Skip to main content

New approaches to (digital) social networks

Corpus Linguistics
Thursday 15 December 2022, 12:00

New approaches to (digital) social networks’

Prof Mikko Laitinen (University of Eastern Finland)

To view the recording, please click here.

Social networks play a considerable role in language variation and change, and social network theory has offered a powerful tool in modeling how linguistic innovations spread into communities (Milroy 1987). However, existing work on networks, with foundations in linguistic anthropology, leaves open various questions. Milroy & Milroy (1992: 5), for instance, show that the participant observation methods effectively limit the analysis to networks with sizes “between 30 and 50 individuals”. The theory, therefore, relies on evidence from networks that are substantially smaller than natural human networks, with sizes around 150 (Dunbar 2020). Second, much of the evidence comes from traditionally close-knit urban working-class settings or peripheral rural communities, and it has been suggested that the model “cannot be easily operationalized in situations where the population is socially and/or geographically mobile”, as is the case in many contemporary societies (Milroy 1992: 177).

This presentation introduces an algorithmic method that utilizes interaction data obtained from digital social networks. This interaction-based information can be quantified using network parameters (e.g. density, connectedness, and node similarity) that are adopted from the graph theory but have not been used in sociolinguistics or in corpus-based study of variation and change.

I first present the method developed by my group (Laitinen & Fatemi 2022) and then in the empirical part establish network parameters to a set of 3,935 randomly selected ego networks, which contain data from 233,774 individuals from the UK and the US. The median network size in this dataset is 51 individuals. The empirical case study utilizes all the texts from all the individuals in the networks and investigates how two linguistic features currently undergoing

change are conditioned by network information. My empirical study concentrates on one orthographic feature (contractions of negatives (e.g. not >n’t) and verbs (e.g. we will > we’ll)) and one grammatical structure (NEED to + V), both of which are undergoing frequency increases in English, but are driven by differing forces of colloquialization and grammaticalization (Leech et al. 2009; Daugs 2017). Our observations using data-intensive methods may lead to rethinking the role of social networks in language change.


Daugs, R. 2017. On the development of modals and semi-modals in American English in the 19th and 20th centuries. In Hiltunen, McVeigh & Säily (eds.), Big and Rich Data in English Corpus Linguistics: Methods and Explorations. (Studies in Variation, Contacts and Change in English, vol. 19).

Dunbar, R. 2020. Structure and function in human and primate social networks: Implications for diffusion, network stability and health. Proc. R. Soc. A. 476A.

Laitinen, M. & M. Fatemi. 2022. Big and rich social networks in computational sociolinguistics. In Rautionaho, Parviainen, Kaunisto & Nurmi (eds.), Social and Regional Variation in World Englishes: Local and Global Perspectives, 166–189. London: Routledge. doi: 10.4324/9781003227342-9.

Leech, G, M. Hundt, C. Mair & N. Smith. 2009. Change in Contemporary English. A Grammatical Study. Cambridge: Cambridge University Press.

Milroy, J. 1992. Linguistic Variation and Change: On the Historical Sociolinguistics. Blackwell.

Milroy, L. 1987. Language and Social Networks. 2nd ed. Blackwell.

Milroy, L & Milroy J. 1992. Social network and social class: Toward an integrated sociolinguistic model. Lang. in Society 21, 1–26.