Skip to main content

Ants: Why the poor communication skills?

Effectiveness in networks is not necessarily synonymous with interconnectedness, as Braess’s Paradox demonstrates in networks where individuals selfishly choose their route. However, the idea that the effectiveness in an interaction network is not always completely positively correlated with interconnectedness seems even more counterintuitive than Braess’s Paradox. A recent study of interactions between members of a colony of ants of the species Temnothorax rugatulus challenges the belief that self-organizing interaction networks conscribe to a universal type that seeks to maximize interconnectedness and thereby transmit information very quickly (think about rumors spreading through a high school).

University of Arizona research Anna Dornhaus and doctoral candidate Benjamin Blonder captured several colonies of the ants, set them up in an artificial nest between two glass slides, painted hundreds of the individual ants to differentiate between them, and then painstakingly analyzed their physical interactions. Their findings—that ants are inefficient at spreading information—are somewhat surprising. The ants did not seem to follow the property of six degrees of separation that allows information to spread so quickly along human social networks. Instead, the ants required more intermediary steps to transmit information among individuals than is expected. What is especially interesting is that even though only physical interactions were counted (they are easier to see and ants have poor vision), the ants’ efficiency of spreading information was lower than if they had randomly bumped into each other, which means that certain ants actively avoid one another.

How can it be that ants, which must have a nearly optimal network strategy because they have survived for millions of years, are bad at spreading information? Blonder and Dornhaus believe that ants naturally form networks so that they are not overly communicative because “If you spend too much time interacting, then you’re not actually getting anything done.” It seems ant networks, which must be optimized to effectively produce action from ants, have sacrificed communication for action because too much communication would slow the network down.

Another interesting point mentioned, but not explored, by the article is the behavior of the individual ant. Ants do not consciously make decisions for the good of colony; instead they seem to employ a pre-programmed mixed-strategy in their decision-making. Ants likely employ mixed strategies that maximize social welfare in games between ants. For example, among a group of foraging ants that find food, some ants will start to bring the food back, and others will go alert other ants to help. If all of the ants began to bring the food back without alerting others or if all of the ants went to alert others, the colony’s total payoff (bringing the food back efficiently and quickly) would be lower.


Leave a Reply

Blogging Calendar

September 2011