How is the size and structure of human social networks influenced by cognition? Existing research has identified a correlation between neocortex size and group size in a variety of primates, including humans, but it is unclear how the deeper structure of our social networks is impacted by cognitive limitations. Such limits have important implications for social network analysis both because they can help us to distinguish network features that can result from the conscious preferences of actors from those that must emerge unintentionally, and because they may contribute to a causal explanation for several observed regularities in human social networks. This project focuses on social memory- the capacity to recall information about social ties- as the bedrock of the human ability to reason about social relations. An experimental method developed specifically for this research will be used to measure the size of human social memory as well as to identify and explore proposed socio-cognitive strategies that act to enhance recall without imposing the need for additional organic memory. It is expected that networks with certain characteristics will be easier to recall even when they contain more ties to be remembered. The results from this experimental evaluation will then be used in a non-experimental component to relate the size of an individual?s memory capacity to several characteristics of their personal network, including its size, density, and degree of structural balance. It is expected that those who possess better social memories will have personal networks that are more complex and more difficult to recall. This project will collect data from over five hundred subjects in eight experimental conditions to provide the first detailed assessment of social memory capacity in humans and explore its connection to personal network structure.
Broader Impacts
This research will have impacts in several areas. First, in an academic sense, it is part of a growing sociological interest in the interaction between evolution, biology, and social life. How are our social structures influenced by, and to what extent do they influence, our biology? Second, this project will add to the study of social networks by helping to produce general predictive theory. A better understanding of the cognitive underpinnings of networks will help to generate expectations about when particular network structures may be observed. Third, an enhanced understanding of how networks are constrained by cognition will aid in improving coordination in high stress environments where efficient teamwork is key. Finally, this project will include both graduate and undergraduate research assistance, thereby providing valuable hands-on training in scientific research for the next generation of scientists.
Social networks are complex structures consisting of the relationships between individuals. Effective social behavior requires individuals to manage their own relationships (e.g., tend to a relationship with one's spouse) as well as to monitor relationships between others (e.g., know whether one's spouse and their sister are feuding). However, the number of relationships possible in a group increases at a rapid rate with group size; for example, a group of ten people contains forty-five possible relationships (if we assume that A->B is the same as B->A), while a group of fifteen people contains one hundred and five potential relationships. Cognitively managing this information is, therefore, a significant challenge. This project investigated whether humans use "compression heuristics" to simplify the recall of social networks. Compression heuristics are patterns used to organize the learning experience such that information on the relationships can largely be discarded, and the networks can be cognitively rebuilt from these simplifying rules at a later time. This reduces the memory demands on the individual, at the cost of a certain degree of inaccuracy in recall. In a series of experiments, this reseach showed that humans do, indeed, use these heuristics and they operate as predicted. Additional work indicates that females exhibit higher overall capability in this are than males, and that this advantage may be biological in origin, though this cannot be determined with certainty at this time. In general this work is consistent with the "Social Brain Hypothesis" that human intelligence evolved to deal with social, rather than physical challenges. Because the complexity of the social problem increases with mean intelligence, social challenges continually increase, driving intelligence upwards. However, the use of compression heuristics by Humans should have encouraged the development of high levels of general problem solving intelligence, useful for identifying, inventing, and applying compression heuristics, rather than simply task-specific intelligence. Thus, human intelligence may be a result of a fundamental interaction between social and biological systems. More specifically, this work has potential relevance to treatments for autism spectrum disorder. ASD is known to involve social cognitive deficits; ASD individuals are often less able to manage complex social interactions than neurotypical individuals. This research helps to make the actual cognitive processes that underlie sociability more clear, and make the methods of social encoding more explicit. While not directly studied here, these may give rise to improved treatments for ASD as well as new avenues of research. This research has to date resulted in three peer-reviewed scientific papers, multiple conference presentations, and several invited talks. It has also sparked a number of follow-up projects, including an ongoing effort to use EEG nets to monitor brain activity during compression heuristic use.
