Transcript Cultural Capital and Social Exclusion: some preliminary

Cultural Capital and Social
Stratification in Britain
Mike Savage, (Sociology & CRESC),
University of Manchester
The fall and rise of class in Britain
•
During the 1980s and early 1990s, British class analysis becomes increasingly
marginalised and loses its way….
•
•
•
•
Since the mid 1990s, there has been a striking revival of class analysis, much
of it with a strongly ‘cultural’ inflection. Key writers include Stephen Ball,
Simon Charlesworth, Tim Butler, Fiona Devine, Diane Reay, Andrew Sayer,
Bev Skeggs, Valerie Walkerdine, Alan Warde. Indicative work includes
•
•
•
Loss of theoretical engagement with issues of social change
Exclusive focus on structural issues, e.g. social mobility, rather than culture and identity
Loss of engagement with American traditions
Rethinking Class, culture, identities and lifestyle (ed Devine, Savage, Scott and
Crompton)
Sociology, special issue on ‘class and identities’, December 2005
Bourdieu is the central theoretical reference point of this new work, but his
claims about cultural capital have not been explicitly tested in the UK using
both quantitative and qualitative research. This is also necessary to allow an
engagement with older ‘class structural’ traditions of research
• This is the remit of the major ESRC funded project, Cultural Capital and
Social Exclusion, directed by Tony Bennett and Elizabeth Silva (Open
University), myself and Alan Warde. Our three aims
1. What is the nature of cultural capital in Britain?
2. How important is cultural capital, vis-à-vis economic and social capital
3. How does the concept of cultural capital address issues of policy concern
around social exclusion
Researching class & cultural capital
Bourdieu’s concept of cultural capital raises a host of difficult issues….
- Is the concept of habitus over determinist?
- Is Bourdieu’s Distinction too rooted in place (France) and time (1960s) to have
contemporary relevance?
- How does the approach relate to the ‘class analysis’ tradition?
.
Most British sociologists have two very different ‘takes’ on Bourdieu:
1. ‘Old-school’ stratification sociologists have generally been critical of Bourdieu
but have never systematically sought to measure cultural capital directly
• Halsey and educational inequality (parental education taken as surrogate for cultural capital)
• Goldthorpe and class inequality (premised on RAT approaches)
2. British cultural sociologists have emphasised (some) theoretical aspects and not
his methodological and empirical endeavours
• Featherstone, Lash & Urry on post-modernism and the new petty bourgoisie
• the almost exclusive use of qualitative case studies (Charlesworth, Skeggs etc)
Tony Bennett, Elizabeth Silva, Alan Warde & I set out to measure cultural capital
through emphasising the need to place habitus and cultural capital within the
contexts of cultural fields, namely music, reading, film, TV, sports, art, leisure,
eating out.
How do we define cultural capital?
Bourdieu sees cultural capital as implicated in the
reproduction of class privilege. It involves a distance from
‘everyday life’, and some kind of tension between high and
low culture, but there are different forms this might take:..
• The ‘Kantian aesthetic’, distance from the necessary
• ‘Snob’ culture or ‘the leisured aesthetic’
• The institutionalised ‘canon’ (through familiarity with cultural artefacts
legitimised in education curriculum)
• The ‘cultural omnivore’ (Petersen, Bryson, Erickson)
• ‘national cultural capital’ (Hage)
These definitions relates to different understandings of the
middle class. Is there a divide between ‘industrialists’ and
‘intellectuals’, or can we see them united in a broader
‘service class’.
CCSE research
• Project includes 25 focus groups, 60 in depth interviews,
and national survey of 1564 respondents, plus ethnic boost
of 200.
• Survey questions on 8 cultural subfields, tapping taste,
participation and knowledge
TV – stations watched, programmes (dis)liked, frequency of
viewing
Film - genres & 7irectors (dis)liked, frequency of attendance
Music - genres & artists (dis)liked, attendance at musical event
Reading – genres and writers (dis)liked, books read,
Visual arts – genres & artists (dis)liked, works possessed,
Eating out – kind of venues (dis)liked
Embodiment – sport, body modification, clothes, household style
• Also questions on economic and social ‘capital’; domestic
division of labour; parents’ cultural interests and
background; respondents’ social, cultural and political
attitudes
How do we analyse this data?
• Recent cultural sociology increasingly applies orthodox
modelling procedures in recent work, seeking to predict
specific kinds of cultural outcomes (see examples in
Poetics)
• The danger is that this focuses on crude indicators, and the
complexity of cultural taste, participation and knowledge is
not explored.
• We were therefore attracted to Abbott’s (2001) call for the
revival of descriptive quantitative approaches (e.g. factor
and cluster analysis, social network analysis) over
explanatory methods (e.g regression) which rely on
problematic assumptions of ‘general linear reality’.
• This tallies with Bourdieu (1984) and Peterson’s (1983)
call for descriptive quantitative methods as a means of
understanding the ‘patterning of culture’
Using orthodox modelling approaches….
1. (Pre?) define variables which indicate ‘high culture’ and see which
groups are predisposed to them? Table 1 indicates some of the
predictable relationships that can be delineated in this way…..
Table 1: Logistic Regression on liking of musical genres
Rock
Jazz
World
Classic
Country
Electronic
Heavy M
Urban
Female
.535***
.822
.787
.853
.949
.785
.417***
1.302
Age
.952***
.996
.969***
1.063***
1.051***
.928***
.958***
.921***
Ethnic minority
.226***
.685
5.208***
1.867*
1.935*
.606
.285**
1.672
O level
.959
.978
1.187
1.979***
.563**
.840
1.237
1.340
City & Guilds
1.168
1.673*
1.223
1.821**
.734
.783
2.103*
1.433
A level
1.160
1.231
1.243
2.431***
.575*
.466**
1.493
.809
University
1.740**
1.834*
1.630
6.040***
.498**
.806
1.724
.924
Employers
3.404**
2.775*
.468
1.147
.673
1.290
1.493
1.095
Self employed
1.583
1.866*
.605
.745
1.003
.330
2.497**
.856
Hi profs
1.308
1.785
.698
1.708
.556
1.338
1.123
664
Lo profs
1.012
1.603
.839
1.175
.605*
.636
1.386
1.163
Lo mgrs
1.690
1.285
.649
1.050
.675
.665
.883
1.202
Hi super’
.705
1.421
1.164
1.212
1.452
.692
1.615
1.475
Intermed
1.323
1.116
.729
1.211
.938
1.488
1.675
.717
Lo super’
1.461
1.126
1.103
.987
.717
.973
1.910
.766
Lo technic
1.881*
1.506
1.446
.502
1.125
1.443
1.704
.925
Routine
.884
.930
.914
.798
.709
1.819
.873
.806
Using orthodox modelling approaches….
1. (Pre?) define variables which indicate ‘high culture’ and see which
groups are predisposed to them? Table 1 indicates some of the
predictable relationships that can be delineated in this way
age effects are the most powerful
education effects are more powerful than class effects
the middle classes like most genres,
ethnic effects are powerful in several cases
But, lots of odd results, consider Heavy Metal & self employed, Rock
music and employers, etc:
2. We don’t get an easy sense of the overall ‘patterning of culture’ through
such methods. Therefore we are keen to describe the nature of cultural
taste more fully so that we can interpret inductively which cultural
practices, if any, are relationally demarcated from others…. Which
explains why we turned to ‘Geometric Data Analysis’
Geometric Data Analysis (GDA)
• Aided by Johs Hjellbrekke (Bergen), Brigitte LeRoux, and
Henry Rouanet (Paris V).
• Adapts a form of principal components analysis to locate
responses as coordinates in geometric space according to
categorized responses to 168 modalities, (derived from 41
questions) across a range of cultural practices
• We can then inductively interpret the axes to empirically
assess which cultural practices are related and to consider
what forces separate practices out.
• By superimposing socio-demographic variables on this
‘space of lifestyles’ we can assess how far the cultural
patterns revealed appear to socially structured
….We can read the figures as cultural maps
Table . MCA cloud of contributing modalities, axis 1 and 2.
The cultural field
Mostly likes and engagement
Overwhelmingly dislikes and avoidances
Electronic, urban & rock music, modern art,
science fiction, TV comedy, playing football,
eating Indian and Chinese food
Especially ‘established’ cultural forms
Watch 5+ hours TV each weekday
Mostly likes and engagement
Mostly dislikes and avoidances
Impressionist painting, modern literature,
French restaurants, Opera, art galleries,
museums, orchestral concerts, stately homes
Especially modern music and art. Also sport
and going out. Likes country and western
music, musicals, fish & chips
Table . MCA cloud of individuals: preferences for classical music lit up, axis 1 and 2.
Table . MCA cloud of individuals: frequency of opera attendance lit up, axis 1 and 2.
Cultural capital/education/ class
The cultural field
Mostly likes and engagement
Electronic, urban & rock music, modern art,
science fiction, TV comedy, playing football, eating
indian and chinese food
Professional/managerial, well-educated, young
Mostly likes and engagement
Impressionist painting, modern literature, French
restaurants, Opera, art galleries, museums,
orchestral concerts, stately homes
Professional/managerial, well-educated, older
Overwhelmingly dislikes and avoidances
Especially ‘high’ cultural forms
Watch 5+ hours TV each weekday
Routine jobs, poorly educated, young
Mostly dislikes and avoidance
Especially modern music and art. Also sport and
going out. Likes country and western music,
musicals, fish & chips
Routine jobs, poorly educated, older
So, is there cultural capital in the UK?
• No obvious divide between the Kantian and the ‘leisured’
aesthetic:
• No evidence for distinction between ‘intellectuals’ and ‘industrialists’
• The wealthy are attracted to established high culture
• We do see the power of the cultural ‘canon’ and powerful
circuits linking education and class position
• The middle classes are increasingly attracted to omnivore
taste, meaning that the divide between ‘high’ and ‘popular’
culture is replaced by that between ‘multiply engaged’ and
‘disengaged
• Multiple frames of ‘national’ cultural capital operate, with
strong emphasis on British culture, and lesser appeal of
American cultural forms.
How does this relate to class stratification?
• It might appear that our findings endorse the view that the
divide between ‘service class’ and other classes are
primary, but it is more complex than this.
–
GDA shows that the most efficient way of placing people into classes means that the most
advantaged class includes higher employers and managers and professionals, but not lower
managers.
We can distinguish three main classes: a professional class
of 24% of population, an intermediate class of 30% and a
large working class, including supervisors and technicians,
of 46%
• Class effects are closely related to education effects, and if
forced to chose, education appears the main force.
-1
P lane 1-2: 2 Higher profes s ional oc c .
1
P lane 1-2: 10 S emi-routine
0.5
0
-0.5
-1
-1
-0.5
0
0.5
plane 1-2:1-E mployers large org/Higher managerial
1
-0.5
0
0.5
1
P l ane 1-2: 3Lower prof/hi gh tec hn
0.5
0
-0.5
-1
1
1
Plan des axes 1-2 (3ème classification)
0.5
0
-0.5
-1
-1
-0.5
0
0.5
1
Plane 1-2 : 4- University
palen 1-2 1 :no education
1
1
0.5
0.5
0
0
-0.5
-0.5
-1
-1
-1
-0.5
0
0.5
1
-1
-0.5
0
0.5
1
Some general conclusions
• Looking at issues of consumption and lifestyle, class divisions are
profound. It still makes sense to see Britain as having three fairly
strongly culturally bounded classes, with a large working class.
• Class divisions need to be seen as articulating educational and
employment, and as demonstrating the power of institutionalised
cultural capital.
• Little evidence for the central importance of the divide between ‘high’
and ‘low/ popular’ culture: things have moved on from Distinction.
Should we distinguish ‘primitive’ from ‘mature’ cultural capital?
• The social capital debate rightly alerts us to the dynamics of
engagement and disengagement as now fundamental to stratification
and inequality.
• Age (and to a lesser extent, gender and ethnicity) prove to be of
fundamental importance in shaping cultural practices
• Methodologically, we argue for the importance of ‘descriptive’ over
‘variable centred’ methods. The visualisation possibilities opened up
by new computing increases the scope and potential for such methods