Social groups vary considerably in form and function. Their social networks should reflect this variation. We are interested in characterizing the networks of cohesive groups as teams. In classic social networks, associations are built around individual gain. However, team networks coordinate around a common goal or outcome. In teams, individual connections should be built around the function of the whole network, rather than the network built around the individual. In biological systems, this should reflect selection operating at the group versus individual levels. Our projects include characterization of social insect networks and analysis of NBA basketball – a game where team coordination and cohesion is critical to success.
1Structure and function of social insect networks
We are using structural analyses of antennal contact networks to ask: how does network structure in ant colonies relate to colony function? Ants use antennations in virtually all of their inter-individual communication, making it useful for analyzing communication networks. In recent work, we dissected the networks of harvester ant colonies to their most basic structures, motifs or 3 node subgraphs. Colonies have an abundance of feed-forward loops, which are associated with regulatory networks and with resampling. They lack the “social clique” motifs generally associated with social networks such as the world wide web. Read more in Waters and Fewell, 2012.
Workers in a social insect colony do not seem to establish long-term associations with each other, in contrast to the ties found in many vertebrate social networks. Instead, worker-worker interactions are ephemeral and often built around work. This allows us to diagram and test division of labor as a network by collapsing the network down, considering tasks and information centers as nodes. This allows us to ask how information flow mediates task regulation. In honey bees, for example, up- or down-regulation of pollen is dependent on information flow from nurses performing brood care, from the brood (which are the primary consumers of pollen) and from pollen foragers dancing to recruit new workers into foraging. These strong connections between tasks allow the colony to respond rapidly to changes in task need. Read more in Fewell 2003; Fewell et al. 2009.
2Basketball teams as strategic networks
We are also interested in other contexts in which individuals coordinate around a common goal – well, and Dr Fewell just likes basketball. Basketball teams are beautiful examples of how coordination among group members plays an important role in group success. However, most current statistical analyses of the game focus on individual metrics. We apply network-level analyses of NBA games to evaluate differences in team strategy. This graph shows the passing network of the 2010 LA Lakers and reflects the teams use of the “triangle offense”, which involves players outside the point guard in distributing the ball. Read more in Fewell et al. 2013.