How much conversation content is actually social

Abstract

How much of what we say to each other is of a social nature? In a pioneering study, Dunbar et al. (1997) estimated that ‘gossip’ – understood loosely as conversation about social and personal topics – accounted for about two-thirds of time spent on conversation. This result has been extremely consequential: in addition to its considerable popular impact, it was instrumental in motivating some of the most influential theories of human cognitive evolution (including language evolution - the ‘gossip’ theory of language origins, Dunbar 1998). However, it is not clear that this theoretical importance rests on a sufficiently strong evidential basis, as from today's perspective we can point to several limitations of both the dataset and approach of the 1997 study.

In our study, we revisited Dunbar's research question with methodological solutions made possible by corpus tools. We used Spokes (Pęzik 2014), a corpus of Polish informal conversations (N = 669; >2.6 million word tokens), based on live recordings of casual speech, recorded in private as well as public places, with speakers from a variety of Polish demographic backgrounds, such as age and education. We used 535 conversations, and coded each of the 274,954 lines of text therein as either carrying 'social' content (further subdivided into three main categories of social, and two sub-categories), or 'non-social' content. Our findings indicate an even larger portion, approximately 85% being of a social nature.

We should however stress that - in line with our evolutionarily motivated research question - we defined ‘social content’ very inclusively. Hence, we conclude with a crucial, if obvious, caveat: the proportion of ‘social’ to ‘non-social’ content in conversation depends mostly on how ‘social content’ is defined. This apparently self-evident reservation should not be overlooked in interpretations of numerical results.

Code of Conduct for ZAS events: The ZAS is committed to fair, respectful, and professional interaction at its events. Therefore, please observe the Code of Conduct for this event.