sociology clusters, 1900-2010

As a follow up to my look at contemporary sociology, I thought I would look at where we’ve been.

I grabbed all the available articles published in the four major general interest sociology journals: AJS (starting in 1900), ASR (starting in 1920 ), Social Forces (starting in 1922) and Social Problems (starting in 1956). I used my RefCliq program to analyze clusters of citations by decade.

Below, I’ve included the major clusters and central author. If you click on the decade, you can see the full reports. I’ve would say the results are strongly correlated with trends in elite sociology, but with a couple of caveats. In addition to problems with how references are reported and recorded, works/areas that span clusters are often misplaced. This problem is likely made worse studying this set of journals, which often have editors that specifically ask that articles, “speak more widely to different areas of the discipline.” Also, articles from the first thirty years of sociology don’t really look like modern articles, so I’m not sure what the reference analysis tells us. That said, I think it’s a quick and easy way to look at major trends in our field.


  • Movements (McAdam)
  • Neighborhoods (Wilson)
  • Networks (Granovetter)
  • Gender/Employment (Kanter)
  • Religion (Steensland)
  • Family/education (McLanahan)
  • Organizations (DiMaggio)
  • Political sociology (Esping-Anderen)
  • Gender (West)
  • Market/class (Bourdieu)
  • Immigration (Gordon)
  • Race (Blalock)
  • Work (Burawoy)
  • Environment (Beck)


  • Movements (McAdam)
  • Race (Wilson)
  • Organizations/markets (Granovetter)
  • Mobility/inequality (Blau)
  • Theory (Giddens)
  • Gender (Kanter)
  • Crime (Hirschi)
  • Political sociology (Skocpol)
  • Theory (Cook)
  • Religion (Durkheim)
  • Networks (Graovetter)
  • Cross-national (Wallterstein)
  • Family (Becker)
  • Race/Ethniciti (Blalock)
  • Social Psych (Kohn)


  • Earnings/Inequalities (Beck)
  • Mobility (Duncan)
  • Crime/delinquency (Hirschi)
  • Marital/family (Mason)
  • Theory (Mead)
  • Theory (Parsons/Garfinkel)
  • Cross-national (Wallterstein)
  • Organizations (Aldrich)
  • Crime (Collins)
  • Segregation/ethnic (Gordon)
  • Movements (McCarthy)
  • Community (Wirth)
  • Suicide (Durkheim)
  • Media/culture (Becker)


  • Occupational mobility (Duncan)
  • Crime/deviance (Becker)
  • Theory (Parsons)
  • Urban (Taeuber)
  • Religion/politics (Campbell)
  • Mental health (Gurin)
  • Community (Mills)
  • Measurement (Blalock)
  • Organization (Blau)
  • Riots (Spilerman)
  • Mobility/inconsistency (Lenski)
  • Development (Wallterstein)
  • Alienation (Kohn)


  • Theory (Merton)
  • Occupations/mobility (Lipset)
  • Political (Adorno)
  • Family/neighborhood (Duncan)
  • Delinquency (Cloward)
  • Mental illness (Hollingshead)
  • Organizations (March)
  • Status Inconsistency (Lenski)
  • Community (Dahl)

Post-war (1945-1959)

  • Theory (Parsons)
  • Occupation/stratification (Centers)
  • Community (Warner)
  • Social Psych (Kardiner)
  • Metropolitan (Mckenzie)
  • Marital/family (Burgess)
  • Personality (Mead)
  • Attitudes (Adorno)
  • Associations (Komarovsky)

Pre-war (1920-1945)

  • Sociology (Linton)
  • Social Psychology (Allport)
  • Community (Park)
  • Family/Delinquency (Burgess)
  • Culture (Ogburn)

Early days (1900 – 1919)

  • Social Psychology (Baldwin)
  • Sociology (Giddings)

A couple of haphazard impressions from someone who has never studied the history of sociology:

  • Social psychology was at the core of the discipline since the very beginning.
  • References to Marx, Weber, Durkheim et al. have never not very been central.
  • What I recognize as the modern sociology coalition seems to have emerged in the 1950s, although you can see the roots of it in the earlier period.
  • Two people dominate their eras: Merton in the 1960s and Duncan in the 1970s. I count about 500 cites to Duncan in the 2,800 articles published in the 1970s. That is about twice the rate of Parsons in his heyday.
  • Gender and race don’t show up as central to distinct cluster until the 1990s. If you are interested in sociology of gender from that period, probably a better way to go is the first decade of Gender & Society (1987 – 1996).  Compare that, for example, to the last five years of the journal.
  • It would be great to turn this into a table showing how topics evolved over time. Perhaps someone could put that together for an ASA Poster Session.

4 thoughts on “sociology clusters, 1900-2010”

  1. It would also be fascinating to compare these cluster analyses based on citations to topic models. It would be a nice way to see what can be learned from one that can’t be learned from the other. I’ve only done a little topic modeling – on economics journals, not sociology – but one thing that emerges are “methods” topics. So you’ll see the emergence of regression discourse as a big topic that links a lot of empirical subfields. In sociology, this exercise might be even more interesting since the methodological range is a bit broader.


  2. This is fascinating! I am particularly intrigued by my own sense that these trends in these topics fail to map onto the jobs in sociology. What we talk about is not, I am guessing, what we hire people to do.

    For example, attending grad school in the 90s as I did, you might forgive me if I thought that doing sociology meant crafting great narratives of political history from case studies. Those jobs in comparative/historical never did seem to materialize, though, did they?


  3. Could you explain the keywords? (I didn’t see an explanation in the original post, either, but perhaps I just missed it. If so, sorry.) Where do they come from? How are they used in generating the clusters, if at all?

    I ask because a non-trivial number of articles have keywords like “yet”, “well”, “subtle,” “large,” and “ways”. The culture cluster for contemporary US sociology has 4 articles with “yet” as a keyword. This makes me wonder if the clusters are combining authors who have similar writing tics. (Or, maybe there’s a contagion of “yet” across author networks. Or, maybe subfields select for authors who overuse “yet.” (“Yetis?”)


    1. Sorry for the confusion. Clusters and order within cluster is entirely based on citation co-occurences. Keywords are only for descriptive purposes–a shortcut to figure out what is going on in the cluster and if a certain work really belongs there.

      Keywords are from article abstracts that cite those works. They are the most common words in those abstracts that don’t occur in more than 20% of all abstracts. Appearing in less then 20% is a pretty good way of getting rid of non informative words like “of”, “and”, “the”, and “an” but it isn’t perfect. I excluded an additional set of words that aren’t useful but don’t occur very often, but probably should add the words you mention to the automatic exclude list.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.