Whose Social Capital?: Citation and Co-citation Patterns of a Fragmented Concept

Abstract

Despite its wide use across the social sciences, social capital research has been regularly criticized as having fuzzy conceptualization, resulting in multiple (potentially conflicting) operationalizations. Our analyses employ bibliometric analytic tools to demonstrate the disconnected structure of social capital research. The resulting visualization illustrates one potential source of the confusion surrounding social capital—that there are many, potentially nonoverlapping versions of the idea rather than one unified social capital concept.

There are few widely used concepts in the social sciences with as much variability in its theorization and operationalization as social capital. In fact, some have argued that “social capital” is used to mean so many different things that it is left without much analytic utility at all (Fine 2002). We argue that this view throws a particularly promising theoretical baby out with the admittedly murky empirical bathwater. Anyone desiring conceptual clarity when using social capital in their own work—or conveying its utility to others—likely has their own mental map of the social capital concept and preferred sources for establishing its bases. Often, these mappings seem incompatible across their uses. The resulting confusion unnecessarily complicates the understanding of social capital’s potential utility for the social sciences.

Given the highly variable and potentially particularized nature of those conceptualizations, we draw on an emerging set of bibliometric analytic tools to provide systematic means for describing the conceptual terrain represented in the myriad ways social capital has been deployed in the literature.

The Mess We’re in

This visualization examines the bibliometric patterns in the citation and co-citation networks (Yan and Ding 2012) of social capital literature by asking how “key papers” are differentially deployed by scholarship in the field. Specifically, we (1) construct a citation network among papers in a corpus of social capital literature (where each paper “receives” a tie from the papers it cites), ¹ (2) estimate the community structure of that network (“communities” identify subsegments of the network that are more likely to cite others within their community and less likely to cite those outside their community), and then (3) assess how papers from each of those identified communities deploy the key papers as indicated by both (a) the distribution of citations those key papers receive and (b) their joint usage as indicated by community-level co-citation rates (“co-citations” enumerate how frequently papers in the corpus jointly cite each pair of papers from the key papers list).

Data and Approach

Our corpus includes 21,160 publications that (1) used “social capital” in the title, keywords, or abstract; (2) were in the ISI Web of Science “Core Collection”; and (3) were published between January 1, 1974, and October 31, 2020. ² These papers receive 0 to 4,095 citations from other publications within the corpus. ³ From that distribution, we identified 12 key papers that received more than 1,000 citations (i.e., are each cited by 5 percent to 19 percent of the papers in the analyzed corpus). ⁴

Next, we identified the community structure of the citation network among our full corpus. ⁵ We then asked (1) how many citations each of the identified key papers received from papers within each community (see the histograms at the top of each panel in Figure 1) and (2) how frequently combinations of those key papers were jointly cited by other papers within each community (see the tile plots at the bottom of each panel in Figure 1). ⁶ In combination, these serve to illustrate how different communities within the social capital literature differentially (1) evaluate the importance of the set of key papers and (2) see the ideas of those key papers as related to one another.

Figure 1.

Key paper citation and co-citation patterns by citation community.

Does This Help Make Sense of Social Capital?

Is social capital a unified concept, or do we have so many conceptualizations, operationalizations, and uses of “social capital” because the authors using the term are fragmented into different communities who each see “social capital” through different—potentially incompatible—lenses?

The evidence presented in Figure 1 suggests that some communities draw on the key papers in similar patterns (e.g., see the similarities in both panels between clusters B and C) and that what differentiates them arises instead because of the citation differences in how they use literature other than social capital (in that particular pairing, given their foci, respectively, on governance and health-related topics). ⁷

However, other clusters demonstrate patterns that substantially differ from one another. Cluster A (with a focus on management) is dominated by a handful of the key citations (Adler and Kwon 2002; Burt 1992; Coleman 1988; and especially Nahapiet and Ghoshal 1998), whereas other clusters seem to be dominated by a single key citation (Coleman 1988 in cluster D with a focus on education and childhood, while Putnam 2000 is highly cited in cluster E, seemingly related to civic engagement topics). Moreover, some clusters also are substantially different from others in the key citations co-cited with others at significantly lower levels (e.g., cluster D papers tend to use a limited number of the key citations per paper, leading to dramatically infrequent co-citation rates within this cluster for many—but not all—of the key paper pairs). Clusters A, E, and F (seemingly related to migration and community cohesion) appear to be relatively ecumenical in the combinations of papers that they jointly draw on to cite social capital ideas despite some differences in the distribution of which papers are cited at all within them. In contrast, while both Coleman (1988) and Putnam (2000) are relatively common citations within Cluster D (see 5th and 10th bars in the top of panel 4), they are significantly unlikely to be cited together within that cluster (see red cell between cited papers 5 and 10 in the bottom of that panel).

This visualization provides initial evidence that there is not a single unified concept of social capital as indicated by citation patterns. Instead, there appear to be multiple social capital concepts in the literature, and those are differentially drawn on by researchers in somewhat disconnected communities. An open question for future research is how and why those different usage patterns arose and perhaps whether there are subdimensions of social capital as an analytic concept that may help account for how scholars have employed the concept with such a wide array of meanings and uses.

Supplemental Material

sj-pdf-1-srd-10.1177_23780231231184766 – Supplemental material for Whose Social Capital?: Citation and Co-citation Patterns of a Fragmented Concept

Supplemental material, sj-pdf-1-srd-10.1177_23780231231184766 for Whose Social Capital?: Citation and Co-citation Patterns of a Fragmented Concept by jimi adams and Kate Vinita Fitch in Socius