcript

The collective behaviors of animals differ in their details from one species to another, but they largely adhere to principles of collective motion that physicists have worked out over centuries. Now, using technologies that only recently became available, researchers have been able to study these patterns of behavior more closely than ever before.

COUZIN: Placozoa, yeah. This little creature was found crawling around on the glass of aquaria, tropical aquaria. You can see it with the naked eye. It’s about a millimeter, maybe a millimeter and a half if it’s very big. And, you know, looking into this remarkable creature has only really recently sort of drawn the attention of scientists. And that is largely because this strange little quirky swarm of cells actually has the genetic complexity that you would associate with a much more sophisticated organism. For example, it has a large range of neurotransmitters, yet it doesn’t have neurons. wikipedia pnas

Whereas the animals, the interactions between birds in a flock, they’re invisible. They have no physical form. And so one may initially think, well, then it’s only an analogy. In fact, I would say until about five to 10 years ago, I thought it was just an analogy too. I thought that these differences must be very important. But what we’re beginning to understand is that the common feature that they share is computation.

It’s a little bit like, you know, a neuron transmitting information via electrical signals. In this case, it’s not electrical signals. It’s really the density and the turning of the individuals that percolates across the group, but it gives those individuals afar information where the threat is, so they can begin to move away from it very quickly.

This again relates to what we’ve learned from physical systems, specifically physical systems close to a phase transition.

STROGATZ: So I’m just wondering, with now in the case of collective behavior, if nature tunes a flock to be near some kind of point of instability or criticality. Are you suggesting that’s part of what makes it so effective?

COUZIN: Yeah, that’s exactly what I’m suggesting. And so, for example, you know, again, a very recent paper within the last couple of years that we published, we asked, you know, what about getting the best of all worlds? What about if, you know, under general conditions you want to be stable, you want to be robust. But sometimes, you want to become hypersensitive. And so in natural selection, biological systems have to balance this amazing, sort of seemingly contradictory status of being both robust and sensitive. How can you be both robust and sensitive at the same time? pnas

And so what we found was that individuals do not change. What happens is the network changes. The individuals move to change the structure of that network, and it’s that that causes the group to suddenly become more sensitive and more flexible.

And by working with physicists and mathematicians and biologists, and by conducting experiments on animals in virtual reality we can completely control the input. We can completely control the causal relationships.

So, how does the brain represent space-time? And how does that matter in terms of decisions? And what on earth does that have to do with collective behavior of animals? What I realized about five years ago, is that I think there’s a deep mathematical similarity, and I think there are deep geometric principles, about how the brain represents space and also time.

It represents space in a non-Euclidean coordinate system. And we can then show mathematically why this is so important, which is that when you start dealing with three or more options, then actually warping spacetime, making space non-Euclidean, can dramatically reduce the complexity of the world into a series of bifurcations. And close to each bifurcation, it amplifies differences between the remaining options. So there’s this beautiful internal structure.

At the Department of Collective Behaviour, part of the Max Planck Institute of Animal Behavior, researchers are putting locusts into simulated worlds, both virtual and physical, in the hope that they can figure out how devastating swarms form and move.