This presentation revisits the weak social tie model in sociolinguistics and explores to what extent big data can be utilized to operationalize network ties. As is widely known, this model predicts that loose-knit social networks promote diffusion of innovations, whereas dense multiplex networks form norm-enforcing communities (Milroy & Milroy 1985). This model is based on evidence from small networks and has been criticized for its lack of statistical precision (cf. Murray’s (1993) argument that network values are a continuous variable).
We tackle both of these issues. Firstly, we make use of a large database of 199,832 Twitter accounts (c. 7.8 million messages) from a real-time corpus of tweets and their metadata from five Nordic countries. Secondly, we operationalize network ties as a continuous variable. The number of followers indexes truly weak ties (i.e. requires no action from a user). The number of friends is an indication of slightly stronger links (i.e. requires user effort). We do not consider friends and followers to be fully equivalent to weak and strong ties a priori. The dependent variable is the proportional share of tweets in English per account in the region where English is used as a non-native resource.
The results show that for small networks (<100 individuals) the independent factors behave like weak and strong ties: the truly weak ties show strong positive correlations of network size and the dependent variable, and the stronger link shows no correlation. This changes when the network size becomes larger (>100–130 individuals), and the distinction between the independent factors disappears. This finding corresponds to Larsen et al’s (2006) argument that the distinction between weak and strong ties is unclear in mobile societies, and had we restricted ourselves to small networks, this observation could not have been established. Our findings call for new sociolinguistic studies that focus on networks of varying sizes.
References:
Larsen, Jonas, John Urry, Kay Axhausen. 2006. Mobilities, Networks, Geographies. London: Ashgate Publishing.
Milroy, James & Lesley Milroy. 1985. Linguistic change, social network and speaker innovation. Journal of Linguistics 21. 339–384.
Murray, Stephen O. 1993. Network determination of linguistic variables? American Speech 68(2). 161–177.
