From Porter to Bourdieu: the evolving specialty structure of English Canadian sociology, 1966 to 2014.

Stokes, Allyson ; McLevey, John

BEGINNING IN THE LATE 1960s, English Canadian sociology was strongly shaped by the publication of John Porter's (1965) The Vertical Mosaic, as well as collective scholarly efforts to build a sociology of Canadian society, to train and hire Canadians, and to develop uniquely Canadian "explanatory stances" (Cormier 2004). Porter's classic work inspired a vibrant set of closely related research agendas, including ethnic stratification, elite studies, political sociology, and new political economy. By the early 2000s, however, sociologists began to debate the institutional viability of the discipline and the seemingly declining levels of intellectual coherence and consensus (e.g., Brym 2003; Curtis and Weir 2002; Davies 2009; Johnston 2006; McLaughlin 2005). Recently, some senior Canadian scholars have sought to bring intellectual coherence and a more or less common agenda back to the discipline, for example, by promoting critical realism (Carroll 2013) or a renewed staples theory (Matthews 2014) as solutions. Their calls have been met with varied responses (Puddephatt and McLaughlin 2015; Stanbridge 2014; Tindall 2014). (1)

Debates about intellectual cohesion and fragmentation have been going on for over a decade. Yet, the current state of the discipline remains unclear, as does the trajectory leading to this point. Of course, Canadian scholars have long been interested in the history of sociology in this country and in what it means to be a Canadian sociologist (Brym and Fox 1989; Felt 1975; Helmes-Hayes 2010; Hiller 2001; Nock 1993; Platt 2006; Riggins 2014). For example, there is research on PhD training and hiring (Goyder 2009; Wilkinson et al. 2013), the effects of policy initiatives on disciplinary reward structures (Siler and McLaughlin 2008), and the relationships between Canadian and other national sociologies (Fournier 2002; Gingras and Warren 2006; Warren 2014). Many have focused on the development of specialties that are, or have been, especially strong in this country, such as political economy and class analysis (Brym and Fox 1989; Clement 1998, 2001; Langford 2013), feminist sociology (Armstrong 1998, 2013; Eichler 2001, 2002), race and ethnicity (Driedger 2001; Ramos 2013), social network analysis (Tindall and Wellman 2001), demography (Wargon 2001), and symbolic interactionism (Helmes-Hayes R manuscript in progress).

While this research tells us much about specific aspects of English Canadian sociology, we know little about changes in the macrolevel intellectual structure of the discipline, or about the origins and extent of fragmentation. Empirical research on the evolution of intellectual networks of English Canadian sociology would go a long way toward resolving ongoing debates about the current state of affairs in the discipline. As relevant work in the sociology of science has demonstrated (Abbott 2001; Crane 1972; Ennis 1992; Evans and Foster 2011; Frickel and Gross 2005; Jacobs 2014; Moody 2004; Moody and Light 2006; Small and Crane 1979), there is much to be gained by bringing changes in the macrolevel intellectual structure of the discipline into focus. Taking a wider view enables us to see beyond the specialties we work in and read, the conferences we go to, and the cultures of our own research teams and departments.

This article directly addresses the evolution of the intellectual structure of English Canadian sociology from 1966 to 2014. Specifically, we ask the following: How did the specialty structure of English Canadian sociology change over time? When did once-dominant specialties disappear or get reinvented, and when did new ones emerge? Did English Canadian sociology fragment? If so, has it continued to fragment more over time, or has it reorganized around a new set of connected specialties? To answer these questions, we use cocitation network analysis (described below), a method that was developed specifically for analyzing the structure of intellectual communities using metadata on publications and citations. We analyze networks extracted from 7,141 sociology articles that were published between 1966 and 2014, and where at least one author was employed by a Canadian university. Our analysis reveals a general trajectory of growth around a core set of specialties centered on Porter's work (1966 to 1984), followed by periods of diversification (1985 to 1994), fragmentation (1995 to 2004), and then reorganization around a new set of dominant specialties that are mostly knit together by the work of Pierre Bourdieu. Overall, we find that the discipline has been most intellectually fragmented in periods where multiple specialties were emerging or decline concurrently (i.e., 1975 to 1984 and 1995 to 2004). Post-2005 English Canadian sociology is more structurally cohesive than any previous period.

In what follows, we provide additional context for our cocitation network approach by describing some relevant work in the sociology of knowledge. We then describe our data and methods, and present findings from each of the time periods. We conclude with a bigger-picture discussion of why English Canadian sociology has evolved this way, and what it means for the way we understand contemporary scholarship in the discipline.

SOCIOLOGY OF KNOWLEDGE

Sociologists and historians of knowledge have long been interested in disciplines, and in identifying and comparing how groups of specialists produce expert knowledge (e.g., Abbott 2001; Camic, Gross, and Lamont 2012; Chen et al. 2014; Collins 2010; Fourcade 2009; Jacobs 2014; Lamont 2009; Whitley 1984). Kuhn's (1962) classic book, which set the agenda for much scholarship on disciplinary paradigm shifts and revolutions, was motivated in part by his efforts to understand why disputes about the legitimacy of fundamental questions and methods are rare in physics, biology, and chemistry while being seemingly "endemic" in sociology and psychology. Kuhn's (1962) argument was that physicists, biologists, and chemists are socialized into "paradigms" that strongly shape the way they perceive and think about the world. In these fields, paradigms changes are revolutionary, with one being replaced by another.

More recently, sociologists of science and knowledge have used the concept of "epistemic culture" to empirically examine shared epistemic beliefs and practices in specialist communities. As Knorr Cetina (1999) and Lamont (2009) have argued, these epistemic cultures can be strongly or weakly held, and can operate at multiple levels, from small collaborative circles to large-scale intellectual networks. Most of this research has been qualitative, enabling careful interpretations and systematic comparisons.

A second line of work, building on pioneering studies of science by Crane (1972), Merton (1973), and de Solia Price (1963), has focused on the structure and evolution of scientific communities (e.g., Bellotti 2012; Kronegger et al. 2011; Leahey and Moody 2014), and even efforts to map connections across all scientific fields (Borner et al. 2012; Boyack, Klavans, and Borner 2005). Almost all of this work has used quantitative and/or network methods to analyze metadata on publications and citations across hundreds or thousands of articles. Cocitation analysis, for example, connects articles and books that are repeatedly cited together in reference lists, with repeated cocitations suggesting relationships between books and articles. Clusters of repeated cocitations suggest specialties, revealing a kind of "titers' consensus" of the state of an intellectual field and helping identify "invisible colleges" of researchers with shared interests and ideas (Small 1973; Small and Garfield 1985; Wallace, Gingras, and Duhon 2009). (2)

Cocitation network analysis has been used to examine the evolution of specialties in cases as different as information science, organizational studies, and the philosophy of science (Kreuzman 2001; Usdiken and Pasadeos 1995; White and McCain 1998), and for predicting future areas of growth in scientific research (Small 2006). It has become the dominant method for mapping scientific communities, in part because of its consistent validity across many cases (de Beilis 2009; see also White 2011).

For the most part, research on epistemic cultures and research on macrolevel intellectual networks have developed independently of one another. While the goals and methods of our article are more aligned with the literature on intellectual networks, we know that changes in the macrolevel intellectual structure of the discipline include but are not limited to changes in specific epistemic cultures; the rise and fall of specialties and invisible colleges; the introduction of new substantive areas; and of course broader political, demographic, and institutional changes. In the analysis that follows, we discuss changes in epistemic cultures where appropriate, and suggest some questions that would be better suited to qualitative work. (3)

DATA AND METHODS

To collect the most comprehensive data possible on research published by English Canadian sociologists between 1966 and 2014, we searched the Web of Science (WoS) for sociology articles that had at least one author affiliated with a Canadian university. (4) We had two options for how to proceed. One seemingly straightforward option would be to identify all scholars who were affiliated with Canadian sociology departments from 1966 to 2014, and collect metadata on all of their publications regardless of the journals in which they publish. Getting such a comprehensive list would be challenging, and would inevitably be biased toward more-permanent members of departments (e.g., tenured professors) over less-permanent members (e.g., PhD candidates, postdocs, contractual faculty, cross/joint appointed faculty). Even with such a list, the citation databases do a poor job of author disambiguation. Searching for articles by a specific person returns articles by anyone sharing the same name. There are a variety of tools for filtering the results down to a set of articles by the intended author, but doing this effectively often requires filtering based on the disciplines that people publish in. This unfortunate reality negates the benefits of using a list of authors to find sociological work published in nonsociological journals.

Given these limitations, we chose to define the boundaries around sociological work as research published in sociology journals, rather than what sociologists publish in any academic journal. However, the WoS indexes sociology somewhat arbitrarily, (5) so we modified the list of journals that are classified as sociology by removing journals that are not primarily sociological (e.g., Anthrozoos), and by adding journals that are sociological but are not indexed as such (e.g., Criminology). Many of our additions to the list were taken from Jerry Jacobs (2015) bibliometric analysis of sociology journal rankings. While some relevant journals are likely still missing, our list includes all of the most well-known English-language generalist and specialist sociology journals, as well as many others. (6) We excluded book reviews, conference papers, and comments, leaving metadata (including full bibliographies) for the resulting 7,141 articles. A full list of the 169 journals in our analysis is too long to print, but is available as an online supplement (at http://www.johnmclevey.com/crs2016supplement.pdf).

Figure 1 shows the number of published articles in our analysis over time. We divided our data into five time periods--1966 to 1974, 1975 to 1984, 1985 to 1994, 1995 to 2004, and 2005 to 2014. After trying a variety of alternatives, we chose these windows because they were long enough for repeated cocitations to occur across many research areas, but short enough to still reveal changes over time. Each time period is indicated in Figure 1 with dashed vertical lines.

We used specialized open-source software (McIlroy-Young and McLevey 2015) to extract cocitation networks for each time period. As previously discussed, this involved assigning a relationship between books and articles if they appear together in the same bibliographies, and then assigning a weight for the number of times each pair of items is cocited. The raw cocitation networks for English Canadian sociology were very large--a combined total of 233,528 nodes (i.e., books and articles) and 8,807,092 edges (i.e., relationships) between them--but the vast majority of these cocitations only occurred once or twice, and are therefore better thought of as noise. To reveal strong relationships while also being inclusive, we removed cocitations that occurred fewer than three times, and then removed isolated nodes. The number of nodes and edges in the resulting networks are reported in Figure 2, alongside the results of a structural cohesion analysis (which is described below).

[FIGURE 1 OMITTED]

Once the networks were constructed, we identified clusters of books and articles that were densely cocited with one another. In cocitation networks, these clusters represent intellectual communities such as subfields and specialties, invisible colleges, intellectual movements, and epistemic cultures. We identified these communities using methods that network analysts have developed for detecting cohesive subgroups, or "communities." We chose the Louvain community detection algorithm developed by Blondel et al. (2008) because a recent methodological article by Wallace et al. (2009) demonstrated that it identifies specialties in cocitation networks more accurately than other community detection methods. Once we had detected communities, we assigned labels by reading the titles and abstracts for publications in the networks, and when necessary by consulting the titles and abstracts for the original articles that cited them. We computed the network layouts using the Ramada and Kawai (1989) algorithm implemented in igraph for R, and differentiated between specialties in the network visualizations using color.

[FIGURE 2 OMITTED]

We used methods for studying cohesion in social networks in order to determine the extent to which the intellectual structure of English Canadian sociology became more or less fragmented over time. We chose to use structural cohesion analysis (Mani and Moody 2014; Moody and White 2003), which Moody (2004) used in his analysis of cohesion in American sociology coauthorship networks. Inspired by Simmel's work on dyads and triads (Simmel and Wolff 1950), structural cohesion analysis focuses on how networks hold together or break apart when nodes are removed. At the highest level, a network consists of one or more "components," which are isolated parts of the network within which all nodes are connected, and where there are no connections to nodes outside the component (Wasserman and Faust 1994). The cohesive blocking algorithm developed by Moody and White (2003) recursively disconnects the network by determining the minimum number of nodes that have to be removed to break apart each component, and then deleting those nodes. The result is a collection of k-components, where k is the minimum number of nodes required for disconnection. In other words, a minimum of two nodes must be removed to disconnect a 2-component, three nodes to disconnect a 3-component, and so on. As the algorithm progressively disconnects the network, it uncovers a nested hierarchy of cohesive groups, each one more deeply embedded in the network. Moody and White's (2003) classic paper provides a full discussion of the sociological theory and mathematics behind structural cohesion analysis and embeddedness.

This method can identify changes in the cohesion of our cocitation networks. If the networks become more cohesive, we should see increases in the percentage of nodes that are deeply embedded in cohesive groups such as 3-, 4-, 5-, and 6-components. On the other hand, if the networks are becoming more fragmented, we should see decreases in the percentage of nodes that are embedded in cohesive groups, and increases in the percentage that are embedded only in 1- and 2-components. We can look for these changes in network cohesion by observing changes in node embeddedness scores. Node embeddedness is equal to the most cohesive k-component that a node is embedded in (Moody and White 2003). In other words, a node with an embeddedness of 4 is embedded in a 4-component. The more cohesive a network is, the more deeply embedded many nodes will be. The less cohesive it is, the less embedded nodes will be.

While our approach has many strengths, there are some limitations of the data that are beyond our control. First, the WoS has a strong English bias, which makes it inappropriate for mapping the intellectual structure of French Canadian sociology. While there are some French articles in our data set, our analysis can only speak to developments in English Canadian sociology. Second, the WoS indexes reference lists from articles, but not books. This may have an impact on specialties that are more book-heavy than article-heavy. However, most sociologists who write books also write articles, and specialties that produce many books (e.g., culture) are very well represented in the cocitation networks. To save space, we do not include in the reference list for this article the books and articles from the cocitation networks.

FINDINGS

Is the Discipline Becoming More Fragmented?

Over time, did English Canadian sociology become more intellectually fragmented, cohesive, or some combination of the two? If the network became more fragmented, the structural cohesion analysis should show an increase in the percentage of nodes (i.e., books and articles) that are embedded only in weakly connected components, such as 1- and 2-components, and that therefore have embeddedness scores of 1 and 2. If, on the other hand, the network became more cohesive, we should see an increase in the percentage of nodes that are embedded in more-cohesive components, such as 3-, 4-, and 5-components, and that therefore have embeddedness scores or 3, 4, 5, and so on.

Figure 2 shows the breakdown of embeddedness scores in each of the cocitation networks. The period 1966 to 1974 is not shown because the cocitation network was very small and disconnected. (7) The other four cocitation networks met the most basic requirement of cohesion: being connected. In each case, more than 95 percent of the nodes in the network were part of a single connected component. However, in the 1975 to 1984 network, 55 percent of the nodes only had an embeddedness score of 1 or 2, and where therefore only weakly connected. The intellectual structure became more cohesive from 1985 to 1994, as the number of nodes with an embeddedness of 1 or 2 dropped to 38 percent. In 1995 to 2004, the intellectual network became more fragmented, as the percentage of nodes with an embeddedness of 1 or 2 increased to 42 percent. Then, in 2005 to 2014, this dropped down to under 35 percent, making the last decade of English Canadian sociology the most structurally cohesive to date (although extremely close to the 1985 to 1994 period).

In short, the cohesion analysis suggests that the discipline was most fragmented in its early decades. As we will discuss below, this is likely due to the emergence of multiple paradigms and specialties that were, in some cases, very weakly connected to one another. The intellectual network became much more cohesive from 1985 to 1994, when, as we discuss below, research agendas diversified but also become more densely connected in the core of the network.

The intellectual network did indeed become more fragmented between 1995 and 2004, going from 38 percent of nodes with an embeddedness of only 1 or 2 to 42 percent. As previously discussed, it was toward the end of this period that sociologists began to debate the intellectual fragmentation and institutional viability of the discipline. It is possible that this increasing fragmentation was due to the gradual decline of many once-dominant specialties, and the uncertain nascent stages of new ones. In other words, the fragmentation of the late 1990s and early 2000s was likely due to a mismatch between the once-dominant English Canadian specialties and those of a new generation.

Perhaps the most interesting finding is that the intellectual network of English Canadian sociology post 2005 was actually more structurally cohesive than in any of the previous time periods. However, there is a distinction between the structural cohesion of network connections and the ideational aspects of cohesion, including individual perceptions, collective identities, and feelings of social solidarity (Moody and White 2003). We do not know if sociologists perceived these changes in the intellectual network, or if increased structural cohesion was accompanied by a renewed sense of collective intellectual identity.

In the sections that follow, we look more closely at the specific time periods to better understand the evolution of the intellectual network, to track changes in specialties and invisible colleges over time, and to try and figure out what is responsible for the increased fragmentation of the late 1990s and early 2000s and the increased structural cohesion of post-2005 English Canadian sociology.

Emergence and Expansion, 1966 to 1974 and 1975 to 1984

Before the 1970s, Canadian sociology was very small and most departments were just being founded (Helmes-Hayes and Curtis 1998). There were 829 faculty members teaching sociology in Canadian universities by 1971 compared to just 61 faculty members teaching in 1961 to 1962 (Helmes-Hayes 2010; see also chapter 1 of Hiller 1982). With this expansion, pressures to publish in the social sciences increased and a diverse range of specialties began developing, both in sociology and in the system of Canadian higher education more generally. (8)

From 1975 to 1984, following the rapid growth of sociology and the other social sciences in Canada, we see the emergence of a number of major specialty areas, including occupational mobility, political economy, race and ethnic stratification, and political sociology. The relationships among these areas are shown in Figure 3. Nodes are colored according to their membership in the clusters identified by the community detection algorithm, and nodes are sized by their degree. At the very center of the network, linking together many of the dominant areas, was John Porter's The Vertical Mosaic and Wallace Clement's The Canadian Corporate Elite: An Analysis of Economic Power.

[FIGURE 3 OMITTED]

Much research produced in the core of this network was macrosociological, focused on class, power, and politics, and was part of the effort to develop both a sociology of Canada and a Canadian sociology. The split between the occupational mobility and political economy specialties seems to reflect both political differences and orientations to American sociology. Accounts by scholars occupying positions in this network (e.g., in Helmes-Hayes and Curtis 1998) emphasize these differences. For example, political economists critiqued occupational mobility scholars for focusing on individual mobility and resource distributions, and argued for a more politically radical view of class relations (Clement 1998, 2001). Occupational mobility and status attainment scholars (e.g., Monica Boyd) used more theory and methods from American sociology (Helmes-Hayes and Curtis 1998). Helmes-Hayes (2010) discusses these political and intellectual tensions in depth in chapter 9 of Measuring the Mosaic.

The cocitation network shows these core specialties in relation to other developing specialties that are rarely discussed as part of early Canadian sociology. Close to the bottom of Figure 3, there is a cluster that represents work on crime, which was weakly connected to symbolic interactionist ethnographies, which were in turn connected to ethnomethodology and phenomenological sociology. The symbolic interactionism cluster was also linked (via Goffman) to German microsociological theory (especially Simmel) and debates about socialization. In short, interpretive sociology in Canada developed alongside the mainstream quantitative and comparative-historical sociology of the day from the start, but is rarely, if ever, discussed in accounts of the discipline at the time (e.g., Brym and Fox 1989).

Diversification, 1985 to 1994

From 1985 to 1994, specialties almost everywhere in the intellectual network diversified, and yet, as shown in Figure 2, became more connected than they were in the previous two decades. Figure 4 shows that the core of the intellectual network was still the new political economy and political sociology, but with a stronger focus on elites, segmented labor markets, and work, occupations, and professions. There were also new connections to research on organizations and to a very weakly knit chain of social theory, including Dorothy Smith, Jurgen Habermas, and Michel Foucault. Political sociologists continued researching party systems, class, and voting, but also conducted more explicitly comparative scholarship on Canada and the United States that was, more often than not, connected to the work of Seymour Martin Lipset.

[FIGURE 4 OMITTED]

Research on crime also diversified and connected to scholarship on law that was being conducted within a broader research agenda on work, occupations, professions, and gender inequality. Other new specialties that developed include work on aging and quality of life, organizations, leisure, social network analysis, and scholarship on families with an emphasis on spousal abuse.

Social theory, especially the works of Habermas, Foucault, and postmodernists such as Lyotard, became much more prominent. They were connected to the core of the intellectual network mainly by cocitations with Bourdieu, who was part of debates about inequality and class analysis, and Dorothy Smith, whose feminist sociology combined (and therefore cocited) Marxism with ethnomethodology. As with the previous period, there was very little overlap between ethnomethodology and symbolic interactionism, except through cocitations of a few well-known microsociologists, including Goffman, Blumer, and Garfinkel.

Fragmentation: Opening Up of a Discipline-Wide Opportunity Structure, 1995 to 2004

By 1995 to 2004, the disciplinary network had completely transformed. In the face of generational change and continued diversification, the connections within and between specialties became much sparser. Specialties from previous decades remained active, but were no longer central. In short, the period between 1995 and 2004 was a time when once-dominant research agendas moved out of the core of the intellectual network, contributing to the fragmented structure shown in Figure 5. Some already existing fields, such as research on crime and juvenile delinquency and leisure continued to develop but did not move into the core of the disciplinary network, likely because many scholars were more oriented to the broader intellectual fields of criminology and leisure studies beyond sociology. The once-dominant research agendas of political sociology were replaced by social movements research (which is not shown because it was not attached to the giant component), with an emphasis on middle range theories such as resource mobilization, framing, and political process. Social capital research and social network analysis were growing and developing into a more diverse research agenda, and social theory was in the early stages of becoming its own area, completely disconnected from a cluster of scholarship on Bourdieu, which emphasized culture and class.

The largest component for 1995 to 2004 was a set of specialties that were knit together by only a few cocited books and articles. Clusters of scholarship on leisure and gender inequality, social movements, organizations, and families were completely disconnected from the giant component (and therefore not shown in Figure 5).

There was a collection of clusters of scholarship on risk society and modernity, power and governmentality, and actor-network theory that could be considered a weakly knit social theory specialty. Some of this social theory scholarship is linked to substantive research areas (e.g., the environment and religious cults), but for the most part it consists of cocitations of Ulrich Beck, Anthony Giddens, Michel Foucault, and Bruno Latour.

[FIGURE 5 OMITTED]

There were considerable changes in the microsociological research cluster, which in previous decades were split primarily between symbolic interactionism and ethnomethodology. In 1995 to 2004, there was more feminist literature being cocited with qualitative data analysis books (e.g., on grounded theory), while classics in symbolic interactionism were much less central. As in previous decades, ethnomethodology was very weakly connected to other microsociological research agendas and qualitative data analysis. In previous decades, these connections were between Garfinkel and Goffman, but in 1995 to 2004, they were between Garfinkel and Dorothy Smith. The most central qualitative methods book was, unsurprisingly, Glaser and Strauss' The Discovery of Grounded Theory.

In the early and middle 2000s, a series of articles on the state of sociology reacted to this fragmentation; the intellectual and institutional future of Canadian sociology was understandably uncertain in the face of generational change (e.g., Brym 2003; Curtis and Weir 2002; McLaughlin 2005; Myles 2003). (9) Yet, as we describe below, this fragmentation was relatively short-lived. The 1995 to 2004 period can perhaps best be thought of as an opening up of a discipline-wide opportunity structure, enabling the development of a new core of English Canadian sociology.

A Reorganized English Canadian Sociology, 2005 to 2014

The fragmentation of English Canadian sociology between 1995 and 2004 provided an opportunity for new specialties to move into the core of the disciplinary network. From 2005 to 2014, the network became more cohesive than it had been at any previous point (see Figure 2), this time centered on the work of Pierre Bourdieu. Distinction is the largest node in the culture cluster at the very center of Figure 6. In addition to many other works by Bourdieu, the sociology of culture cluster at the core of the intellectual network includes classics by Michele Lamont, Paul DiMaggio, Ann Swidler, Richard Peterson, and research by Josee Johnston and Shyon Baumann. This suggests that despite its diversity, the sociology of culture in Canada is focused on culture and inequality, culture "as practice" or "in action," and the production of culture. The centrality of the sociology of culture is surprising in part because in the previous decade it existed mostly as a chain of books by Bourdieu, weakly connected to a vibrant research agenda on social capital.

[FIGURE 6 OMITTED]

Social capital and social networks research also moved into the core of English Canadian sociology. Classic books and articles, including Putnam's Bowling Alone; Nan Lin's Social Capital', Diani and McAdam's edited Social Movements and Networks', Granovetter's "The Strength of Weak Ties"; and Erikson's "Culture, Class, and Connections" form the core of this cluster. Some are cocited with work in the culture cluster, especially "The Forms of Capital," and an emerging body of work on "relational sociology," pioneered by scholars such as Nick Crossley.

There is a considerable amount of scholarship on gender in the core of the network, connected to work on bodies and health, sexualities and sports, social theory, and qualitative feminist sociology. Completely disconnected from this set of research agendas is research on gender inequality at work, and work-life boundaries (located in the upper left of Figure 6). A wide variety of ways of thinking about and researching gender and gender inequality are clearly central to contemporary English Canadian sociology.

The emergence of a social theory specialty was well under way in the 1995 to 2004 period, when works by well-known theorists of risk and modernity, globalization, and governmentality were beginning to be cocited. In the 2005 to 2014 period, these cocitations were primarily among Ulrich Beck, Michel Foucault, Nikolas Rose, and Anthony Giddens. Most cocitations are between multiple works in a theorist's oeuvre, with some cocitations to work in environmental sociology (although less frequently than in 1995 to 2004). Beyond occasional cocitations with social theory, we see the steady growth of environmental sociology in other parts of the discipline, for example, in work on ecology. Culture is also closely connected to scholarship in a wide variety of other specialties, including the sociology of education, debates about public sociology and Canadian sociology, social movements, institutional theory, and the sociology of the professions.

Research on race and ethnicity continues to be one of the largest specialties in the intellectual network. In the 2005 to 2014 period, it appears to be primarily focused on immigration and acculturation. As Ramos (2013) proposes in his discussion of the shift from talking about race to talking about ethnicity in the pages of Canadian Review of Sociology, this could be due to major demographic changes in Canadian society and due to research on the effects of immigration policy reforms in the 1980s.

Classic symbolic interactionist work has once again become central to a cluster focused on qualitative methods. Publications closest to the core are by Blumer, Goffman, Glaser and Strauss, and Corbin. The latter two books, which are on grounded theory, are frequently cocited with research in the sociology of culture. Goffman also plays a key role connecting interactionism and qualitative methods to culture through repeated cocitations with Bourdieu. These cocitations help bring some interactionist scholarship into the core of the network.

It is clear that English Canadian sociology has reorganized around a different set of specialties, that these specialties are more connected than has been the case previously, and that Bourdieu's work knits many but not all of the central specialties and research agendas together. This represents a shift from Porter's dominance in early Canadian sociology to Bourdieu's current dominance. Indeed, Bourdieu's increasing influence is not unique to Canada, and is also the case in the United States (Lamont 2012; Sallaz and Zavisca 2007), and globally (Medvetz and Sallaz forthcoming).

DISCUSSION AND CONCLUSION

We have shown how the disciplinary network of English Canadian sociology has evolved according to a general pattern: emergence and growth, diversification, fragmentation, and reorganization. This general trajectory suggests that the most fragmented periods of English Canadian sociology were 1975 to 1984, when a number of closely related research agendas developed concurrently in the core of the intellectual network, and 1995 to 2004, when those research agendas moved out of the core and a new set of specialties and research agendas were still developing. It is clear that English Canadian sociology was most fragmented at times when multiple research agendas were declining and emerging simultaneously.

The core of contemporary English Canadian sociology is culture, social capital and social networks, gender and sexualities, social theory, education and professions, institutional theories, and qualitative methods. There are other vibrant lines of research as well, including on immigration, environmental sociology, gender inequality, work-life balance, sports and leisure, and the sociology of religion. Whereas many of the early dominant specialties of English Canadian sociology were anchored around John Porter's The Vertical Mosaic, many of the current core specialties are knit together by Pierre Bourdieu's works. These specialties are better connected than many lines of research from previous decades of English Canadian sociology.

The sudden dominance of these new fields is part of several epistemic shifts apparent in English Canadian sociology. For example, the decline of the new political economy and the rise of the sociology of culture, with Bourdieu are the core, could signal a shift in the way that many Canadian sociologists think about class and culture. Similarly, the growth of research on social capital and social networks suggests changes in how sociologists think about social structure and inequality. Additionally, classic political sociology studies of party systems, class, and voting have declined as new research agendas on social movements and contentious politics have developed. Other changes in the intellectual network appear to represent the gradual evolution of specific invisible colleges. The symbolic interactionists, for example, have generally maintained a stable core of classic books by Goffman, Blumer, and Glaser and Strauss, while connecting, at various points in the history of the discipline, to ethnomethodologists, feminist sociologists, social theorists, and sociologists of culture.

The concurrent rise and decline of multiple specialties and research agendas clearly suggests that these changes are at least in part generational, and are likely shaped by the institutional contexts that different generations have faced (see also McLevey, Stokes, and Howard forthcoming). While early generations built a national sociology focused on scholarship about Canada by Canadians, later generations are more likely to experience a "second shift" (Johnston 2006) requiring them to conduct both national and international research, to publish in national and international journals, and to develop both national and international reputations. This would be consistent with Warren's (2014) recent empirical work demonstrating the increasingly international orientation of anglophone and francophone Canadian scholarship. Indeed, Bourdieu's work is increasingly prominent in global sociology (Medvetz and Sallaz forthcoming), and the sociology of culture, gender, social networks, and race and ethnicity are also thriving globally.

Despite dramatic changes, this new set of core specialties retain what many have argued is most distinctive about English Canadian sociology: strong relationalism and a focus on social inequality. Certainly the sociology of culture, social networks and social capital, social movements, and the sociology of gender, race, and ethnicity fit this profile. Perhaps the most notable thing about these specialties, however, is the connections between them. In the most mundane sense, no specialty can be central in an intellectual network without being connected to other specialties. But in this particular case, where the whole intellectual network transformed, it was the emerging linkages between these growing areas that facilitated their concurrent rise. They became dominant together in a newly connected English Canadian sociology.

Understanding these epistemic shifts and substantive changes in the intellectual network of sociology is important for several reasons. First, in the context of ongoing debates about the state of the field, it is important to have a broad understanding that transcends the personal perspectives that any individual person has within their own specialties, research teams, departments, and institutions. This ability to see a broader structure beyond one's personal position has long been one of the strengths of network analysis (Kadushin 2012), and applied to intellectual networks it might help fuel some of the "atypical combinations" of ideas that drive innovative and high-impact research (Uzzi et al. 2013). Relatedly, as Bourdieu (1988) himself wrote about "reflexive sociology," research on knowledge production can be an important tool for developing more rigorous and informed knowledge, and for expanding the scope of sociological expertise.

Knowledge of the evolving specialty structure is also helpful in mapping the contemporary "lay of the land," which is especially useful for junior scholars who are choosing specialties and mentors and planning their careers. It can also help facilitate the formation and further development of research clusters, which the Canadian Sociological Association has been promoting for several years now. Finally, knowledge about the historical trajectory and key strengths of Canadian sociology is important as it becomes increasingly international (Warren 2014), and as sociologists develop research agendas at both the national and international levels.

This article is part of ongoing efforts to understand the nature of sociological research in our country, and the contributions Canadians can make to global sociology. These efforts require empirical knowledge about our discipline's current shape, how it has changed over time, and how it continues to evolve as we move into the future. While our approach has many strengths, it should be considered alongside historical, interview, and survey data. Surveys and interviews could be especially useful because they could tell us whether or not sociologists have developed stronger collective identities as the intellectual network has become more structurally cohesive. Similarly, studies of the evolution of coauthorship networks--perhaps using Siena models (Snijders, van de Bunt, and Steglich 2010)--could provide a view of changes in the social cohesion of English Canadian sociology that complements our analysis of intellectual structure in this article.

References

Abbott, A. 2001. Chaos of Disciplines. Chicago, IL: University of Chicago Press.

Armstrong, P. 1998. "Missing Women: A Feminist Perspective on the Vertical Mosaic." Pp. 116-44 in The Vertical Mosaic Revisited, edited by R. Helmes-Hayes and J. Curtis. Toronto: University of Toronto Press.

Armstrong, P. 2013. "Puzzling Skills: Feminist Political Economy Approaches." Canadian Sociological Review 50(3):256-83.

Bellotti, E. 2012. "Getting Funded. Multi-Level Network of Physicists in Italy." Social Networks 34(2):215-29.

Blondel, V.D., J.-L. Guillaume, R. Lambiotte and E. Lefebvre. 2008. "Fast Unfolding of Communities in Large Networks." Journal of Statistical Mechanics 2008(10):1-12.

Borner, K., R. Klavans, M. Patek, A.M. Zoss, J.R. Biberstine, R.P. Light, V. Lariviere and K. Boyack. 2012. "Design and Update of a Classification System: The UCSD Map of Science." PloS One 7(7):e39464.

Bourdieu, P. 1988. Homo Academicus. Stanford: Stanford University Press.

Boyack, K, R. Klavans and K. Borner. 2005. "Mapping the Backbone of Science." Scientometrics 64(3):351-74.

Brym, R. 2003. "The Decline of the Canadian Sociology and Anthropology Association." Canadian Journal of Sociology 28(3):411-16.

Brym, R. and B. Fox. 1989. From Culture to Power: The Sociology of English Canada. Toronto: Oxford University Press.

Burawoy, M. 2005. "For Public Sociology." American Sociological Review 70(1):4-28.

Camic, C., N. Gross and M. Lamont. 2012. Social Knowledge in the Making. Chicago, IL: University of Chicago Press.

Carroll, W. 2013. "Discipline, Field, Nexus: Re-Visioning Sociology." Canadian Review of Sociology 50(1):1-26.

Chen, S., C. Arsenault, Y. Gingras and V. Lariviere. 2014. "Exploring the Interdisciplinary Evolution of a Discipline: The Case of Biochemistry and Molecular Biology." Scientometrics 102(2):1307-23.

Clement, W. 1998. "Power, Ethnicity, and Class: Reflections Thirty Years after the Vertical Mosaic." Pp. 34-59 in The Vertical Mosaic Revisited, edited by R. Helmes-Hayes and J. Curtis. Toronto: University of Toronto Press.

Clement, W. 2001. "Canadian Political Economy's Legacy for Sociology." Canadian Journal of Sociology 26(3):405-20.

Collins, R. 1986. "Is 1980s Sociology in the Doldrums?" American Journal of Sociology 91(6):1336-55. Retrieved March 25, 2011 (http://www.jstor.org.libaccess. lib.mcmaster.ca/stable/2779799).

Collins, R. 2010. Gravity's Shadow: The Search for Gravitational Waves. Chicago: University of Chicago Press.

Connell, R.W. 2000. "Sociology and World Market Society." Contemporary Sociology 29(2):291-96.

Cormier, J. 2004. The Canadianization Movement: Emergence, Survival, and Success. Toronto: University of Toronto Press.

Crane, D. 1972. Invisible Colleges: Diffusion of Knowledge in Scientific Communities. Chicago: University of Chicago Press.

Cronin, B. and C. Sugimoto. 2014. Beyond Bibliometrics: Harnessing Multidimensional Indicators of Scholarly Impact. Cambridge, MA: MIT Press.

Curtis, B. and L. Weir. 2002. "The Succession Question in Canadian Sociology." Society 26(3):313.

Davies, S. 2009. "Drifting Apart: The Institutional Dynamics Awaiting Public Sociology in Canada." Canadian Journal of Sociology 34(3):623-54.

de Beilis, N. 2009. Bibliometrics and Citation Analysis: From the Science Citation Index to Cybermetrics. Lanham, MD: Scarecrow Press.

de Solia Price, D.J. 1963. Little Science, Big Science. New York: Columbia University Press.

Driedger, L. 2001. "Changing Visions in Ethnic Relations." Canadian Journal of Sociology 26(3):421-51.

Eichler, M. 2001. "Women Pioneers in Canadian Sociology: The Effects of a Politics of Gender and a Politics of Knowledge." Canadian Journal of Sociology 26(3):375-403.

Eichler, M. 2002. "The Impact of Feminism on Canadian Sociology." American Sociologist 33(1):27-41.

Ennis, J.G. 1992. "The Social Organization of Sociological Knowledge: Modeling the Intersection of Specialties." American Sociological Review 57(2):259-65.

Evans, J. and J. Foster. 2011. "Metaknowledge." Science 331:721-25.

Felt, L.F. 1975. "Nationalism and the Possibility of a Relevant Anglo-Canadian Sociology." Canadian Journal of Sociology l(3):377-85.

Fourcade, M. 2009. Economists and Societies: Discipline and Profession in the United States, Britain, and France, 1890s to 1990s. Princeton, NJ: Princeton University Press.

Fournier, M. 2002. "Quebec Sociology: A Discipline and Its Object." American Sociologist 33(l):42-54.

Frickel, S. and N. Gross. 2005. "A General Theory of Scientific/Intellectual Movements." American Sociological Review 70(2):204-32.

Gingras, Y. and J.-P. Warren. 2006. "A British Connexion? A Quantitative Analysis of the Changing Relations between American, British and Canadian Sociologists." Canadian Journal of Sociology 31(4):509-22.

Goyder, J. 2009. "A Behavioural Approach to Ranking Academic Departments by Prestige: The Case of Sociology in English-Speaking Canada." Bulletin of Sociological Methodology 101(1):28-41.

Helmes-Hayes, R. 2010. Measuring the Mosaic: An Intellectual Biography of John Porter. Toronto: University of Toronto Press.

Helmes-Hayes, R. and J. Curtis, eds. 1998. The Vertical Mosaic Revisited. Toronto: University of Toronto Press.

Helmes-Hayes, R. and N. McLaughlin. 2009. "Public Sociology in Canada: Debates, Research, and Historical Context." Canadian Journal of Sociology 34(3):573-600.

Hiller, H. 1982. Society and Change: SD Clark and the Development of Canadian Sociology. Toronto: University of Toronto Press.

Hiller, H. 2001. "Legacy for a New Millennium: Canadian Sociology in the Twentieth Century as Seen through Its Publications." Canadian Journal of Sociology 26(3):257-63.

Jacobs, J. 2014. In Defence of Disciplines: Interdisciplinarity and Specialization in the Research University. Chicago, IL: Chicago University Press.

Jacobs, J. 2015. "Journal Rankings in Sociology: Using the H Index with Google Scholar." The American Sociologist. Retrieved April 16, 2016 (http://link.springer.com/ article/10.1007/s12108-015-9292-7).

Johnston, J. 2006. "The 'Second Shift' of Canadian Sociology: Setting Sociological Standards in a Global Era." Canadian Journal of Sociology 30(1):1-40.

Kadushin, C. 2012. Understanding Social Networks: Theories, Concepts, and Findings. New York: Oxford University Press.

Ramada, T. and S. Kawai. 1989. "An Algorithm for Drawing General Undirected Graphs." Information Processing Letters 31(1):7-15

Knorr Cetina, K. 1999. Epistemic Cultures: How the Sciences Make Knowledge. Cambridge, MA: Harvard University Press.

Kreuzman, H. 2001. "A Co-Citation Analysis of Representative Authors in Philosophy: Examining the Relationship between Epistemologists and Philosophers of Science." Scientometrics 50(3):525-39.

Kronegger, L., F. Mali, A. Ferligoj and P. Doreian. 2011. "Collaboration Structures in Slovenian Scientific Communities." Scientometrics 90(2):631-47.

Kuhn, T. 1962. The Structure of Scientific Revolutions. Chicago, IL: University of Chicago Press.

Lamont, M. 2009. How Professors Think: Inside the Curious World of Academic Judgment. Cambridge, MA: Harvard University Press.

Lamont, M. 2012. "How Has Bourdieu Been Good to Think With? The Case of the United States." Sociological Forum 27(l):228-37.

Langford, T. 2013. "Five Decades of Class Analysis in the Canadian Review of Sociology." Canadian Review of Sociology 50(3):306-36.

Leahey, E. and J. Moody. 2014. "Sociological Innovation through Subfield Integration." Social Currents 1(3):228-56.

Mani, D. and J. Moody. 2014. "Moving beyond Stylized Economic Network Models: The Hybrid World of the Indian Firm Ownership Network 1." American Journal of Sociology 119(6):1629-69.

Matthews, R. 2014. "Committing Canadian Sociology: Developing a Canadian Sociology and a Sociology of Canada." Canadian Review of Sociology 51(2):107-27.

Mayhew, B.H. and R.L. Levinger. 1976. "Size and the Density of Interaction in Human Aggregates." American Journal of Sociology 82(1):86-110.

Mcllroy-Young, R. and J. McLevey. 2015. Metaknowledge: A Python Library for Quantitative Sociology of Knowledge and Scientometric Research. Retrieved March 15, 2016 (http://networkslab.org/metaknowledge/).

McLaughlin, N. 2005. "Canada's Impossible Science: Historical and Institutional Origins of the Coming Crisis in Anglo-Canadian Sociology." Canadian Journal of Sociology 30(1):1-40.

McLevey, J. A. Stokes and A. Howard. Forthcoming. "The Uneven Influence of Pierre Bourdieu in Canada." In The Oxford Handbook of Pierre Bourdieu, edited by T. Medvetz and J. Sallaz. Oxford: Oxford University Press.

Medvetz, T. and J. Sallaz, eds. Forthcoming. The Oxford Handbook of Pierre Bourdieu. Oxford: Oxford University Press.

Merton, R. 1973. The Sociology of Science: Theoretical and Empirical Investigations. Chicago, IL: University of Chicago Press.

Moody, J. 2004. "The Structure of a Social Science Collaboration Network: Disciplinary Cohesion from 1963 to 1999." American Sociological Review 69(2):213-38.

Moody, J. and R. Light. 2006. "A View from Above: The Evolving Sociological Landscape." American Sociologist 37(2):67-86.

Moody, J. and D. White. 2003. "Structural Cohesion and Embeddedness: A Hierarchical Concept of Social Groups." American Sociological Review 68(1):103-27.

Myles, J. 2003. "Where Have All the Sociologists Gone? Explaining Economic Inequality." Canadian Journal of Sociology 28(4):551-59.

Nock, D. 1993. Star Wars in Canadian Sociology: Exploring the Social Construction of Knowledge. Black Point, Nova Scotia: Fernwood Publishing.

Platt, J. 2006. "How Distinctive Are Canadian Research Methods?" Canadian Review of Sociology 43(2):205-31.

Porter, J. 1965. The Vertical Mosaic: An Analysis of Social Class and Power in Canada. Toronto: University of Toronto Press.

Puddephatt, A.J. and N. McLaughlin. 2015. "Critical Nexus or Pluralist Discipline? Institutional Ambivalence and the Future of Canadian Sociology." Canadian Review of Sociology 52(3):310-32.

Ramos, H. 2013. "From Ethnicity to Race in the Canadian Review of Sociology, 1964 to 2010." Canadian Review of Sociology 50(3):337-56.

Riggins, S.H. 2014. "Memorial University's First Sociologist: The Dilemmas of a Bureaucratic Intellectual." Newfoundland and Labrador Studies 29(l):47-83.

Sallaz, J. and J. Zavisca. 2007. "Bourdieu in American Sociology, 1980-2004." Annual Review of Sociology 33:21-41.

Siler, K. and N. McLaughlin. 2008. "The Canada Research Chairs Program and Social Science Reward Structures." Canadian Review of Sociology /Revue canadienne de sociologie 45(1):93-119.

Simmel, G. and K.H. Wolff. 1950. The Sociology of Georg Simmel. Vol. 92892. New York: The Free Press.

Small, H. 1973. "Co-Citation in the Scientific Literature: A New Measure of the Relationship between Two Documents." Journal of the American Society for Information Science 24(4):265-69.

Small, H. 2006. "Tracking and Predicting Growth Areas in Science." Scientometrics 68(3):595-610.

Small, H. and D. Crane. 1979. "Specialties and Disciplines in Science and Social Science: An Examination of Their Structure Using Citation Indexes." Scientometrics l(5-6):445-61.

Small, H. and E. Garfield. 1985. "The Geography of Science: Disciplinary and National Mappings." Journal of Information Science 11(4): 147-59.

Smith-Lovin, L. 2000. "Simplicity, Uncertainty, and the Power of Generative Theories." Contemporary Sociology 29(2):300-306.

Snijders, T., G. van de Bunt and C. Steglich. 2010. "Introduction to Stochastic Actor-Based Models for Network Dynamics." Social Networks 32(l):44-60.

Stanbridge, K. 2014. "How to Commit Canadian Sociology, or 'What Would Innis Do?'" Canadian Review of Sociology 51(4):389-94.

Tindall, D. 2014. "A Distinctive Canadian Sociology?" Canadian Review of Sociology 51(4):395-401.

Tindall, D. and B. Wellman. 2001. "Canada as Social Structure: Social Network Analysis and Canadian Sociology." Canadian Journal of Sociology 26(3):265-308.

Usdiken, B. and Y. Pasadeos. 1995. "Organizational Analysis in North America and Europe: A Comparison of Co-Citation Networks." Organization Studies 16(3):503-26.

Uzzi, B., S. Mukheijee, M. Stringer and B. Jones. 2013. "Atypical Combinations and Scientific Impact." Science 342(6157):468-72.

Wallace, M.L., Y. Gingras and R. Duhon. 2009. "A New Approach for Detecting Scientific Specialties from Raw Cocitation Networks." Journal of the American Society for Information Science and Technology 60(2):240-46.

Wallerstein, I. and N.D. Young. 1997. "Open the Social Sciences: Report of the Gulbenkian Commission on the Restructuring of the Social Sciences." College and Research Libraries 58(4):392-94.

Wargon, S.T. 2001. "Connections: Demography and Sociology in Twentieth Century Canada." Canadian Journal of Sociology 26(3):309-31.

Warren, J.-P. 2009. "The Three Axes of Sociological Practice: The Case of French Quebec" Canadian Journal of Sociology 34(3):803-29.

Warren, J.-P. 2014. "The End of National Sociological Traditions? The Fates of Sociology in English Canada and French Quebec in a Globalized Field of Science." International Journal of Canadian Studies 50(1):87-109.

Wasserman, S. and K. Faust. 1994. Social Network Analysis: Concepts and Applications. New York: Cambridge University Press.

White, H. 2011. "The Sage Handbook of Social Network Analysis." Pp. 271-85 in The SAGE Handbook of Social Network Analysis, edited by J. Scott and P. Carrington. London: SAGE Publications.

White, H. and K. McCain. 1998. "Visualizing a Discipline: An Author Co-Citation Analysis of Information Science, 1972-1995." Journal of the American Society for Information Science 49(4):327-55.

Whitley, R. 1984. The Intellectual and Social Organization of the Sciences. New York: Oxford University Press.

Wilkinson, L., J. Bramadat, R. Dolynchuk and Z.T. St. Aubin. 2013. "Are Canadian-Trained PhDs Disadvantaged in the Academic Labor Market?" Canadian Review of Sociology 50(31:357-70.

Allyson Stokes

University of Texas at Austin

John McLevey

University of Waterloo

Both authors contributed equally to this article. Thanks to Peter Carrington, Neil McLaughlin, Howard Ramos, John Goyder, Kyle Siler, Amelia Howard, Marion Blute, Katja Neves, and Alexander Graham for comments on previous versions of this article, whether in draft form or as a conference talk. Thanks also to Rima Wilkes and the three anonymous reviewers.

John McLevey, Knowledge Integration, Sociology & Legal Studies, University of Waterloo, Waterloo ON, N2L3G1. E-mail: john.mclevey@uwaterloo.ca

(1.) While the debates are clearly about Canada, they tap into longstanding concerns about intellectual coherence in sociology, and its survival and relevance in changing university systems and political climates around the world (Burawoy 2005; e.g., Collins 1986; Connell 2000; Helmes-Hayes and McLaughlin 2009; Smith-Lovin 2000; Wallerstein and Young 1997; Warren 2009).

(2.) This emphasis on intellectual structure differentiates it from other types of citation analysis, which are typically used for assessing influence or information flows (Cronin and Sugimoto 2014).

(3.) Cocitation network analysis trades depth for breadth, and so is less suited to studies of specific epistemic changes than qualitative studies.

(4.) A reviewer pointed out that a better way of defining a sociologist would be "someone who teaches in a university department of sociology." However, for reasons described in detail below, this would introduce a new set of problems.

(5.) We thank an anonymous reviewer for pointing out the arbitrariness of how the WoS indexes sociology.

(6.) A reviewer pointed out that our approach misses research that sociologists publish in nonsociology journals, and might therefore reflect the evolution of sociology journals rather than what sociologists are actually doing. While we are sympathetic to the reviewer's argument, we do not think that it makes sense to emphasize changes in sociological research that occur outside of the discipline's recognized publishing venues. If the content of all the major generalist and sociology journals change over time, then sociology is changing over time. If the content of the sociology journals stay the same but sociologists stop publishing in them, and instead publish in history, philosophy, political science, or cultural studies journals, then something else is happening. The trickiest cases are interdisciplinary social science journals that publish sociological content. In these cases, it would be helpful to differentiate between articles that are published by sociologists and those written by other types of social scientists. We considered screening articles in the more interdisciplinary journals to select those affiliated with a sociology department, but came to the conclusion that it was not a useful approach. First, the institutional affiliation data from the WoS are inconsistent, especially for older articles. It does not always include data on departmental affiliations. Second, even the mainstream generalist and specialist journals sometimes publish articles by social scientists in other disciplines. Third, some authors are trained as sociologists but work in business schools, communications programs, or women's studies programs, etc. While it is relatively easy (if time consuming) to figure out which contemporary authors are sociologists, it is very difficult to do with authors who are no longer active. Given that our focus is 1966 to 2014, the number of authors who are difficult to verify is quite high. Given these limitations, we decided that it was better to be consistent and include all articles that were published in sociological journals where at least one author was affiliated with a Canadian university. Finally, cocitation network analysis strongly prioritizes repeated cocitations. Individual articles that make it into the data set but are not really part of English Canadian sociology are less likely to meet our threshold for repeated cocitations, and would therefore have little if any effect on the construction of the networks.

(7.) The network is available from the authors on request.

(8.) The networks are so sparse and disconnected that there is little value in visualizing them. They are available on request.

(9.) Based on these articles, it is clear that Canadian sociologists were aware of a generational change occurring. We think this is a better explanation for the fragmentation than random disconnections between specialties because the connections within specialties were also becoming sparser. We would not expect to see this, or the development of entirely new specialties, if disconnections between specialties were the cause. It is also unlikely that the fragmentation is the result of growth and natural limits on nodes' degree (Mayhew and Levinger 1976) because the nodes in this network are documents, not people.