Article,
Persistence of social signatures in human communication
Proceedings of the National Academy of Sciences, 111 (3): 942--947 (Jan 6, 2014)
DOI: 10.1073/pnas.1308540110
Abstract
The social brain hypothesis has suggested that social network size (and
structure) is constrained by a combination of cognitive processes and the time
required to service social relationships. We test this hypothesis in humans
using a unique 18-month mobile phone dataset by examining changes in the
structure of social networks across a major change in subjects' social and
geographical circumstances. Our analysis reveals that the time allocation
patterns of call frequency by participants to network members have a
distinctive overall shape, where a small number of top-ranked network members
received a disproportionately large fraction of calls, with some individual
variation. However, importantly, whilst there was a large turnover of
individual network members, these changes have little effect on the time
allocation patterns of each individual: individuals thus displayed a
distinctive "social signature" that was both persistent over time and
independent of the identities of the network members. This provides the first
direct evidence that social networks are constrained by a combination of
cognitive constraints and the time individuals have available for social
interaction, confirming one of the key assumptions of the social brain
hypothesis.
