'The social origins and schooling of a scientific elite: Fellows of the royal society born from 1900' by Erzsébet Bukodi, John H. Goldthorpe and Inga Steinberg in (2022) British Journal of Sociology comments
We present an analysis of the social composition of the UK scientific elite, as represented by Fellows of the Royal Society, in terms of Fellows' social class origins and type of secondary schooling. From various sources, we have assembled data for 1691 Fellows, representing 80% of our target population of all Fellows born from 1900 onwards whose scientific careers were spent predominantly in the UK. We find that while these elite scientists come largely from more advantaged class backgrounds, it is professional rather than business or managerial families that are the main source of their recruitment—and, increasingly, such families where a parent is in a STEM occupation. Recruitment from working-class families has declined and for most recent birth cohorts almost ceased. The scientific elite is thus now more homogeneous as regards the social origins of its members than it was in the second half of the twentieth century. At the same time, little change is evident in the secondary schooling of Fellows. In all birth cohorts, between two-fifths and a half of all—and over two-thirds of those from more advantaged class backgrounds—were privately educated, although the proportion attending Clarendon schools would seem low compared with that in other elites. A further finding of interest is that some variation in Fellows' class origins and type of schooling exists across different research fields.
The authors state
This paper is the first in a planned series on the sociology of the UK scientific elite, as represented by Fellows of the Royal Society. In this research, we aim to exemplify a new approach to elite studies that we have outlined elsewhere (Bukodi & Goldthorpe, 2021) and that, we believe, treats questions of the social composition, openness, and power or, more generally, effectiveness of elites in a more integrated way than hitherto. We plan to extend this approach to a range of other—including the UK political and economic—elites. But our initial focus on a scientific elite reflects our view that in the context of rising concerns over such issues as global warming, pandemics, and the threats as well as the potential of AI, greater research attention needs to center on scientific elites in view of the increasingly important sociopolitical role that they would seem destined to play.
The study of scientific elites has in fact a long, though rather discontinuous, history, and has been driven by significantly differing interests.
In the nineteenth century, a central concern was with the relative importance of heredity and environment in the creation of scientific genius. A notable controversy on this issue occurred between Francis Galton and the Swiss botanist, Alphonse de Candolle. Relying mainly on the analysis of pedigrees, including those of Fellows of the Royal Society, Galton (1869, 1874; Galton & Schuster, 1906) persisted in the view that heredity was the dominant factor in outstanding intellectual achievement in all fields, science included. Such achievement ‘ran in families’, and those few men from hitherto undistinguished families who rose to intellectual eminence—as, say, through being elected Fellows of the Royal Society—were to be regarded as ‘sports’, whose characteristics would in turn be ‘transmissible by heredity’, as animal and plant breeders well knew (Galton & Schuster, 1906, p. xlii). De Candolle, on the basis of a study (1885) of the social backgrounds of over 300 scientists who were foreign members of European scientific academies—that is, who had gained international recognition—challenged Galton's almost exclusive emphasis on heredity. While accepting that scientific excellence did have some tendency to run in families, he pointed to the part played by ‘family traditions’, and also stressed that the number of those attaining such excellence was strongly conditioned by national cultural and institutional environments that favored or disfavored the promotion of scientific ability.
In this regard, Galton came to agree with de Candolle on at least one point1: the importance of the character and quality of national educational systems. Education in English ‘public schools’ and at Oxford and Cambridge, Galton maintained, was ‘conducted in the interests of the clergy’ and was strongly opposed to science. And although this ‘gigantic monopoly’ of the clergy was slowly yielding, a notable contrast remained with Scotland, where schools and universities were far more conducive to the development of ‘a taste for science’ (Galton, 1874, pp. 259–260).
In the earlier twentieth century, interest in, and the study of, scientific elites became increasingly centered in the United States, and with a changing focus. Pioneering work of influence was that of the German-trained psychologist, James Cattell (see esp., 1915). Cattell thought that the issue of the relative importance of heredity and environment in scientific ability was misconceived: ‘it is like asking whether the extension of a spiral spring is due to the spring or the force applied to it. Some springs cannot be extended by a foot by any force; no spring can be extended without force’ (1915, p. 6). Cattell's main concern was with how far in the US scientific potential was being adequately exploited in the national interest. He established a list of 1154 ‘leading scientific men’ in the US, based on the rankings of others in their fields, and sent these men a questionnaire dealing with various aspects of their families of origin. Analyzing data from the 975 completed questionnaires that were returned to him, he found that the American scientific elite was predominantly recruited from ‘the professional classes’—from the sons (daughters did not figure) of physicians, lawyers, clergy and teachers. Far fewer came from business or farm families, fewer still from the families of skilled workers, and none at all from the families of domestic servants and unskilled labourers. Overall, the professional classes contributed, in proportion to their numbers in the American population, 14 times more leading scientists than did all other classes together. Cattell concluded (1915, p. 8) that while the number of individuals ‘capable of exhibiting genius is limited’, it was evident that ‘many of them are lost through lack of opportunity’, and he underlined the damaging consequences for national economic and social development and prestige.
In what was essentially a follow-up study to Cattell's, Stephen Visher (1947), a social geographer, exploited the American Men of Science directory, which had taken over Cattell's method of identifying the leaders in a field. He focused on the 2607 scientists who were thus ‘starred’ in successive editions of the directory from 1903 to 1943, and as well as accumulating published biographical material on these individuals, sent a questionnaire, similar to Cattell's, to the 1443 who were still alive, obtaining responses from 905. Visher's analyses of this body of data led to results that he regarded as disappointing so far as the recruitment of the scientific elite was concerned. He found little change in the situation described by Cattell. The children of professionals, and especially of professionals living in larger urban areas, continued to be quite disproportionately represented among leading scientists, while there was little increase in the proportion coming from the broadly defined working class.
In Britain, the interest in the scientific elite shown by Galton was not sustained. So far as we are aware, it was not until a study by Choobbasti (2007) that any research at all comparable to that of Cattell and Visher was undertaken. In treating the scientific elite as being constituted by Fellows of the Royal Society, Choobasti's study could be regarded as a direct precursor of that on which we here report. It has, however, major limitations. Choobbasti invited all current Fellows whose email addresses were available on the Society's website as of 2005—525 in total—to complete a web-based questionnaire relating to their social backgrounds. Unfortunately, after two follow-ups, still only 102 Fellows had responded, and since at the time rather less than half of all Fellows provided an email address, it has to be reckoned that Choobbasti obtained information on less than 10% of his target population. Using the Registrar-General's Social Classes as a basis for analysis—despite this categorization being for some time outmoded—Choobbasti felt able to claim (2007, p. 142) that ‘a remarkable proportion of intergenerational mobility’ was revealed in that 65% of Fellows' fathers did not have professional occupations.
In our own research, we do wish, eventually, to take up the issues raised in this earlier work of the extent to which social ‘skews’ in the composition of the scientific elite reflect inequalities of opportunity and thus a potential wastage of talent. However, as we have previously observed (Bukodi & Goldthorpe, 2021), in addressing such issues, in the case of any elite, simply comparing distributions of its members' social origins or of their type of education with distributions estimated for national populations at large, although a common practice, is unduly limited. Comparisons need also to be made with distributions in the ‘pool’ from which elite members are generally drawn. And how far skews result from the processes of selection or election from the pool and how far from the social character of the pool itself is a matter of evident importance.
In the present paper, we take as being of interest, and as prior to other concerns, essentially descriptive questions relating to the social composition of the British scientific elite, as this is represented by Fellows of the Royal Society. In treating the scientific elite in this way, we follow the approach for which we have earlier argued (Bukodi & Goldthorpe, 2021) of defining elites by some kind of institutional or associational reference, so as to avoid making what might be regarded as our own, more or less arbitrary, judgments. We here focus on the degree of social homogeneity of the scientific elite, thus understood, in terms of the social origins, specifically the social class origins, of Fellows of the Royal Society and of their type of schooling. We analyze the extent to which the composition of the elite in these respects has changed over time, and also on the extent to which variation in this composition may exist within the elite—that is, across different scientific fields. In a forthcoming companion paper, we will then extend these analyses to Fellows' university education, at both undergraduate and postgraduate levels.
The social composition of elites has of late attracted growing attention in that an undue degree of social homogeneity has been seen as a source of problems of stasis, ‘group think’ and ‘cultural disconnect’ in elite outlook and policy. In the British case, see, for example, on the governmental elite, King and Crewe (2013) and on elites more generally, Social Mobility Commission (2019). The Royal Society has in fact been itself subject to criticism in this regard (Røstvik & Fyfe, 2018). That is, over its difficulties in moving on from being a ‘gentlemanly space’—or more caustically ‘a musty old boys club’—in which an ethos of gentlemanly conduct proved, if only inadvertently, to be in various respects damaging to women scientists, both within and outside of the Society.
Finally, it will also emerge that starting out from questions of composition serves to bring out issues in the study of elites that seem not always to have been appreciated in earlier work. In particular, the connection between the social homogeneity of elites and the openness of access to them can be more complex than has generally been recognised.
