Humans might not seem to have much in common with swarming ants or schooling fish or flocking birds. But, according to Iain Couzin, the animals' coordination and cooperation can teach us about human behavior -- and even our voting patterns.
I spoke last month with Couzin, an assistant professor of evolutionary biology at Princeton University. Below are excerpts from our interview.
Why should we care about how animals coordinate behavior?
Everyone has seen bird flocks, fish schools and ant swarms. Understanding these types of processes is about understanding our world and environment. There are lots of implications for humans. Our bodies are a collective of cells, mostly cooperating. With tumors, we want to understand these rogue cells. Our brain is a collective. Each neuron is only capable of relatively simple information processing. But in the billions, [they allow us to feel] love and hate and have consciousness. There are some big questions about collective behavior in brain dynamic. Finally, we ourselves are embedded in our own collective human society. Many of the features we see in animal groups can have implications in understanding our own society: why and how we coordinate our behavior, why stock prices suddenly crash, how people choose what to buy.
Because we're interested in understanding the fundamental principles of collective behavior, we end up studying these types of processes with a wide range of organisms. We did laboratory studies of fish, for example. You might think that can't be related to humans, but the models we've created to understand how fish make decisions inform us about how humans make decisions. We also study humans. We've tracked the gazes thousands of pedestrians looking at socially-contagious behavior in human crowds.
How did you come to do this work?
When I was a kid, I loved animals. I was particularly fascinated by ants. How do these little creatures coordinate their behavior and make decisions? It was an interest in the natural history of animals at first. In college, I studied biology and got more involved in trying to understand the science of animal behavior.
Many people assume insects move in groups because they're working together. But your work on locusts turned that theory on its head. Talk about that study.
Ants are closely related to each other. They're sisters working together for the benefit of the colony. When we look at an ant colony, that view of them cooperating is appropriate. We have this tendency when we see other animal groups, such as locust swarms, to think the highly-coordinated nature of the interactions equals cooperation. But there's a dramatic difference between coordination and cooperation.
When individual locusts run out of essential nutrients -- they've eaten all the vegetation -- what do they do? The only thing they can do is turn on each other. They're highly cannibalistic. They start trying to bite each other. It's very dangerous to be bitten in such an environment, so they try to avoid it. The outcome is that individuals start chasing after those moving away, while trying to avoid others. The outcome is what looks like a highly cooperative behavior. But it's not. It's a forced march where everyone is trying to eat others and avoid being eaten. That shaped our perspective on one of the driving forces for collective behavior.
How did your study on schooling fish differ?
When people see two types of behaviors that look similar, such as locusts swarming and fish schooling, they tend to assume the reasons for those behaviors are the same. That's not the case. In schooling fish, there's not necessarily a leader within the group. Individuals follow local attractions and coordinate their behavior to create the fantastic synchronized patterns we also see in flocking birds. This has important consequences on the evolution of these groups and how these individuals function as a collective.
What can your work in animals tell us about how humans coordinate behavior?
In the fish study, we were trying to understand how animal groups make decisions and whether animals can effectively vote. Voting among humans requires explicit counting, but in animal groups that's not possible. Fish can store information for weeks, months or years. We used the fact that fish can remember things to give them different preferences to associate rewards with food choices. We can manipulate not just the number of individuals that associated these locations with food sources, but we can also set up competition between individuals. We've shown that the individuals in the majority can effectively dictate what the group does. They're voting without counting.
If we put uninformed individuals into the group that haven't been trained and don't have any preference, they spontaneously reinforced the majority view. Individuals that don't have strong preferences have the same type of result -- promoting democratic consensus -- not just in the fish group, but also in human opinion dynamic. There's an underlying mathematical principle here. What we've discovered with fish might be related to how people come to consensus. We found a mathematical parallel between what appeared to be disparate systems.
Talk about your research on how humans follow each other's gazes, which took you outside the laboratory setting.
We're biologists, so we're used to studying animals in their natural environment. The majority of studies on human social interactions have been conducted in controlled laboratory conditions. We were interested in how people behave in natural environments. There was a wonderful study conducted about 40 years ago by Stanley Milgram showing that when people walked down the street and suddenly looked up, other people had a propensity to copy that behavior. As the stimulus group size increased, you'd have an increasing propensity to copy. For the past four decades, people have interpreted this to be that there's a strong copying behavior within the human environment.
We replicated that experiment. Instead of using slide film like Milgram, we used computer imaging and computational techniques to record the motion of pedestrians and the direction in which they were looking. We found that the Milgram study has been completely misinterpreted. Human copying behavior is weaker than expected. There have been hypotheses that people copy the behavior of crowds because of social conformity. That's not the case. People in front of those looking up who could see the people looking up did not tend to copy the gaze. But individuals walking behind those looking up did. That was validated in a further study at Princeton.
People are more likely to change their gaze when taking cues from the back of someone's head. That's unexpected. Every psychological study I've seen involved gaze copying while looking face-to-face. We show in a natural context that face-to-face contact inhibits gaze copying. The gaze copying moves backward through a crowd. This has biological consequences. By copying the back of people's heads, our attention is being drawn to places that might be relevant to us in the future because we're walking in the same direction as those pedestrians.
Which of your studies has most surprised you?
The study of locusts was one of the most surprising. We weren't looking for it. It was a chance observation. We found it so interesting that we started studying it. That's one of the great joys of doing biology.
Watch Couzin's PopTech talk.
Photo: Iain Couzin in Mauritania studying locust swarms