The role of literary fiction in facilitating social science research

The paper begins by highlighting the often-discussed need for interdisciplinary research, questioning the assumed novelty of such collaborations given the ambiguous boundaries between disciplines. It draws on C.P. Snow's concept of "two cultures" to illustrate the historical chasm between the humanities and sciences, but focuses on the nuanced relationship between social sciences and literary studies. While both fields share an interest in human behavior and social relationships, their historical development has fostered distinct methodologies and perspectives. Social scientists, often adopting positivism and quantitative methods, have historically contrasted with literary scholars who initially focused on cultural ideals and elite artistic expression. The paper argues for a deeper understanding of how social scientists already utilize literary fiction, acknowledging the frequent citation of social science theories by literary scholars but questioning the reciprocal influence. The authors aim to move beyond simply advocating for greater interdisciplinary dialogue and delve into the actual ways social scientists engage with literary works in their research and teaching.

The paper acknowledges the relative scarcity of research on the topic. It cites examples of social scientists, like Robert J. Shiller, who have called for greater engagement with literature and fiction, highlighting Cathie Jo Martin's work on the insights offered by literary sources in imagining policy and shaping narratives of governance. However, the authors note a lack of empirical evidence showcasing the extent of this engagement. This lack of readily available data motivates their study, aiming to provide a comprehensive account of how social scientists utilize literary sources, going beyond the limitations of only examining explicit citations in published work.

The study employed a two-step approach. First, a qualitative analysis of social science articles from JSTOR was conducted to identify and categorize references to literary fiction. A list of 30 prominent Anglophone authors and two literary characters (Robinson Crusoe and Sherlock Holmes) was used as search terms. This initial analysis revealed limitations stemming from the reliance on predetermined search terms and the inherent bias of published materials—often omitting the less formal, informal engagements with literary works. The findings from this analysis were compared with Michael Watts's typology of literary uses in economics. Building upon Watt's typology, the authors expanded the categories to encompass a wider range of applications, including literature as argument, literature as context, literature as metonym, literature as decoration, and additional subcategories to accommodate the diversity observed in the JSTOR articles. This analysis highlighted the limitations of relying solely on published works for understanding the true scope of literary influence. Second, a survey was designed and distributed to 13,784 social science researchers at 25 top universities to gather self-reported data on their engagement with literary fiction. The survey was pre-tested, and 875 responses were received, representing a 7% response rate. The sample included researchers from various career stages and across major social science disciplines (economics, sociology, political science, and psychology). This combination of approaches was intended to provide a more comprehensive view of the topic.

The survey results indicated that a vast majority (93%) of respondents acknowledged the important insights literature offers into society and social life. Significantly, 46% affirmed that reading literature played a role in their research questions or project development, and nearly 40% reported using literary sources in their publications. However, the extent and nature of this use varied. The analysis revealed differences across disciplines, with economists expressing slightly more skepticism toward literature's usefulness than researchers in other fields. A striking pattern emerged regarding career stage, with early-career researchers (PhD students and postdocs) exhibiting greater skepticism about using literature in their publications compared to senior researchers. This difference might reflect institutional pressures and the need for early-career researchers to adhere to stricter disciplinary norms to advance their careers. The survey also highlighted the prevalent use of literature in teaching (only 30% never using literary sources in their teaching). The study identified the most frequently cited authors (Orwell, Tolstoy, Dickens, Dostoevsky, Shakespeare among the top five), often reflecting canonical works of realist fiction. More contemporary authors, many female, also appeared, showcasing a diversity of genres. The top ten most-frequently cited works included classics like Orwell's *1984* and *Animal Farm*, Huxley's *Brave New World*, and novels frequently featured in standardized curricula like *To Kill a Mockingbird* and *The Grapes of Wrath*. This suggests the influence of both established literary canons and widely-read educational materials.

The findings address the research question by demonstrating that while explicit citations in social science publications may be limited, a substantial portion of social scientists actively engage with literary fiction in various aspects of their work. The survey findings offer valuable insight into the unspoken influences that shape research trajectories and teaching approaches, which are often overlooked in the formal academic record. The observed generational and career-stage differences highlight the impact of institutional structures and incentives within academia. The preference for canonical realist novels in the recommendations may reflect both established scholarly traditions and the influence of standardized curricula. Overall, the study reveals a more complex and multifaceted relationship between literature and social science than previously documented.

The study provides a nuanced understanding of how social scientists utilize literary fiction, demonstrating its significant yet often implicit role in research and teaching. The contrast between survey responses and the analysis of published work underscores the limitations of solely relying on formal publications for comprehending interdisciplinary influences. Future research could delve deeper into specific disciplinary differences, examine the impact of diverse literary genres beyond realist fiction, and investigate the practical strategies for facilitating greater cross-disciplinary engagement. This could include incorporating explicit training in literary analysis methods within social science programs and providing institutional support for interdisciplinary scholarship. The study advocates for acknowledging and valuing the less formal and less overtly explicit forms of interdisciplinary exchange.

The study's reliance on a self-selected sample of researchers from a specific set of top universities might limit the generalizability of findings to other institutions or geographic areas. The reliance on self-reported data in the survey could introduce recall bias or subjective interpretations. While the JSTOR analysis attempted to be comprehensive, the choice of search terms inevitably imposed limitations on the scope of literature included. Further, the study primarily focused on Anglophone literature, potentially neglecting the contributions of non-English literary traditions.