Finally, it is still unclear what impacts computational social sciences will have for universities and disciplines.
First, the question about organizing our institutions for computational social sciences. Should computational social sciences be its own centre or institution or part of established departments? If they are in departments, in which of them: computer science, sociology, political science etc. Academia is not alone with this question: companies and organizations are similarly looking for different models on how to bring in data science skills to support work in organizations: should it be centralized into a core data science team, de-centralized to develop data science competencies in all teams or semi-centralized with data science competencies being embedded into teams, but forming their own centre (Patil and Loukides, 2011). King (2014) reflected the establishment of the Institute for Quantitative Social Science at Harvard University, which represents a centralised approach to this work. One of his core messages was the need to foster and incentivise collaboration. He also highlights that disciplinary silos may limit computational social sciences, both within social sciences and even across social science and disciplines like social sciences. However, in many universities traditional methodologies are decentralised: quantitative and qualitative methods are thought and explored by people who work in disciplinary departments. Maybe computational social science will similarly live within departments?
Second, computational social sciences changes who are seen as computational social scientists. Grimmer (2015) highlights how everyone can become social scientists as social data sources are increasingly available for researchers. Similarly Savage and Burrows (2009) discuss how interesting work in sociology takes place not in sociology departments but in corporations. While many highlight that social scientists provide valuable skills and knowledge in these investigations (Wallach, 2018; Grimmer, 2015), they are clearly not the only actors involved in the investigation of the society: computational social sciences invite engineers, computer scientists and physicists to think about the society. Bartlett et al. (2018) examines who are in the locus of legitimate interpretation: who are allowed to interpret results. They note how in biology and high-energy physics the interpretation role remains at domain experts. However, Bartlett et al. (2018) suggests that in computational social sciences that also non-social scientists can provide valid interpretations about the society.
Third, social scientists are deeply concerned about different epistemologies, referring to different understandings how we can gain knowledge from the society. Often these are discussed as a continuum, where on the other side are positivists-arguing that the reality is only what is observed (and measured) - and relativists-arguing how the reality is interpretations. Many scholars position themselves somewhere between these two positions-and often the two endings are used as caricatures in the discussions. However, several concern has been raised that “big data” and computational methods challenge epistemological positions (Kitchin, 2014; Boyd and Crawford, 2012). That is to say they allow particular types of research more-often more in the positivist side of the continuum. Partly the bias towards positivist research is true: computational methods like network analysis and algorithmic data analysis work only on what can be observed. They cannot “read between the lines” or conduct extensive interpretations from the data-like is done with some critical methods. However, as we showed in both of these chapters, researchers have a lot of freedom how they choose what to seek to to observe. This allows these methods to be used in more interpretation-focused research approaches as well (Ylä-Anttila, 2018).
To conclude, the impact of computational social sciences is still evolving. These three aspects all require collective consideration and work in shaping the institutional and disciplinary settings to achieve the impact - both academic and societal - we hope to have. Further years will formulate many answers to topics raised in these paragraphs. The ball is now in your court: think how to use skills developed through the book to formulate good research questions, form collaborative teams, and shape the future of scholarship.