Defining Digital Social Sciences 7

In this post, Lisa Spiro (Rice University) provides an overview of the different facets of digital social sciences, observing points of connection with digital humanities.

As a member of a research team investigating the skills and competencies important to digital scholarship, I’ve become interested in what “digital scholarship” means in different disciplines, particularly the social sciences and humanities. Perhaps not surprisingly, I’m finding some significant points of intersection between digital humanities and digital social sciences. For example, the Digging into Data Challenge promotes innovative research using computational methods across the humanities and social sciences, funding projects in literature, political science, law, and other domains. CLIR’s 2012 report on the results of the first round of Digging into Data, One Culture: Computationally Intensive Research in the Humanities and Social Sciences, recommends embracing interdisciplinarity and developing more inclusive models for collaboration. Reflecting this call for interdisciplinary collaboration, several digitally-oriented research centers explicitly encompass both the humanities and social sciences, including Northeastern’s NULab, the University of Illinois at Urbana-Champaign’s Institute for Computing in Humanities, Arts, and Social Sciences (I-CHASS), and Michigan State’s Matrix. What are we to make of the connections between humanities and social science research? And what does digital research in social sciences entail, anyway?

[pullquote]By developing a deeper awareness of how social scientists use computational methods to address research questions, humanists might gain new insights into how they can apply similar techniques to their work—and vice versa.[/pullquote] I ask these questions from the perspective of someone with a background in digital humanities interested in connections to (and differences from) digital social sciences. Of course, the social sciences and humanities have long been associated with each other, particularly fields such as history (classified as a social science, humanities, or both) and anthropology, owing to a common interest in culture, material objects, and interpretation. Indeed, interpretive social sciences are often brought under the broad umbrella of digital humanities. But the increasing significance of data-driven methods to the humanities as well as the social sciences seems to be sparking new connections, particularly between computational social science and computational humanities. 

While “digital humanities” is itself a fuzzy term, “digital social sciences” seems to be even less well defined. In the social sciences, several related terms fit under the general category of “digital social sciences,” including e-social science, computational social science, digital cultural heritage, and Internet studies. These terms differ in how they conceive of the role of digital technologies in the social sciences, whether in providing the infrastructure for computationally-intense, distributed research; fueling data-driven or computational analysis; supporting the curation, analysis and dissemination of cultural heritage materials; or providing the basis for the social, political, cultural, technical, psychological, and economic studies of the Internet. For ease of discussion, I’ll use “digital social sciences” as an umbrella term to refer to these different approaches, while acknowledging that the term lacks coherence and currency. In this post, I will briefly describe these different facets of digital social sciences and note points of connection with digital humanities. I hope that these general observations will spark further discussion about opportunities for deeper collaboration as well as about what makes each field unique.

E-Social Science/ Digital Social Research

In focusing on the infrastructure for distributed, computationally-intensive research in the social sciences, e-social science resembles e-science and humanities cyberinfrastructure initiatives. The term “e-social science” appears to be most common in the UK, probably because of the UK Economic and Social Research Council’s (ESRC) program to create a National Centre for e-Social Science as part of the nation’s e-science efforts. Established in 2004, this distributed center funded 12 nodes (plus a hub) that created “innovative tools, techniques and services.”[1. Peter Halfpenny and Rob Procter, “The E-Social Science Research Agenda,” Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 368, no. 1925 (July 18, 2010): 3762, http://dx.doi.org/10.1098/rsta.2010.0154.] Most of the nodes developed applications in domains such as simulation, qualitative analysis, statistical analysis, data management, and GIS, while one conducted “social shaping” studies to understand the impact of these technologies on research. While the UK e-science program aimed to “facilitate bigger, faster and more collaborative science” through grid computing, this model was found to be less applicable to social science research, which typically occurs on a smaller scale, uses a mix of qualitative and quantitative methods, and already has good research tools.[2. Ibid, 3765.]

Now ESRC prefers the more general term “digital social research,” which focuses on “the application of a new generation of distributed, digital technologies to social science research problems.”[3.  “Digital Social Research,” Economic and Social Research Council. Accessed March 11, 2014.] Digital social research encompasses both quantitative and qualitative approaches; it involves new data sources (such as social networking data), methods (such as social network analysis), capability (such as collaboration tools), scholarly practices (such as new publishing models), areas of study (such as Internet studies), and scale (such as global collaborations).

Computational Social Science

A 2009 article in Science by David Lazer, et al. declared the emergence of computational social science, as researchers use massive data sources (such as data from email, blogs, and web searches) to study human behavior.[4. David Lazer, Alex Pentland, Lada Adamic, Sinan Aral, Albert-László Barabási, Devon Brewer, Nicholas Christakis, et al., “Computational Social Science,” Science 323, no. 5915 (February 6, 2009): 721–723, http://dx.doi.org/10.1126/science.1167742.] According to Claudio Cioffi-Revilla, “Computational social science is the integrated, interdisciplinary pursuit of social inquiry with emphasis on information processing and through the medium of advanced computation.”[5. Claudio Cioffi-Revilla, “Computational Social Science,” Wiley Interdisciplinary Reviews: Computational Statistics 2, no. 3 (2010): 259–271.] Lev Manovich invokes Lazer, et al.’s article in calling for a computational humanities that uses data to understand cultural phenomena, such as visualizing patterns across a million manga pages. The intersection between computational social sciences and computational humanities is demonstrated by Northeastern’s NULab for Text, Maps and Networks, which is co-directed by Lazer, a political scientist, and English professor Elizabeth Maddock Dillon (NULab also develops digital cultural heritage projects such as Our Marathon).

“Text, maps and networks” provides an apt summary of three key approaches to computational social science. In his overview of the field, Cioffi-Revilla identifies five main methods. As I  provide short definitions of these methods based on Cioffi-Revilla’s work, I’ll also offer one or two examples of how they are applied in digital humanities.

• Automated information extraction. Computational social scientists use content analysis and text mining to monitor trends, extract event information, and study political rhetoric, just as digital humanists employ text mining to understand the characteristics of literary genres or social and political changes revealed in historical newspapers.
• Social network analysis. In social science, social network analysis is used to design better transportation and public health networks, understand terrorist networks, and gain insights into how organizations work.[6. Ibid.] Likewise, recent digital humanities work involving social network analysis includes mapping correspondence networks.
• Geospatial analysis. The spatial turn cuts across disciplines. Using GIS, researchers bring together data with mapping tools, investigating the spatial dimensions of phenomena such as the relationships between obesity and mobility, the spread of disease, or the dissemination of ideas.[7. Paul M. Torrens, “Geography and Computational Social Science,” GeoJournal 75, no. 2 (2010): 133–148.]
• Complexity modeling. Complexity modeling applies mathematical techniques to understand the interactions among elements in a system and disturbances to equilibrium such as violent conflicts, market fluctuations, and natural disasters. While complexity modeling seems less common in digital humanities than the three methods mentioned above, there are efforts to apply it to the humanities. For example, in 2012 Anthony Beavers, Mirsad Hadzikadic, and Paul Youngman chaired a conference on “Modeling Complexity in the Humanities and Social Sciences.” Accepted papers explored topics such as the social transmission of language and political change.
• Social simulation models. Through Agent Based Modeling (ABM), researchers simulate how autonomous agents interact within an environment, investigating phenomena such as environmental changes and the emergence of organizations. Although ABM is less common in digital humanities, Michael Gavin (my former colleague at Rice) presented a fascinating paper at Digital Humanities 2013 that explored its potential for understanding the circulation of texts and ideas in seventeenth-century England.

These methods are employed in the sciences as well as social science and humanities (it would be fascinating to consider how and why these methods are used in different disciplines, but that is beyond the scope of this brief overview). Of course, many aspects of digital humanities do not fit under the rubric of computational social science as described by Cioffi-Revilla, such as digital editing, scholarly communication, critical making, and 3D modeling. However, many of these approaches can be associated with the third category, digital cultural heritage.

Digital Cultural Heritage

Digital cultural heritage explores the significance of digital technologies for representing, disseminating, and preserving cultures, drawing upon archaeology, art history, museum and library studies, history, literary studies, cultural studies, and other disciplines. We see an emphasis on digital cultural heritage at Michigan State’s Matrix, which characterizes itself as a “center for digital humanities and social sciences.” In a presentation on digital cultural heritage, Ethan Watrall, Assistant Professor of Anthropology at Michigan State University and Matrix’s associate director, notes the center’s support for interdisciplinary collaboration, educational programs like the Cultural Heritage Informatics Initiative, and the development of tools such as the digital repository and publishing platform Kora. Examples of digital cultural heritage initiatives include Mukurtu, a platform that empowers native peoples to manage and share their cultural heritage, and Open Context, which reviews, documents, and publishes open archaeological data.

Internet Studies

As an interdisciplinary field, internet studies builds upon political science, communications, library and information science, sociology, anthropology, psychology, economics, cultural studies, computer science, and other disciplines to explore “the social and cultural implications of the widespread diffusion and diverse uses of the Internet, the Web, and related information and communication technologies.”[8. William H. Dutton, ed. The Oxford Handbook of Internet Studies (Oxford Handbooks in Business and Management), Oxford UP, 2013.] As Dutton notes, areas of investigation include the development of technologies (such as social factors informing the design of Internet technologies); how people use these technologies (such as online communities, e-commerce, social media, and Internet culture); and policies governing the development and use of the Internet (such as intellectual property and privacy). Research centers in internet studies include Harvard’s Berkman Center for Internet & Society and the Oxford Internet Institute. Internet studies bears some resemblance to new media studies, although the latter draws more from the arts, humanities and design.

Connecting Digital Social Science and Digital Humanities

[pullquote]It’s also important to recognize and respect what distinguishes the humanities and social sciences, such as their approaches to evidence and argumentation and their areas of focus.[/pullquote]

This short survey of digital social sciences is by no means comprehensive; we could also consider scholarly communication, digital ethnography, and fields such as digital anthropology and digital sociology. In any case, it’s clear that there are significant points of intersection between digital social sciences and digital humanities. While digital humanists and digital social scientists already work together in centers such as NULab and Matrix, employ similar methods, and use common tools such as R and Gephi, I wonder if there might be opportunities to deepen collaborations in order to share knowledge and build interdisciplinary community. For example, perhaps a digital humanities journal could run a special issue that explores the two domains, or a conference session on social network analysis could draw from both the humanities and social sciences, or digital humanities centers could host visiting scholars with expertise in digital social science. By developing a richer awareness of how social scientists use digital methods to address research questions, humanists might gain new insights into how they can apply similar techniques to their work—and vice versa.

Deeper collaborations might also enable the two communities to band together in confronting common challenges. Like digital humanities, computational social science faces challenges such as training new scholars, ensuring that research is recognized and rewarded by tenure and review committees, and fostering collaborations between domain experts and computer scientists.[9. David Lazer, et al., “Computational Social Science.”] On the campus level, perhaps digital humanists and digital social scientists can pool resources such as hardware and programming expertise. Likewise, training on tools like Gephi can be targeted at scholars in both the humanities and social sciences (as has happened at places such as UC Berkeley and Fordham), reaching larger audiences and fostering interdisciplinary dialogue. Researchers from one domain could adapt standards, models, or protocols developed in an another.

Even as both communities could benefit from even more contact with each other, I think it’s also important to recognize and respect what distinguishes the humanities and social sciences, such as their approaches to evidence and argumentation and their areas of focus. Humanists could offer important perspectives about the ethical implications of social research or ways to conceive of data, while social scientists could provide insights into statistical methods or suggest how ethnography could add another dimension to a study.  The point is to enhance what discipline each does and open up new areas of inquiry, not to turn the humanities into the social sciences or vice versa. Conversations among digital humanists and digital social scientists could also deepen disciplinary self-awareness, since your own thinking often gets clearer when you explain your processes to someone with a different perspective.

Image Credit: “Social Science,” by Anders Sandberg, used under CC BY 2.0 | Cropped from original.

Note: This post was updated on April 11, 2014, with a minor edit to clarify an institutional affiliation.

This work is licensed under a Creative Commons Attribution 4.0 International License.

---