Power, Innovation, and Language Creativity in an Online Community of Practice

Student thesis: Phd


This research utilises a mixed-methods approach and a near-complete corpus of all posts made to Popheads, a Reddit-based online music-orientated Community of Practice (CofP), to analyse two key unresolved topics in the intersection between CofP theory and sociolinguistics. The first topic centres on the applicability of the weak-tie theory of language change (Milroy & Milroy 1985) to an online CofP. The principles of this theory have been found to hold in studies of offline CofPs (Bucholtz 1999; Eckert 1988; Eckert 2000; Mendoza-Denton 2008; Moore 2010), yet the only study to explore the theory in an online CofP was grounded in an ungeneralisable linguistic context (Stewart et al. 2017). The first research question subsequently asks if the weak tie theory's prediction that the innovators of linguistic forms will be peripheral members holds in an online CofP. On the contrary, three of the four forms studied were innovated by non-peripheral members who scored highly across multiple markers of status. This departure from previous findings may be related to the more positive attitudes towards linguistic creativity in online environments. The second research question asks if the weak tie theory's prediction that the early adopters of linguistic forms would be high-status members holds in an online CofP. This research concluded that while there were four status markers that the majority of early adopters scored higher than the community median on, there were five others that did not appear to be linked to early adoption. Therefore, the monodimensional concept of being 'high-status' was shown to be an overly simplistic way of conceptualising the multidimensional concept of status. The most satisfactory answer to this research question is that members who are early adopters tend to score highly across four markers of status ('comments contributed', 'replies received', 'submissions contributed', and 'months active'). The third research question uses principal component analysis and logistic regression analysis to determine which status markers are statistically significant predictors of early adopter status. All four case studies showed that 'commenting behaviour' was a statistically significant predictor, with early adopters tending to be highly prolific commenters. Three of the four case studies demonstrated that a measure which approximates if members are on inbound or outbound trajectories in the community was also a statistically significant predictor of early adopter status, with early adopters tending to be dedicated members on inbound trajectories. Finally, two of the four case studies also confirmed an interaction between these two variables: being a prolific commenter and being on an inbound trajectory is a significant predictor of early adopters status. This research concludes that what is important when considering early adopter status is where members sit on multiple hierarchies and their position in the intersection of those hierarchies. Meanwhile, the second unresolved topic explores the relationship between power, status and hierarchy by looking at the stages of linguistic content moderation, in which community-salient linguistic forms were banned from usage in the Popheads community. I conclude that the relationship between power, status and hierarchy is more complex and multidimensional than suggested in much of the current CofP literature (Davies 2005; Kerno 2008; Moore 2006; Wenger 2010). There are multiple hierarchies at work in this CofP and having high status in certain hierarchies seems to correlate with power to control and influence linguistic content moderation policies. For instance, having moderator status gives members the power to implement and enforce content moderation policies, while prolific comment contribution seems to correlate with the power to take back control of the lexicon and violate the linguistic content moderation policies.
Date of Award1 Aug 2021
Original languageEnglish
Awarding Institution
  • The University of Manchester
SupervisorMaj-Britt Mosegaard Hansen (Supervisor) & Andrea Nini (Supervisor)


  • censorship
  • sociolinguistics
  • corpus linguistics
  • community of practice
  • power
  • weak-tie theory of language change

Cite this