Quantifying the exclusionary process of canonisation, or How to become a classic of the social sciences

ABSTRACT

This article theoretically and empirically analyses the allocation of recognition and status in a scientific community and of the rank ‘classic’ to better comprehend the selective and reputational mechanisms at work since the beginning of the discipline. The aim is to examine the process of canonisation and the exclusionary logic, which systematically eliminates certain knowledge of specific scholars over time. The role of cultural capital , such as excellent scientific work and its recognition, is taken into consideration in contrast to social capital, such as influential social positions within the field, and to personal attributes such as gender, regarding their relevance for status. A unique database of 957 scholars representing the field of early German sociology was created, and a quantitative analysis was conducted. The results indicate the field was structured very unequally with high rewards for scholars with social capital. However, other than what one might expect, social power does not appear to be relevant for joining the classics canon in the twentieth century. Regardless of the type of capital, while there is great continuity for the most successful male scholars without social positions, women were largely excluded . Exclusionary mechanisms, eliminating the knowledge produced by minorities, may still be at work.

KEYWORDS:

• History of sociology
• women in sociology
• cultural and social capital in the social sciences
• recognition
• statusinequality

Acknowledgements

The author would like to thank Christian Ebner, Julian Go, Barbara Grüning, Marco Santoro, Philip Smith, and an anonymous reviewer for their helpful comments on the earlier versions of this article. The author is grateful to Nina di Guida, Flora Hoth, and Yevgeniy Martynovych for data assistance, and to Sonja Schnitzler for research on the DGS board members.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 It is possible that wide recognition can also stem from a scientific scandal or renowned disagreement with a scholar.

2 Like William Outhwaite (2009), I am using the terms ‘classic’, ‘canon’, and ‘canonisation’ in an informal sense. Here, the aim is not to discuss what the canon is or should be in general and what rules should apply to create a canon or a collection of classics, but to investigate the empirical fact of certain canons made explicit in publications that provide collections of classics, which influence their consumption by students, meaning future scholars. By the same logic, canonisation is thought of as the empirical fact of a process taking place over time in which scholars become canonised, resulting in them becoming classics or members of the (however selected) canon.

3 Here, a ‘curated canon’ is defined as a collection of so-called classics arranged by one or few peer experts.

4 The last reprint of Kaesler (1999a, 1999b) was the 6th revised and updated edition published in 2012.

5 I start by understanding recognition as a synonym to the public mentioning of a person in references, which can result from a high amount of cultural or social capital (or both) in possession of said person. In this sense, recognition is used synonymously to the term impact. Further, I understand status to be to some degree related to recognition/impact.

6 To complete the picture, historically, economic capital or financial independence (which will not be investigated here) were also relevant in terms of scientific success (many scientists had no or very little income from their academic positions or even had to provide private money to buy research instruments) shaping the field and its structure in a certain way.

7 Unfortunately, in the absence of adequate space, I will have to limit myself to describe only the results for a few attributes such as gender.

8 Under ‘populations’, I understand the set of all people listed in the index of people of the ‘Handwörterbuch’ for the initial population, and all classics published in the investigated classical canons for the final populations.

9 This edition was republished multiple times until 2012 and is used widely in German sociology classrooms (Reckwitz, 2002).

10 ‘Recognition’ is understood as the fact of being noted (making an impact) in the form of a reference, which can be but not necessarily must be associated with social capital, as defined in this section. Here, recognition is not meant as literally included in Bourdieu’s testimony on social capital with recognition as a mostly social act of appreciation.

11 The index gives the numbers of all pages a person is named on, including self-citations and editorial mentions such as names of authors on the header of pages. So, authors who have written many articles receive more recognition in the form of mentions just for editorial reasons. To rule out this problem, I have subtracted the pages with authors’ self-citations and editorial mentions. This article is based on these adjusted numbers unless explicitly stated otherwise.

12 It would be desirable to be able to explicitly investigate cultural capital; however, as the number of publications a person has published is not necessarily particularly related to the impact of a person and because it is quite costly in terms of time to gather the concerned data for each of the nearly 1000 people, I limit myself here to this useful reduction (which will be shown capable of revealing much relevant information).

13 I concentrate on institutionalised relationships, since it is very complex to retrospectively investigate and research informal relationships in a systematic empirical way due to widely missing data. For instance, I had to limit the analysis to the board members of the DGS and not membership of the association itself, since membership can only be evidentially stated for certain individual members of early German sociology and systematic data is missing.

14 Still, there may be some interfering variable attributes such as gender, age, ethnicity, or religion, which may play a crucial role and which may alter the outcome of two people in comparison, even though both may have had the same amount of cultural capital to exchange for recognition.

15 Of course, the act of being noted as a cited person by an author simultaneously entails a social aspect since it establishes a social relationship between the author and the cited author; however, this relationship is not necessarily connected to social capital, because a scientific work may be cited by itself, even if the author theoretically may be anonymous or unknown or, if known, may not be relevant to the process of citation.

16 Not to be confused with simple membership. Board membership is a position which is distributed through elections within the association and is a powerful leading position within the community, for instance, such as the president of a sociological association.

17 Within the book, authors recognise peers individually in every article and as a result collectively across the entire book. Some authors are more generous in distributing mentions naming more people on average than others. For the ‘Handwörterbuch’, most of the people are named in the literature sections at the end of the articles.

18 There are some authors whose turnout in terms of recognition is 0 without self-references because they are only named by themselves in their own articles.

19 For instance, author Benno von Wiese was the son of editor Leopold von Wiese, and Rudolf Heberle was the son-in-law of Ferdinand Tönnies, to name only two obvious examples of (in these two cases very close) social relationships.

20 In case of the ‘Handwörterbuch’, they have written roughly one third of the articles themselves.

21 In contrast, probably, although this is not historically certain, the authors were prominently visible to each other only after the book was published, not before, since the writing process is a solitary business and the authors did not have social dependencies of each other. Hence, it is not to be expected that they would cite each other more often based on their mutual authorship.

22 This is a technical term for a relation that can be applied to a set of categories for which not only an equivalence relation can be established but also a relation ordering the categories in a certain way (indicating ‘more’, ‘higher’, ‘larger’, etc. expressions of an attribute).

23 The data bases were set up and (double)controlled by hand. For additional attributes such as gender, I use author information data from the ‘Deutsche Nationalbibliothek’ (German National Library, DNB.de).

24 Not reported here, the attributes age and nationality were analysed for all (sub)sets and I also gathered additional data such as emigration status.

25 For professorship, I distinguished between (a) no professorship, (b) a professorship, but not in sociology, and (c) a professorship explicitly in sociology.

26 For each person listed, the DNB data offers standardised categorical information. Only in the rare case that a person-related DNB dataset does not exist did I consult other sources.

27 For the variable board membership on a nominal scale, it is important to note that I did assign true/false (1/0) to each person who was a board member at some point in her life – even if the person was a board member 20 years after the publication of the ‘Handwörterbuch’. This may complicate the interpretation for the initial population, but it is relevant for the interpretation from the perspective of the later canons.

28 For another 80 the status could not be clarified.

29 Max Weber, who with 40 references is the most cited person, is named 40 times more often than the bottom 642 people. While the average person is mentioned in a maximum of two articles and the median person only in one of the 62 articles, winner Weber is named in 23 articles.

30 The mean of the 80 with unclear status is 1 (median: 1; standard deviation: 0).

31 The high performers have their mean birth year in 1864 (median: 1864, standard deviation 7 years), the underachievers in 1885 (median: 1893, standard deviation: 17 years).

32 Since there are only 35 women in the population (and 13 people with unclear identity), there is only a very small deviation in the results.

33 ‘Only’ as in ‘no other than’ this position.

34 As no women have made it to the ‘Top 20’ or the ‘Top 50’, they would look the same for all people.

35 It also contains men who became classics in other fields, such as Wilhelm Wundt (psychology), Wilhelm Schmidt (theology), George W. F. Hegel (philosophy), or Immanuel Kant (philosophy and political theory).

36 In fact, this subset is aggregated consisting of six subsets, with 40 as the highest number of pages.

37 In this subset, I include authors with 0 mentions, which are special cases originating from the exclusion of self-citations for the adjusted page numbers.

38 Only six are mentioned above the mean of all people.

39 Walther is a special case, since she was cited in three articles by two authors (Geiger and Sombart), of whom one author cited her on three pages in two articles. She thus might as well have been a woman with two mentions (Holzhauser, 2018, p. 103).

40 There may be exceptions concerning social capital, such as Marianne Weber, who was Max Weber’s wife.

41 With one exception, somehow unexpected, the only-authors with adjusted page numbers have no significantly higher recognition level than the people cited only. (Of course, there is a small difference of 2.2 for the authors to 1.9 for the people cited only; however, this difference disappears upon rounding.) One reason for this can be found in the adjustment of the page counts, which has the negative side-effect of not only eliminating self-citations but also relevant citations of these sociologists, who unfortunately are themselves the authors of perhaps the only article in which they would be cited for their cultural capital as experts in the field.

42 And probably also: or Editor-Board-Members – this case did not appear empirically.

43 As two positions have already made a major difference in terms of recognition, it is assumed that two positions can already be counted as a rather high amount of social capital.

44 Since it was argued that authorship or future board membership is not more clearly associated with social capital than with cultural capital, a margin is considered.

45 Those who were not in the handbook are very young in comparison to the average ‘Handwörterbuch’-person. Most of the new people were too young to have been included in 1931.

46 If one calculates only with the people without any positions it is even higher, with 5 out of 10 of subset 1 and a 50% success rate in the 1970s-canon.

47 In subset 2 there are only about 5% and in subset 3 only about 0,2% ending up canonised in the Käsler/Kaesler canons.

48 For the case of Mead, Carreira da Silva and Brito Vieira (2011) discuss the singularity of his canonisation, emphasising that most of his classical work was published posthumously based on a transcript (also Shalin, 2015). Additionally, both men, Mead and Schütz, were quite young in comparison to the common personnel of the ‘Handwörterbuch’.

49 Another highly improbable explanation would be that German sociology changed its contents so profoundly from the 1930s to the second half of the century that the editors, who in this case would have been required to be representatives of a certain sociology and the divide of different sociologies would have had its divide not only over time but also within the cohort between those with and those without social positions, were later no longer relevant. Long story short, this does not seem very plausible.

Additional information

Notes on contributors

Nicole Holzhauser

Dr. Nicole Holzhauser is a postdoc at the Institute of Sociology at the Technische Universität Braunschweig. Her main research interests are theoretical sociology and methodology, social science studies and the history of sociology.