Credit: Steve Jurvetson
Scientists at the University of Bristol have a strange new tool for studying the brain: ant colonies. Ants and neurons, in spite of obvious physical differences, have a similar complexity in that their emergent group intelligence is far greater than that of their components. Both systems exist simultaneously as an individual and a collective, working without central control to make group decisions that look, from colony- or brain-level, like the product of a single mind. With a grant from the UK’s Biotechnology and Biological Sciences Research Council, computer scientist James Marshall and biologist Nigel Franks are taking advantage of the systemic similarities and physical incongruences of these two systems to illuminate general principles of group decision making, a process that pervades biology and society—from ants and neurons to people and governments.
Through this research Marshall says he hopes to “elucidate general principles of decision making by comparing the processes and structures of social insects with those in vertebrate brains, such as in the primate visual cortex.” Though the idea of comparing brains and ant colonies is not entirely new—Douglas Hofstadter used the metaphor as early as 1979 in his book, Gödel, Escher, Bach—Marshall and Franks are the first to formally explore it. Their research focuses on two specific types of decisions: house hunting in ant colonies and visual discrimination in monkeys.
Choosing a new home, or house hunting, is the most complicated decision an ant colony makes. When an ant nest is overcrowded or damaged, scout ants begin searching for a new building site by making independent evaluations of different spots and reporting back to the colony. A decision is made when a “quorum” is reached, when a certain number of ants agree on a location.
This same process occurs among neurons in a monkey’s visual cortex when the animal performs a visual discrimination task. In the task, a monkey is flashed an image of dots moving in different directions and must decide which way the majority of them are going. When the image appears, neurons in the monkey’s visual cortex gather bits of information from the monkey’s eyes, much like ants evaluating a nest site. As more data is gathered, the neurons with the correct answer gradually increase their firing rate. When their activity reaches a certain threshold level, the monkey makes a decision.
Thus, decisions in both brains and ant colonies are based on thresholds that can be adjusted for either speed or accuracy. Understanding how these mechanisms work “will not only advance our understanding of collective decision making by social insects and individual decision making by vertebrates, but could potentially give us ways to design machines for tackling these kinds of problems,” Marshall says.
According to Cornell University neurobiologist Thomas Seeley, the social insect metaphor can be scaled up even further, all the way to government. He is currently writing a book entitled Honeybee Democracy (Princeton University Press, 2010) about how we can improve our group decision making and government by imitating honeybees, which, like ants, are social insects. Honeybees have been making group decisions for tens of millions of years. And as Seeley says, “Democracy among humans traces back only 2,500 years or so to Athens, Greece, so it is perhaps not surprising that we are still fumbling around.”
The efficiency of a social insect democracy is largely a product of the simplicity of its constituents. “Human groups have factions with conflicting agendas, whereas social insects in a colony have much less self-interest and genuinely ‘want’ to converge on the best option,” Marshall says. Still, Seeley believes that there is much to be learned about decision making from honeybees. “I have learned several things from the bees that I have employed as department chair,” he says. Seeley cites a number of what he terms “swarm smarts,” such as the honeybee’s ability to avoid groupthink by having each “voter” assess options independently. Furthermore, when disaster strikes, the honeybees adapt their decision making by lowering their threshold levels and deliberation time. While we, as individuals, do this constantly (rushing to meet deadlines, ordering quickly in a crowded deli instead of weighing our options), our democracy can be slow and ungainly, endlessly logjammed over issues of immediate need.
The similarities between ants and neurons “suggest there are general principles of organization for building groups far smarter than the smartest individuals in them,” Seeley says. Group decision making is occurring constantly at different levels of complexity and across different scales of space and time; understanding how it works in one context can inform our understanding of other systems that might be more complicated or difficult to explore. By looking at decision making in all of its diverse incarnations, we can step outside of standard modes of reasoning and find new ways to talk about complexity in our ecosystems, our communities, our governments, and our minds.
Originally published July 23, 2009