Class and status: on the misconstrual of the conceptual distinction and a neo‐Bourdieusian alternative

Key takeaways

content: "@paalgardflemmenClassStatusMisconstrual2019" -file:@paalgardflemmenClassStatusMisconstrual2019

Reading notes

Imported on 2024-06-10 13:09

⭐ Important

• & We discuss in particular Chan and Goldthorpe’s influential, yet largely unchallenged, attempt to reinstate a strict version of the class-status distinction, mounted as an attack on ‘Bourdieusian’ accounts. We argue that this is unconvincing in two respects: There are fundamental problems with their conceptualization of status, producing a peculiar account where one expression of status honour explains the other; in addition, (p. 816)
• & their portrayal of the Bourdieusian approach as one-dimensional is highly questionable. In contradiction of a reading of Bourdieu as discarding the class-status distinction, we develop an alternative, neo-Bourdieusian account that recognizes class and status as distinct aspects of stratification, thereby allowing for a subtle analysis of their empirical entwinement. (p. 817)
• & fruitfulness of this approach is demonstrated by analysing the homology between the space of lifestyles and the social space through Multiple Correspondence Analysis of unusually rich data about lifestyles. Importantly, we highlight the relative autonomy of these spaces: Although they exhibit a similar structure, they do not overlap completely. (p. 817)
• & Although Weber’s (2010) tripartite model of stratification in terms of class, Stand and party is widely known, very few accounts of contemporary stratification abide by it. (p. 817)
• & The sociology of Bourdieu has emerged as a central point of reference in the research field, purporting to ‘rethink Max Weber’s opposition’ (Bourdieu 1984: xii) and unite economic and symbolic analyses in studying stratification. However, critics have debated whether this amounts to a new form of reductionist ‘economism’ (Alexander 2003; see also the discussion in Lebaron 2003), or whether it involves an account of economy and culture that is ‘too unitary’ (Crompton and Scott 2005). (p. 817)
• & They argue that Bourdieu presents a particularly influential, if highly problematic, one-dimensional account of stratification that discards Weber’s distinction by pure ‘theoretical fiat’ (Chan and Goldthorpe 2010: 14). But despite the significance and impact of these points, there has been remarkably little systematic engagement with them (though see Atkinson 2015; Wuggenig 2007). (p. 817)
• & We demonstrate that their rendition of the class-status distinction is unconvincing, producing a peculiar account where one expression of status honour explains the other. Turning to Bourdieu, we show how their portrayal of his position is equally problematic. Although we recognize that Bourdieu could be read as suggesting that ‘“status groups” based on a “life-style” are [...] dominant classes that have denied or, so to speak, sublimated themselves’ (Bourdieu 1990b: 139), we develop an alternative account from his writings. (p. 817)
• & Rather than discarding the class-status distinction, we show how Bourdieusian concepts allow for clear analytical distinctions to be drawn between the economic and symbolic dimensions of inequality, as well as the means to unpack their empirical entwinement. (p. 817)
• & we want to characterize the Stände situation as resulting from the typical integral part of life, in which the fate of men depends on a specific positive or negative social assessment of honor’ (Weber 2010: 142). (p. 818)
• & Weber was less definite about their actual interrelationship: Although status could stand in opposition to property, it could also be derived from it, so that ‘the differences between the classes can be combined with the differences between the Stände in numerous ways’ (Weber 2010: 143). (p. 818)
• & Focusing particularly on the occupational structure, Blau and Duncan have studied social stratification by deploying a unidimensional, continuous measure of occupational prestige – the Socio-Economic Index (SEI) (Blau and Duncan 1967; Duncan 1961). (p. 818)
• & Related to this, Laumann has measured social stratification by mapping inductively the associations between the occupations (p. 818)
• & of one’s friends, partners and parents-in-law (Laumann and Guttman 1966). (p. 819)
• & Similarly, the CAMSIS scale, developed by the Cambridge school (Bottero and Prandy 2003; Prandy 1990; Prandy and Lambert 2003), involves a mapping of networks of social interaction (i.e., patterns of friendships and partnerships) that are regarded as giving rise to relations of social closeness and distance. Notwithstanding important differences between these approaches, they all share a propensity to merge class and status in constructing hierarchically ordered scales that are thought to reflect the primary bases for structuring social inequalities. (p. 819)
• & There have always been difficulties with the Weberian concept of status, as it would appear to simultaneously refer to pre-capitalist principles of stratification, as well as the role of ‘culture’ in capitalist systems of stratification. For this reason, Giddens (1981: 80, 109) has suggested replacing the concept with an emphasis on the role of ‘distributive groupings’ in class structuration, that is, the extent to which similarities in lifestyles shape the formation of social classes as distinct groups. (p. 819)
• & The emphasis on status and culture in class analysis waned with the rise of academic Marxism, and the role of culture was effectively sidelined in the renewal and refinement of quantitative analyses of class structure (Goldthorpe and Marshall 1992). Critics have regarded this as a ‘minimalist’, ‘economistic’ programme of class analysis (see especially Devine and Savage 2005; Savage 2000), and the ensuing ‘cultural turn’ in class analysis has in important ways involved drawing more inspiration from Bourdieu’s sociology of class (Reay 2011). (p. 819)
• & Although Bourdieu is not easily squared with any one tradition of classical sociology (Wacquant 2013), some have regarded his approach as clearly connected to the multidimensional Weberian position in stratification theory (Brubaker 1985; Coulangeon and Duval 2015; Weininger 2005). (p. 819)
• & positing the two as distinct independent variables with separate explananda. According to this view, economic security and prospects are stratified more by class than by status, while the opposite is true of cultural consumption and lifestyles. (p. 820)
• & Chan and Goldthorpe view the status order as ‘a structure of relations of perceived, and in some degree accepted, social superiority, equality, and inferiority among individuals’ (2010: 11). They underline that the status order is by definition ‘a set of hierarchical relations’ (2004: 383). (p. 820)
• & Crucially, this hierarchy is regarded as expressed in differential association, especially in more intimate kinds of sociability. (p. 820)
• & Although seeking to reinstate Weber’s distinction, Chan and Goldthorpe devote surprisingly little attention to his writings about status. In fact, the key inspiration for their concept and measuring of status appears to come from the American sociology of stratification. (p. 820)
• & This scaling procedure reveals differences and similarities in friendship patterns (2010: 29), producing a map where occupations appear close if they are similar with respect to the occupation of friends, whereas distance reflects dissimilarity (for an instructive introduction to the method, see Bartholomew, Steele, Moustaki and Galbraith 2008: 55–81). Their multidimensional scaling solution is three-dimensional, yet they believe that only the first dimension reflects status. The second dimension is seen as reflecting segregation by sex on the labour market, and the third differences in situs. (p. 820)
• & The first dimension, the purported status scale, places manual occupations on one side, and professionals, associate professionals in business, specialist managers, scientists and teachers on the other side, while a manual/non-manual continuum can be found in between (Chan and Goldthorpe 2004: 388). However, it is not obvious why these differences in friendships reflect a hierarchy of esteem. (p. 821)
• & Chan and Goldthorpe justify their interpretation of this as hierarchical for three reasons. First, the scale orders occupation according to the degree of manuality, so non-manual occupations are separate from manual ones. Second, within the non-manual occupational groups, those who more often possess professional qualifications are separate from those who do not. Third, it provides clear ‘echoes’, as they put it, of the British status order of times past (2004: 389–90). (p. 821)
• & Chan and Goldthorpe provide no clear reason why the gender divide in differential association should not be considered in terms of status, nor do they argue why differences in ‘situs’ cannot be regarded as differences in status. For them, it is axiomatic that status is a hierarchical phenomenon, but they provide no clear argument as to why their first MDS dimension is hierarchical, whereas the others are not. (p. 821)
• & Moreover, there are fundamental problems with their conceptualization and operationalization of status. First, their conceptualization is very loosely connected to Weber’s account, with only a minimum of direct references to it. The central figure seems to be Lauman. (p. 821)
• & Laumann’s approach has been developed further by the Cambridge school, and Chan and Goldthorpe (2010: 54) state that the CASMIS scale and their own status variable are ‘constructed in very similar ways’. Much like Laumann, the CAMSIS scale does not distinguish between class and status, and it is explicitly premised on the view that ‘economic and social relations are embedded within (p. 821)
• & each other’ (Bottero 2005: 147). What is lacking from Chan and Goldthorpe’s account is any clear argument as to why measures that are widely recognized as powerful synthetic indicators of stratification – combining class and status – without further ado can be used to refer strictly to status honour in the Weberian sense. (p. 822)
• & Second, Chan and Goldthorpe presuppose that status is tied to occupations, and that friendship patterns between occupations reflect estimations of social honour. This is not only poorly supported with argument and evidence, it is also problematic in light of Weber’s original conceptualization. Weber (2010: 148) famously simplified his argument by stating that ‘“Classes” are stratified according to the relations to production and acquisition of goods, “Stände” are stratified according to the principles of their consumption of goods as represented by specific “lifestyles”’. (p. 822)
• & Although occupation is a useful indicator of position in the relations of production, as in Goldthorpe’s own class scheme (Erikson and Goldthorpe 1993), it is less obvious that it should be the unit of choice when operationalizing status situations based on consumption and lifestyle. Chan and Goldthorpe (2004: 385) note that many sociologists ‘of otherwise divergent views’ would agree with Laumann that ‘occupation is one of the most salient characteristics to which status attaches’. It seems, however, that the meaning of ‘status’ is slipping here: what Laumann – and the other sociologists of ‘divergent views’ – mean by it is not status as social honour, but status as social position in a more general sense, as has been common in the American sociology of stratification. (p. 822)
• & A Weberian analysis of status needs to address the specific, distinct contribution of the ‘positive or negative evaluation of social honour’ to social stratification. It is clear that occupations are subject to some form of prestige ranking, notwithstanding the wide-ranging debates about the precise meaning and significance of this (see the discussion in Ollivier 2000). However, Chan and Goldthorpe do not provide good reasons for believing that they can pinpoint the specific role played by social honour by relying on occupations, as occupations obviously involve material, economic and social stratification to the degree that the occupational order is often taken to represent ‘[t]he backbone of the class system, and indeed of the entire reward system of modern Western society’ (Parkin 1971: 18). (p. 822)
• & We noted above that Weber regarded status as primarily expressed in lifestyle, and only secondarily in patterns of association. (p. 822)
• & In defending their focus on differential association, Chan and Goldthorpe (2004: 385) assume that differential association is a good indicator of the status order because ‘recurrent association is a good indicator of a state of social equality between individuals’. (p. 822)
• & It seems reasonable to suppose that people befriend or start partnerships with someone they consider to be their equals, and it seems likely that close relationships are more often maintained with their equals. But for this assumption to adequately tap into status honour as distinct from class, one needs to assume not simply that differential association is influenced by considerations of equality or inferiority, but that this is the primary and dominant cause of differential association. If not, the measure will uncontrovertibly also reflect non-status inequalities. (p. 823)
• & Indeed, as noted by Prandy and Bottero (2003: 190), measures like the CAMSIS scale are not composite measures in the same sense as prestige/status scales, since they map actual social relations of intimacy and similarity. They also tap into the social resources that underpin such interactions, and this is reflected in the fact that they are also well related to class schemes. Differential association is, of course, related to labour market advantage, and to the various economic and material resources to which individuals have access. (p. 823)
• & adherents of scales of differential association advocate them precisely because these scales reflect both status and non-status forms of inequality, as most of these scales are designed to tap into stratification in a general or synthetic sense. (p. 823)
• & it seems plausible that something similar to status honour plays a part in constituting friendships and partnerships, but it does not seem likely that these relationships strictly reflect status honour (p. 823)
• & This problem crystallizes when Chan and Goldthorpe put their status scale into action to explain lifestyles, operationalized as cultural consumption. In this attempt, they actually artificially split in two Weber’s original concept of status: they try to explain the primary expression of status (lifestyles) by its secondary aspects (restrictions on social interaction). (p. 823)
• & It is striking that while purporting to reinstate Weber’s conceptual distinction, they fail to base their measure of status on the two most distinct criteria identified by Weber: (1) That status is principally expressed in lifestyles and, accordingly, (2) that it is founded in the sphere of consumption rather than production. (p. 824)
• & The multidimensionality of social space is fairly well known in the secondary literature (see, e.g., Brubaker 1985; Wacquant 2013; Weininger 2005), but Chan and Goldthorpe mention it in only a footnote in just one of their publications (2010: 14, n.10). (p. 824)
• & First, the distinctiveness of status processes can be seen in and through the concept of embodied cultural capital. This subtype of cultural capital refers to the symbolic mastery of specific lifestyles, which are valued in specific fields, to the extent that some advantage or recognition is conferred on those who master them (Bourdieu 1986: 48–50). (p. 825)
• & Second, through his conceptualization of an analytically distinct space of lifestyles, Bourdieu advances a way of modelling divisions in Weber’s primary expression of status honour. Pioneered in Distinction, the space of lifestyles captures the main lines of difference in the distribution of tastes, preferences and practices (Bonnet, Lebaron and Le Roux 2015; Le Roux, Rouanet, Savage and Warde 2008; Rosenlund 2009). (p. 825)
• & Much like in Weber’s discussion of status, Bourdieu argues that social differences in lifestyles are subject to tacit evaluations of honour, if only through the ‘innocent language of likes and dislikes’: ‘Taste is what brings together things and people that go together’ (1984: 241). (p. 825)
• & Finally, status processes are also captured by the concept of symbolic power. This is perhaps the closest conceptual equivalent to status honour, strictly speaking: the concept refers to the social esteem associated with certain social positions, lifestyles, practices, skills, assets, and so on. (p. 825)
• & the concept of symbolic power tackles how power relations gain efficacy through a process (p. 825)
• & Bourdieu refers to as ‘misrecognition’, meaning that power relations are not seen ‘for what they objectively are’, but in a form ‘which renders them legitimate in the eyes of the beholder’ (Bourdieu and Passeron 1990: xxii). Misrecognition thus arises when capital is not recognized as the true underpinning of certain outcomes (Bourdieu 1990b). Thus, the various forms of capital can take the form of symbolic capital, which is the form assumed ‘when they are perceived and recognized as legitimate’ (1990a: 128). For instance, when excelling in the education system is seen as a reflection of ‘natural abilities’, symbolic capital is produced. In this way, Bourdieu calls attention to the specific contribution of statu (p. 826)
• & A key point in Distinction is that ‘classes increasingly take the form of status groups’ (Brubaker 1985: 756). (p. 826)
• & Everything seems to indicate that Weber opposes class and status group as two types of real unities which would come together more or less frequently according to the type of society...; [however,] to give Weberian analyses all of their force and impact, it is necessary to see them instead as nominal unities...which are always the result of a choice to accent the economic aspect or the symbolic aspect – aspects which always coexist in the same reality... (Bourdieu 1966: 212–13, cited in Weininger 2005: 84, with his additions and modified emphasis) (p. 826)
• & The homology argument thus suggests that actors in a certain region of social space have experienced similar forms of ‘conditioning’ over the life course, tending to share a certain similarity of disposition, manifesting itself in a certain similarity of lifestyle. (p. 827)
• & the extent of absolute social mobility in most capitalist societies would suggest that, in any one region of social space, there will be some significant heterogeneity of social origins which produces heterogeneity of experience, which in turn might lead us to expect that the ‘degree of fit’ between social positions, dispositions and forms of position-taking is less than perfect (p. 827)
• & This is based on our understanding of homology; we do not see lifestyles as a mere reflection or effect of class conditions but as possessing a crucial degree of autonomy. (p. 828)
• & ‘Correspondence analysis is a relational technique of data analysis whose philosophy corresponds exactly to what, in my view, the reality of the social world is’ (Bourdieu and Wacquant 1992: 96; see also the discussions in Mohr 2013; Wuggenig 2007). (p. 828)
• & Similarly, Bourdieu conceptualizes the social space as a system of ‘invisible’, objective relations between the forms of capital so that occupying a position in social space means being involved in objective relations of power, which in turn is only intelligible relationally. This way of understanding social structures has a strong affinity with the geometric properties of Multiple Correspondence Analysis (MCA), where the positions of any individual or category are also strictly relational, and there is no meaning inherent to any single position. (p. 828)
• & Making no assumptions about the distributions or properties of data, MCA aims to detect structure in the data, by extracting a few axes that represent main differences in response profiles (here, lifestyles and social positions). MCA works bottom-up, building a model from the distributions observed, as opposed to testing a predefined model against the observations (Greenacre 2007; Le Roux and Rouanet 2010). (p. 829)
• & The procedure calculates distances between all individuals and all categories: distances between individuals reflect their degree of difference in their response categories, and distances between categories reflect their degree of difference in their composition of individuals. MCA produces two clouds: a cloud of individuals representing the distances between individuals, and a cloud of categories representing the distances between categories. These clouds represent approximations of how (dis)similar individuals are in their responses, and how (dis)similar categories are in their composition of individuals. (p. 829)
• & MCA allows for the use of supplementary variables that are given no mass, meaning that they do not affect the structure of the space. This extension, called Specific MCA, also enables the use of supplementary categories. This can be useful to prevent ‘junk’ categories, like missing data, from affecting the structure of the space. Supplementary categories are positioned as a reflection of how they relate to the active categories used in the construction of the space, but the positions of the active categories remain unaffected. (p. 829)
• & Lebart, Morineau and Warwick (1984) have suggested that when using MCA for the analysis of relationships between socio-demographic and attitude variables, one should utilize a ‘reciprocal’ approach, that is, construct a space of socio-demographic variables with the attitudes as supplementary variables, and then the other way around to compare results. (p. 829)
• & Drawing on Lebart et al.’s work, Rosenlund (2009, 2014) has developed a way of approaching the homology thesis in more detail. He has operationalized it by constructing each space – one social space and one space of lifestyles – with separate correspondence analysis procedures for each, and then using coordinates from the one as supplementary categories in the other to compare their structures. (p. 829)
• & We expand on this by drawing on Le Roux and Rouanet’s (2004) extension of MCA methodology to include both analysis of variance and Euclidean classification, that is, Agglomerative Hierarchical Clustering with Ward’s criterion (p. 829)
• & applied to the axes produced by MCA. This allows us to provide a fuller account of the relationship between the two structures, and a more precise estimation of their interrelationships. (p. 830)
• & Our analytical strategy thus involves four steps. We begin by constructing and interpreting a space of lifestyles. Following Weber and Bourdieu, we see lifestyle as the primary expression of status. Second, we perform a Euclidean classification of this space, producing optimally homogenous lifestyle clusters. Third, we construct a model of the social space based on an independent MCA of indicators of economic and cultural capital. We follow Bourdieu in seeing the social space as society’s multidimensional class structure. Fourth, we use the lifestyle clusters as supplementary categories in the social space. This can be thought of as a form of ‘visual regression analysis’ (Lebart et al. 1984), meaning that we use the social space as the ‘independent variable’ and the lifestyle clusters as the ‘dependent’ one. This allows us to demonstrate the homology between the two spaces and we use ANOVA to gauge the degree to which they correspond. (p. 830)
• & We know of no conventional social science survey capable of offering us this opportunity, so we have opted for the 2011 round of Norsk Monitor, carried out by the Ipsos MMI.1 (p. 830)

⛔ Weaknesses and caveats

• ! The social segregation of neighbourhoods, schools and universities means that the potential pool from which we ‘choose’ friends and partners is highly selective before we ‘choose’ them, that is, before our own considerations of status come into play, so that socioeconomic factors and esteem are intertwined in producing patterns of friendship. (p. 823)