The Social Closure of the Cultural Elite. The Case of Artists in Sweden, 1945–2004

This chapter identifies and analyses elites in a social domain where formal positions of power are few and far between – using the fine arts as an example – by combining Weberian closure theory with Pierre Bourdieu’s concept of ‘fields’. Based on a database of more than 14,000 artists, active during the period of 1945–2008, an elite group of 627 is identified. When their social origins are analysed, a first main result of the inquiry is that this aesthetic elite is strikingly similar to other elites who are defined through formal positions of power: the elite are disproportionally drawn from the upper tiers of society. A second main result is that the recruitment to leading – informal – positions in the Swedish field of art displays a process of social closure. Over time, the elite are increasingly populated by individuals with origins in the ‘intellectual’ or ‘cultural’ fractions of the middle and upper classes.


Introduction: Symbolic Dominance and Social Closure
A central concern in research on elites and ruling classes has always been the degree to which social elites manage to monopolise the resources required to reach the positions they hold. There is a normative element to this concern, revealed for instance by the public 'power surveys' conducted in the Nordic countries. Here, the attention is directed towards trying to gauge which social strata are represented within different elite groups. An open, representative elite is considered to be more democratic -and hence 'better' -than a closed elite (cf. SOU 1990:44;NOU 2003:19;Togeby 2003). However, the concern is also clearly related to classical sociological questions on class formation and class reproduction. How do privileged groups -classes as well as elites -manage to occupy central positions and to what extent do they succeed in retaining power, by limiting the access to such positions?
In this chapter, we discuss ways to study the position-takings and the social reproduction of powerful groups. We are especially interested in whether processes of social closure seem to be in operation. Closure theory describes the processes through which insiders circumvent the possibilities of outsiders to participate in the competition over scarce resources on equal terms. The fact that the monopolised benefits can be 'appropriated' on a more or less 'permanent basis' by certain groups is relevant to our study (Weber 1963, p. 44). We also orient our study towards the 'exclusionary' forms of closure, which is when the term 'insider' is equivalent to different kinds of dominant collectives, or social elites (Parkin 1981, pp. 44-47;cf. Murphy 1988). To define the social areas being 'enclosed', we draw on Bourdieu's concept of 'fields' (Bourdieu 1993). Fields are by definition restricted areas with limited access, and therefore probable sites for exclusionary closure.
The most common research strategy in elite studies is to examine individuals who manage to occupy -and retain -formal positions within institutions and organisations (Mills 1999(Mills [1956 ;Giddens 1972;Scott 2008). Research in this vein has generated important insights into these extremely secluded and interconnected worlds (Stanworth and Giddens 1974;Mintz & Schwartz 1985;Useem 1984;Hartmann 2007;Savage and Williams 2008). However, from a field theoretical point of view, this conceptualisation is inadequate for apprehending the distribution of power in society. In fact, a significant part of the 'field of power' -Bourdieu's conceptualisation of the struggle for power within the ruling, dominant class (Bourdieu 1998(Bourdieu [1989cf. Wacquant 1993) -is characterised by the absence of formal positions. This is especially true for the cultural and intellectual fields, where the most important positions tend to be informal. In other words, a focus solely on formal positions would leave these areas entirely unchartered. Arguably, however, expanding the notion of power to include informal positions is also productive in other parts of the field of power. Elsewhere, we have demonstrated that if the economic elite in Sweden is defined by ownership of large-scale capital instead of by holding CEO-positions etc., then half of this elite has been professionally active in positions other than those with formal control over the economy. Moreover, this hidden half of the economic elite have to a greater extent descended from capitalist dynasties compared to the more visible, formal half (Gustavsson and Melldahl 2018;Melldahl 2018;cf. Hansen 2014).
In this chapter, we will describe and discuss ways to locate and identify the position-takings and the social reproduction of elites that largely cannot be apprehended through formal positions. We focus on cultural elites -through a case study of the fine arts in Sweden -since this condition is particularly conspicuous in the areas where artists compete for social prominence. To account for their specific power, methods and sources adapted to the task are required. For analysing social closure, the Swedish field of art is a sufficiently restricted area to enable us to chart all -formal and informal -dominant positions within it during a rather extensive period of time. Moreover, in social spheres such as these, the relation between position-takings and social reproduction is particularly interesting, since the resources being accumulated are symbolic rather than material and consequently not immediately socially inheritable or possible to monopolise. Under such conditions, social closure should be more complicated to achieve.
Our study deals with three challenges. First, identifying the elite in cultural and artistic fields where formal positions are more or less missing is a challenge in its own right. How are 'informal' positions of dominance to be defined? A second, related challenge consists of finding sources and methods adapted to the demands inherent to a historical perspective. Processes of social closure -where the hierarchisation of the active agents in the field is an outcome of different power struggles -become visible only when time is taken into account. Is a certain group able to successfully both claim and retain key positions? A third, underlying challenge is to relate two parts of Bourdieu's sociology that often are employed separately from one another. How are origins in different classes and class fractions in the social space (the topic in Distinction, Bourdieu [1979) related to the capacity to reach and retain elite positions within specialised fields (the theme in Bourdieu's later works, e.g. Bourdieu 1988Bourdieu ], 1996Bourdieu [1992)? Do incumbents of elite positions, by definition more transitory, typically originate from the same more durable classes (cf. Baltzell 2009Baltzell [1958Bottomore 1993Bottomore [1964, p. 52)?
We argue that this way of widening the research domains for the sociology of elites also enhances the understanding of the real magnitude of class reproduction. If only certain elites are investigated, the degree of monopolisation and social closure in a given society is not sufficiently examined. Hence, both formal and informal positions should be subjected to inquiry, during varying political and economic regimes.

Intellectuals as a Disputed Elite
Within the sociology of elites, the position of intellectuals, and especially dominant cultural producers, is ambiguous and disputed. There are remarkably few -if anyempirical studies that examine (informally defined) intellectuals within the traditional framework of elite studies with focus on social recruitment etc. There are some studies that examines these groups in the making, by focussing the social recruitment to elite art schools (e.g. Börjesson 2012), and the social closure of such schools (e.g. Saner and Vögele 2016), but this is at the stage before some of them has a chance to occupy any 'informal' positions of power in intellectual fields. Rather, these intellectuals are at times discussed in theoretical accounts on the definition of the elite. Anthony Giddens, for instance, recognised that leading groups may be defined 'in any given category of activity', including the arts (1973, pp. 119-120). Similarly, C. Wright Mills and Tom Bottomore both thoroughly discussed the role of intellectuals, or more generally individuals with a large impact on the imagination of the general public ('The Names that need no further identification', Mills 1999Mills [1956, pp. 71-72), focussing on their potential significance for the structure and mediation of power in modern societies (Bottomore 1993(Bottomore [1964Mills 1999Mills [1956, chap. 4). However, Giddens, Bottomore and Mills were ultimately unanimous in their verdict that such groups should not be considered as part of the elite proper. This term, they all agreed, should be reserved for groups with a real capacity to rule, to 'those individuals who occupy positions of formal authority at the head of a social organisation or institution' (Giddens 1973, p. 120;cf. Bottomore 1993cf. Bottomore [1964Mills 1999Mills [1956). Hence, whenever power is tied to the potential to yield formal political, bureaucratic and organisational influence, intellectuals are excluded from the elite.
When a different, composite understanding of power is evoked, the position of the intellectuals within the elite is altered. By allowing for multiple -and, more importantly, competing -forms of power, Pierre Bourdieu defines a central role for intellectuals and cultural producers in the overall 'division in the labour of domination' (Bourdieu 1998(Bourdieu [1989, p. 265). In Distinction, the central theme is the opposition between holders of cultural and economic capital, expressed by their diverging cultural preferences, political stances etc., especially pronounced within the dominant class (cf. 1984 [1979], pp. 114-125; 260-267). In The State Nobility this is further nuanced through the already mentioned concept of the 'field of power', a theorisation of the opposition especially between economic and cultural capital, and the groups that produce, distribute, legitimise or in other ways control these forms of capital (1998 [1989], pp. 263-272, 336-339). Consequently, if cultural capital has gained the status of a form of power able to compete with the dominant economic capital, those who produce culture are as central for an understanding of power as those who control the economy. From this perspective, the lack of empirical studies on the recruitment and reproduction of intellectual elites is troublesome, especially given the amount of scholarly attention that economic and political elites have received. In this chapter, we take Bourdieu's multidimensional understanding of power -where intellectuals play such an integral role -as our point of departure in order to examine how dominant positions within the field of cultural production have been a site for historical processes of social closure.
Following this introduction, we describe Bourdieu's concept of 'field' and how we have employed this concept primarily as a selection device. After a few words on the sources that we have used, we present some empirical results to illustrate this approach. In a concluding section, we relate our approach to discussions on social closure, reproduction and the possibilities of examining hidden elites by illuminating less visible subdivisions (individuals populating informal positions) within the already opaque and inaccessible social elite.

Intellectuals Within Fields of Cultural Production
In Bourdieu's sociology, the field of power is in fact a field of fields, or a meta-field of sorts. Within this field, Bourdieu situates, amongst others, the economic field, the field of the state administration and the fields of cultural production (the local fields of art, literature, journalism and science, etc.) (Bourdieu 1998(Bourdieu [1989Bourdieu 1987, pp. 841-843, 850-853). Although intellectuals and cultural producers occupy a dominated position within the overall power elite, like the 'poor cousins' of the economic bourgeoisie (Bourdieu [1979, p. 176; cf. spiritual elites in Moulin 1987Moulin [1967, p. 105), they have the power to take initiative in the struggle for the dominant interpretation of reality. This specific power that intellectuals and cultural producers possess, the 'essentially symbolic power to make people see and believe' -a capacity to impose a certain interpretation of how society should function -is central when social struggles are conceived primarily as battles over legitimacy (Bourdieu 1990cf. pp. 141, 149;cf. Scott 2008, pp. 30-32;Swartz 1997, p. 223).
Through the concept of the 'field of power', a link is established between the various specialised fields and the central concern in traditional elite studies to focus on those 'at the top.' A crucial feature of this concept is that not all of those involved in the various specialised fields are 'effective agents' in the struggles over legitimacy and over the relative power of the various recognised forms of capital that take place within the field of power. Only agents equipped with 'enough specific capital […] to be able to occupy the dominant positions within their respective fields' are capable of making an impact in the field of power (Bourdieu 1998(Bourdieu [1989, p. 264). In other words, the field of power can be conceived of as a representative body, where the dominant agents from the most important fields, at a given place and time, compete with each other over the 'dominant principle of domination' (Bourdieu 1998(Bourdieu [1989, p. 265).
Consequently, in order to define and identify the key players within the field of power, the distribution of specific capital in the various fields first has to be examined. Our approach here is limited to this second task, in other words to scrutinise a specific field in search of the individuals occupying its dominant positions.

Working with Bourdieu's Concept of the Field of Cultural Production
Fields, in Bourdieu's rendition, are relational constructs, defined by the tensions between oppositional, field-specific polarities. Within the field of cultural production, the primary polarity runs between an autonomous, intellectual pole (where l'art pour l'art or 'science for science's sake' are mottos indicating the invisible borders to other areas of activity, a nomos in Bourdieu's vocabulary) and a heteronymous, commercial pole (following the rule 'business is business'). This opposition is a division between two economies: the subfield for restricted production, driven by claims for long-term recognition, and the subfield for large-scale production, driven by a pursuit for short-term gains. In this context, we will only focus on the aforementioned, most autonomous subfield, where production is directed at a market primarily composed of other producers (colleagues as well as competitors) and other groups equipped with the necessary skills and dispositions (i.e. especially consumers from the cultural and intellectual elite etc.) (Bourdieu 1996(Bourdieu [1992.
Within the subfield of restricted production, a second polarity is present. The 'old', highly consecrated defenders of the established order (the reigning 'mandarins' of culture [cf. Bourdieu 1980, p. 212] and the critics and gallerists etc. that are associated with them) opposes the 'young', less consecrated challengers of the tradition (avant-garde artists and their supporting critics and gallerists etc.). This pattern, Bourdieu argues, is cross-temporal. Hence, the new avant-garde always challenge the avant-garde from an earlier period (Bourdieu 1993, pp. 37-40, 46;Bourdieu 1996Bourdieu [1992. The concept of 'field' is applicable to any relatively autonomous sectors of social activity (cf. Kauppi 2003, p. 778;Swartz 1997, pp. 126-127). The subfield for restricted production within the field of art is definitely such a closed and restricted area to which access is limited to 'old' and 'young' artists renouncing the 'economic' economy (cf. Moulin 1967Moulin /1987. Important for our analysis is also that the subfield is a structured and hierarchised area populated by different types of agents, not only by old and young artists in our example of the field of art, but also by different critics, gallerists, art museums, scholarship committees, etc. performing the necessary role of recognising the artists as artists. In the field, these agents are distributed in the same manner as the artists are: some endorse the new, innovative and provocative avant-garde; others cherish the old, consecrated avant-garde. In other words, the critics, for instance, are not in mutual agreement, but rather position themselves in relation to the choices and strategies taken by other critics (Bourdieu 1993, pp. 76-77;Bourdieu 1996Bourdieu [1992. In this sense, the oppositions in the field are embedded within an institutional order (Swartz 1997, pp. 120-121; cf. the explanatory model in Becker 1982).
An important implication is that there are multiple ways to achieve success and gain recognition in a field. This means that there are different types of recognition and consequently different kinds of dominant positions. An analysis of leading or elite positions within a field must therefore include several modes of success to avoid selecting only one kind of influential agent (cf. Swartz 1997, pp. 221-222; see also empirical studies operationalising this kind of definition, for instance on writers [Sapiro 2002] and on central bankers [Lebaron 2008]). As indicated above, this requires sources fit for the task.

The Elite Within the Field and Their Origins in Social Space: Sources and Selection
Since fields are delimited social arenas where different types of agents are competing for recognition in different ways, we have, for the construction of the prosopographical database from which the data in this chapter is drawn, collected archive material from many different institutions within the field of art. As a starting point we have employed criteria of success endemic to the field of art, used and recognised by the agents in the field themselves. Such criteria are expressed, for instance, in curriculum vitae (CV) of artists or in presentations in exhibition catalogues.
Consider the 'typical' CV of any given artist. Because this is a highly standardised genre, certain items, or rather certain headings, tend to be reproduced over time and across national contexts. The typical artist would include a list of his or her exhibitions, giving special emphasis to any solo shows and including information on the gallery responsible for each exhibition (the greater the artist's public recognition, the greater the probability that the list is rather short, containing only a selection of the most important exhibitions). Moreover, if the artist has received any awards, grants or scholarships, they are -regardless of their pecuniary worthbound to be included in the CV, as are any acquisitions of the artist's work by public art councils, museums or other significant public or private collections. Other headings are also frequently found; for instance, the co-optation by a fine art academy or the publication of a book or a catalogue on the artist by a central art scholar.
However, to rely only on information supplied by some artists does not resolve the issue of making an adequate selection of leading artists. Therefore, we have collected data on all artists exhibited at central galleries (etc.). This required thorough discussions with art historians, to complement and value the headings mentioned above using their expertise, in order to, for instance, distinguish which galleries and grants to include or to determine the most important art critics for different time periods. After this procedure, we complemented the data with information on additional informal (e.g. art reviews) and formal positions (e.g. holding the position of professor at important schools of fine art). In Bourdieu's terminology, each and every item included on the 'typical' CV can be regarded as 'agencies of consecration' (Bourdieu 1993: 112), key elements in the sacred act of recognising, selecting and celebrating the real professionals in a world where the professional title is not legally protected (cf. Filer 1986: 58).
There is an important methodological point to this strategy of identifying the leading artists a posteriori. It allowed us to construct the individuals occupying dominant positions at different points in time based on the actual distribution of field-specific capital, as opposed to identifying them a priori, departing from a certain selection of artists included in Who's who or other similar publications such as published dictionaries of artists or surveys on art history. This strategy permitted the inclusion of artists that made real contributions to the field, but were left out of sources of institutional memory of this kind. It is of significance to note that, in this field, formal positions (heading 8 below) are less important than informal (headings 1-7). Our indicators on the artists' field-specific capital in the prosopographic database are organised in the following headings: 1. Exhibitions 2. Reviews 3. Grants 4. Acquisitions 5. Commissioned Public Art 6. Transnational Investments 7. Publications 8. Positions Each heading involves agencies of consecration that at given points in time occupied different positions within the field. The prosopographic database contains, for instance, information on older as well as younger galleries; domestic exhibitions as well as exhibitions abroad; market-oriented activities and institutions (pieces of work traded as acquisitions, commercial agents such as galleries, etc.) as well as state-supported activities and public institutions (commissioned public art and financial support for artists); and field-specific recognition (academy co-optation, peer-appointed teaching positions) as well as public acknowledgment (large museum exhibitions, publications, reviews, etc.). For every heading, we collected all obtainable data for every year between 1945 and 2004. As an example, the heading 'Exhibitions' contains data on all artists exhibited at central galleries (44 galleries in the database), museums (2) and Swedish Konsthallar (6 in the database).
Due to this archival strategy, the number of artists in the prosopography is substantial. For instance, 3685 artists were exhibited at the 44 galleries alone during the period. In total, the prosopography comprises information on activities of 14,500 artists within the field. In other words, the amount of data to work with was similar to that of Raymonde Moulin in her cross-sectional study of 18,600 French artists in the early 1980s (cf. Moulin 1997Moulin [1992).
To identify the elite within the field, among the 14,500 active artists, we followed principles similar to those employed by Moulin by successively sharpening the criteria. Artists who we defined as'successful' (the 'highly visible' in Moulin's terminology [1997Moulin's terminology [ (1992], or 'integrated professionals' in Howard Becker's [1982]) either upheld the few leading formal positions in the field (e.g. professor at the Royal Institute of Art) or, more importantly, made a mark through informal activities in different areas of the field. The end result was a group of 627 elite artists who have in common that they made a significant impact on the Swedish field of art at some point during the 60-year timeframe.
For an examination of the social closure of dominant positions within cultural fields, however, it is necessary to relate the achieved positions within such fields to origins in the social space. Often when elites are studied, data on the social origin is extracted either from questionnaires or biographical handbooks (such as Who's who and similar publications). We have instead studied the social origins of the elite artists as they were recorded in the birth and baptism registers from the parishes where the artists were born, which were available at regional archives or in some cases still kept at the parish offices. Although it is a time-consuming approach, this strategy has three main advantages. First, it is not biased towards already known artists, who more frequently are to be found in Who's who or in artist dictionaries. Second, the origins of all artists are measured at the same point in time: at the time of their birth. Third, accuracy is drastically improved: so far, we have found the social origins of 89% of the artists, while in studies using Who's who, the numbers are often significantly lower (cf. Bourdieu 1998Bourdieu [1989, pp. 342-343). We used the parents' occupational titles as indicators of the artists' social origin (although in most cases, due to the data, information was only available for the father).

Social Recruitment to Dominant Positions in the Swedish Field of Art, 1945-2004
Social closure is a process where a group monopolises the conditions for its own reproduction. In order to reveal such a process, the social composition of a group should prove to be increasingly distinctive. In this section, we will analyse changes in the recruitment to dominant positions in the field of art during the period 1945-2004. Since careers span over vast periods of time, several artists were active during large parts of this timeframe, especially the selection of artists that we have focused on who are defined by their significant and extensive activity in the field of art.
To facilitate analyses of changes over time, we took two measures. First, the 60-year-long researched timeframe was divided into four shorter periods of 15 years. Second, the artists were allocated to the specific 15-year period during which they first achieved sufficient recognition to reach elite status in the field, in order to capture different artistic generations. Hence, in the following tables and figures, each and every artist is found in one and only one 15-year period, even though most of the artists did in fact surpass our analytical 'thresholds' by being active during the subsequent periods as well (on generations, cf. Mannheim 1952Mannheim [1928; for applied studies of generations in fields of cultural production, see Sapiro 2002 andEkelund andBörjesson 2005).
To retain a solid linkage between achieved positions in the field and origins in the general social space, the social origin of the artists were not measured solely with reference to classes but also to class fractions. This incorporated into the analyses the crucial point in Bourdieu's apprehension of the fundamental lines of division within modern societies, namely that not only the volume of capital (i.e. classes) has a stratifying effect, but also the composition of that capital (i.e. class fractions), and especially so in the upper regions of the social space (Bourdieu [1979. As will be demonstrated below, this two-dimensionality proved vital for our results (for a description of the classification into classes and class fractions, see Gustavsson 2002;cf. Gustavsson and Melldahl 2018). Table 18.1 shows the recruitment to dominant artist positions in the field of art from different class origins in the social space across the entire 60-year timeframe (Table 18.1).
Looking first at the representation of artists with a popular class origin, their share amounts to approximately 24% (of the individuals for whom we could identify the social origin). This is roughly the same as for the overall elite in Sweden, when defined through formal positions (SOU 1990:44, pp. 321-322). Conversely, combining the shares for the dominant class and the middle classes, Table 18.1 reveals that 76% of the artists originated from privileged Swedish households. This figure is similar to, or even surpassing, the share of individuals from privileged backgrounds found in the theoretically vastly different (and far more researched) economic elite, in Sweden as well as in various other countries (on Sweden, see Göransson 2007, p. 126; on elite recruitment in the UK, Denmark, France and Germany, see Ellersgaard et al. 2013;Hartmann 2000).
Turning instead to the question of change, Fig. 18.1 testifies to a remarkable stability in the recruitment pattern of elite artists, when the four 15-year periods are compared. The size of the working class (i.e. popular classes) in Sweden has hovered around 50% throughout the twentieth century, although the composition of the class has changed over time (Ahrne et al. 1996, pp. 61-64;Bengtsson 2010, p. 25). Within the artist elite, however, the share of those with popular class origin (black bars) is at most 28% (in 1975-1989Fig. 18.1c). In other words, even though Sweden had a social democratic government throughout the majority of time under study (all but nine years: 1976-1982 and 1991-1994), and a proclaimed social reformist cultural policy since 1974, individuals originating from the popular classes were constantly underrepresented in dominant positions within the field of art. a. 1945-1959 b. 1960-1974 c. 1975-1989 d. 1990-2004 Note: The figures are based on information regarding the 89% of the artists for whom we have located social origins. Hence, in this field, the social and cultural policies do not seem to have had much impact. However, as argued above, the class origins of the dominant cultural producers only provide a partial linkage between the field and the general social space. To provide a more detailed picture of the artists' origins, we need to also look within the social classes and scrutinise differences between class fractions. In Fig. 18.2, only such horizontal distances are highlighted (the fractions of the dominant class were here conjoined with the fractions of the middle classes). They are illustrated by three fractions: 'cultural' (professors, artists, teachers, etc.), with a capital composition dominated by cultural capital; 'professions' (physicians, engineers, nurses, etc.), with a balanced capital composition; and 'economic' (managers and executives, shop-keepers, bank-clerks, etc.), with a capital composition dominated by economic capital.
This operation dramatically alters the picture. If Fig. 18.1 conveys social stability, Fig. 18.2 reveals a pattern of clear social change with one apparent direction: artists with origins in the cultural fraction (white bars) increase with each successive 15-year period (from 40% in 1945-1959 to 52% in 1990-2004). Simultaneously, artists with origins in the economic fraction (black bars) decrease over time (from 35% 1945-1959 to 18% 1990-2004). This is in line with studies by Raymonde a. 1945-1959 b. 1960-1974 c. 1975-1989 d. 1990-2004 Note: The figures are based on information regarding artists with origins in the dominant class and the middle classes. Moulin who argues that artists in all western countries, to a large extent, originate from what in our terminology could be labelled the cultural fraction of the higher strata of society (Moulin 1997(Moulin [1992, p. 275; cf. results showing a more general tendency of field-specific auto-reproduction in Flemmen et al. 2017).

Narrowing the Eye of the Needle
One way to further nuance the analysis of how the accumulation of field-specific capital is related to social closure and social reproduction is to change the size of the constructed elite, making it successively more restricted. In our data, this is possible by raising the thresholds in the data for being qualified as elite, thereby diminishing the number of artists in the resulting elite. When only those who have accumulated extremely large amounts of field-specific capital are included, the size of the elite decreases from a total of 627 individuals to about half (254, with some 65 highly selected artists making their debut in each period). The results are clear: in all but one period, the representation of elite artists with social origins from the popular classes decreases (from 25% to 15% in the first period; from 19% to 12% in the last) and, in all but one period, the representation of artists originating from the cultural fractions of the dominant class and the middle classes increases (from 42% to 43% in the second period; from 52% to 60% in the last). These results differ from Raymond Moulin's study in France, where popular origins actually were more common in the highest strata than in the lower (Moulin 1997(Moulin [1992, p. 279; cf. the discussion on the 'over-selection' of individuals less equipped with inherited capital at the top of symbolic hierarchies in Passeron 1990 [1970]). In the Swedish case, it seems as if the acquisition of field-specific capital indeed is related to the volume -and especially the kind -of inherited capital, indicating an accentuated pattern of social closure when the elite is further narrowed down.

Making Invisible Elites Visible
We have in this chapter argued for the need to include informal positions in sociological research on power. Through a case study of the Swedish field of art between 1945-2004, we have described and discussed sources and research procedures that have allowed us to identify individuals holding dominant positions, formal as well as informal, in this field. Here, we will briefly summarise our main results and discuss the implications of our findings with regards to understanding the social closure and the social reproduction of elite positions.

Increasing Social Closure of Dominant Positions in the Field of Art
Social reproduction and social closure highlight the processes through which groups manage to sustain an achieved position of certain privilege over time by excluding others. Our results show that the social recruitment to the artist elite in Sweden is similar to other elite spheres in Sweden, as well as in other countries, in terms of the under-representation of individuals with popular class origins. Through a historical perspective often lacking in elite studies, we were able to show that this pattern in the social recruitment from different classes has been remarkably stable over time, indicating the efficacy of the efforts of dominant groups to monopolise access. However, this stability is countered by a marked change when horizontal class fractions are used as a complement to vertical classes. This illustrates the relevance of incorporating a certain precision in the analyses of the social recruitment to areas that are unattainable for most people. As our results show, a solid and stable overrepresentation of individuals with origins in the dominant class may well obscure a pronounced and intensifying closure based on class fraction origins: the people reaching dominant positions within this cultural field are increasingly of the same cultural kind (i.e. auto-reproduction). Thus, in order to be productive, Weberian analyses of social closure benefit from being complemented with a Bourdieusian interest in the differentiation within larger social aggregates, in order to achieve a more precise definition of what characterises the insiders.

The Field Concept as a Tool to Locate a Hidden Elite
In our study, we have tried to match the aforementioned precision in our definition of the insiders with a similar degree of precision regarding the restricted social area being enclosed. For this task, Bourdieu's field concept was utilised. When the incumbents of informal power are sought, a selection device attuned to highlighting different parts of any sphere of activity is needed, since the population cannot be defined a priori. In fields, agents -producers, distributors and consumers alikecompete with each other over recognition, or field-specific capital. Through the forces of attraction and repulsion at different, oppositional poles in the field, Bourdieu sees recognition as something that can be won in several ways. The strength of this approach is that it identifies areas where individuals who potentially occupy dominant positions are active. However, because the dominant positions do not come ready-made but are defined in relation to the factual distribution of fieldspecific capital, the approach is also fairly labour intensive. For this study, we collected information on all artists active at important 'consecrating instances' during a 60-year timeframe, in order to apprehend the distribution of field-specific capital (we used information on 14,500 artists to identify 627 elite artists). In other words, access to productive sources of data is necessary. To define the inhabitants of the field, unprinted material from museums, galleries, scholarship committees etc. was used. Similarly, to study the social closure of the field, unprinted material about the population's social origin from church archives in different parts of Sweden was extracted. The strategy to identify different kinds of elites (e.g. 'old', 'young', 'autonomous', 'heteronymous') within one elite group is applicable more generally and not restricted solely to the field of art or other intellectual fields. Even though the logic of intellectual fields might be peculiar and idiosyncratic in several ways, perhaps especially so in the manner they, as Bourdieu argues, glorify 'great individuals' and in the degree to which this occurs (Bourdieu 1993, p. 29), there are also similarities to be observed in other social arenas that are, at face value, very different from intellectual fields. The struggle between the 'old' and the 'new' is arguably a crucial component in most specialised arenas, whether these are structured around perceptions on religion, bureaucracy, medicine, law or the economy. In certain ways, there are striking similarities between how Bourdieu portrays the temporal logic of intellectual fields (e.g. the transition between and within aesthetic regimes) and Schumpeter's description of structures of transformation in the capitalist economy (e.g. the transition between and within economic phases). This similarity extends even to the cult of the genius, which in Schumpeter's work is present in the guise of the 'young' radical entrepreneur and his/her battles with the 'old' conservative companies (1975 [1942], p. 132).
In other words, the idea of different regulating and legitimising principles that compete with each other within a specialised social sphere is not restricted to the numerically rather insignificant field of art. As we have already argued, this indicates that the basic postulates behind the concept of 'fields' should be regarded as more generally applicable, although the analytical and empirical strategies for apprehending the competing principles, as well as the means for identifying the actors and institutions embodying and representing them, naturally have to be adapted to the sphere in question.

A more Complete Picture of Class Reproduction
Elites are opaque. Consequently, the role for the sociology of power and elites is to identify them and subject them to scientific inquiry in order to illuminate, among other things, the extent to which positions of social power are sites for (upper class) reproduction through processes of social closure -that is, the degree to which the functional division of legitimate authority corresponds with structures of persistent social inequalities. Our results indicate, and the research strategy described in this chapter illustrates, that the degree of social reproduction often encountered in elite studies probably is underestimated. The social recruitment to the informal positions that we have highlighted is more or less identical with the recruitment patterns to the formal positions that elite studies hitherto dominantly have emphasised. Regardless of whether the positions of power are formal or informal, individuals from privileged backgrounds have disproportionate access to them. Hence, elite studies should seek to identify invisible elites in informal positions as well as visible elites in formal positions in all fields of elite activity -not just in the field of art.