We analyze Nordic social media users by clustering them based on their connections on Twitter. The data consists of 15,794 users in the five Nordic countries: Finland, Sweden, Norway, Denmark, and Iceland. We first create an undirected graph from the friendship relations (mutually following each other), then divide the graph into five clusters using a recent M-algorithm, and finally compare the results to users’ locations. The results demonstrate that the users are strongly clustered according to their home country. There is surprisingly little interaction across the countries despite the fact that they are, except for Iceland, physically close to each other and have cultural and linguistic similarities. The main language of the four countries belongs to the Germanic languages, while Finnish is typologically distinct. We further explore content from users in each country, analyzing its alignment with connectivity patterns. Our findings reveal a discrepancy between user-generated content similarity in the Nordic region and their connectivity patterns.

We investigate the uses of fuck in digital social networks from social media, Twitter/X in this case. Social media outlets have so far been predominantly treated as massive text collections, but they can be effectively used to investigate the role of social networks in shaping human communication. We use user-generated texts from 5,660 social networks (with 435,345 users and 7.8 billion words) from three settings (UK, US, and Australia). With embedded network information, this massive dataset enables us to investigate how network properties, that of the size and the strength of the network, influence the use of offensive words in these three settings. Our findings show that Americans use fuck most frequently, while Australians least frequently but they are highly creative with spelling variants of the word. Contrary to prior studies, we observe that people on this social media application swear more with acquaintances than with friends, but only in smaller networks − in larger networks of >100 people, the differences level out. Overall, this study highlights the benefits of using social media data that can be enriched to allow access to the social networks that people interact in.

This article combines the study of large-scale social media data with social network theory in sociolinguistics. Given that the purpose of social media is to form networks and communities, big data from social media applications could have substantial potential in deepening the understanding of the role of networks in language variation and change. The study first presents an algorithmic method suitable for directed-graph ego networks in computer-mediated communication. This method measures network strength and enables us to enrich social media data with a network parameter that indexes how strongly (or loosely) people in a network are connected to each other. We then use a large dataset of c. 4.8 billion words from nearly four thousand networks to study how network strength conditions linguistic change. The results show that online networks are highly similar to traditional offline networks, a finding that enables fixing a major methodological limitation in the study of weak ties, namely that the method is less suited for studying socially and geographically mobile individuals. This finding makes it possible to apply the theory of social networks in sociolinguistics to very large digital networks in social media.

This article looks into using large-scale social media data in SSH research and in particular in studies of language variation and change. It presents a study that investigates the role of social networks in linguistic variability. Previous studies have convincingly shown that networks in which people are connected to each other in loose ties tend to contribute positively to language change. Conversely, networks in which people are closely connected to each other inhibit change. This conclusion is, however, based on small datasets from small networks, and this study tests if the difference is diluted when network size is closer to human average. The results from 3,935 networks suggest this to be the case. Towards the end, the article suggests numerous ways in which large-scale social media data and the use of data intensive methodologies could be increased and encouraged in SSH research.

Previous research has highlighted the importance of network structures in information diffusion on social media. In this study, we explore the role of an individual’s social network structure in predicting publicly announced intention of protest participation. Using the case of ecological protests in Russia and applying machine learning to publicly-available VKontakte data, we classify users into protesters and non-protesters. We have found that personal social networks have a high predictive power allowing user classification with an accuracy of 81%. Meanwhile, using all public VKontakte data, including memberships in activist groups and friendship ties to protesters, we were able to classify users into protesters and non-protesters with a higher accuracy of 96%. Our study contributes to the political-participation literature by demonstrating the importance of personal social networks in predicting protest participation. Our results suggest that in some cases, the likelihood of participating in protests can be significantly influenced by elements of a personal-network structure, inter alia, network density and size. Further explanatory research should be done to explore the mechanisms underlying these relationships.

Social media data have substantially enlarged the potential pools of evidence in the study of variation and change in English. They offer access to language use of large numbers of informants, but the downside is that they contain inadequate social background information. This seriously restricts the theoretical insight, limiting investigations to the study of actuation of change and leaving out those that focus on diffusion. It is widely held that while actuation is largely functional, diffusion requires information on the broader social structures within which speakers belong. We present an algorithmic method for adding social information to Twitter. Our method builds on interaction parameters (size and structure of networks, similarity, and communication frequency). Using such participant-centred information as a proxy for social information increases the empirical validity substantially. In addition, the article presents a case study of how networks of varying strength condition ongoing change. The data are drawn from five metropolitan centres in the UK and from English as a lingua franca setting in the Nordic region. The results suggest that network size plays an important, yet understudied, role in the social network theory in sociolinguistics.

Earlier studies have established that the (perceived) similarity of users is highly subjective and reflects more on how people respect/admire others rather than their characteristics or behavioral similarities. We study this phenomenon among Twitter users, and while confirm that it is indeed the case, we further explore the components of similarity by investigating it using data from three categories (interactions between egos and alters, profile-based activity history, and linguistic content in the messages). We use interactions as estimation for admiration and observe that it has more impact and a higher correlation to the perceived similarity than other objective measures, including similarity based on user profiles and their use of hashtags.

Investigation of social networks formed by individuals in various contexts provides numerous interesting and important challenges for researchers and practitioners in multiple disciplines. Within the field of variationist sociolinguistics, social networks are analyzed in order to reveal the patterns of language variation and change while taking the social, cultural, and geographical aspects into account. In this field, traditional approaches usually focusing on small, manually collected data sets can be complemented with computational methods and large digital data sets extracted from online social network and social media sources. However, increasing data size does not immediately lead to the qualitative improvement in the understanding of such data. In this position paper, we propose to address this issue by a joint effort combining variationist sociolinguistics and computational network analyses with information visualization and visual analytics. In order to lay the foundation for this interdisciplinary collaboration, we analyse the previous relevant work and discuss the challenges related to operationalization, processing, and exploration of such social networks and associated data. As the result, we propose a roadmap towards realization of visual sociolinguistic network analysis.

Social networks play a role in language variation and change, and the social network theory has offered a powerful tool in modeling innovation diffusion. Networks are characterized by ties of varying strength which influence how novel information is accessed. It is widely held that weak-ties promote change, whereas strong ties lead to norm-enforcing communities that resist change. However, the model is primarily suited to investigate small ego networks, and its predictive power remains to be tested in large digital networks of mobile individuals. This article revisits the social network model in sociolinguistics and investigates network size as a crucial component in the theory. We specifically concentrate on whether the distinction between weak and strong ties levels in large networks over 100 nodes. The article presents two computational methods that can handle large and messy social media data and render them usable for analyzing networks, thus expanding the empirical and methodological basis from small-scale ethnographic observations. The first method aims to uncover broad quantitative patterns in data and utilizes a cohort-based approach to network size. The second is an algorithm-based approach that uses mutual interaction parameters on Twitter. Our results gained from both methods suggest that network size plays a role, and that the distinction between weak ties and slightly stronger ties levels out once the network size grows beyond roughly 120 nodes. This finding is closely similar to the findings in other fields of the study of social networks and calls for new research avenues in computational sociolinguistics.