Crossley 2005
Crossley 2005
Crossley 2005
http://bod.sagepub.com/
Published by:
http://www.sagepublications.com
On behalf of:
The TCS Centre, Nottingham Trent University
Additional services and information for Body & Society can be found at:
Subscriptions: http://bod.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
Citations: http://bod.sagepub.com/content/11/1/1.refs.html
What is This?
NICK CROSSLEY
Much work in the sociology of the body has been devoted to an analysis of body
modification and maintenance; that is, to practices such as diet, exercise, body-
building, tattooing, piercing, dress and cosmetic surgery (e.g. Crossley, 2004a,
2004b; DeMello, 2000; Entwistle, 2000; Featherstone, 1982, 2000; Gurney, 2000;
Monaghan, 2001; Pitts, 1998, 2003; Rosenblatt, 1997; Sanders, 1988; Sassatelli,
1999a, 1999b; Smith, 2001; Sweetman, 1999; Turner, 1999). In this paper I seek to
contribute to this work on two fronts. First, developing a theme already present
in the literature, I explore the reflexive and embodied nature of practices of
modification/maintenance. It is very easy when discussing this topic to slip into
a dualistic framework, opposing the body to either self or society and seeming
to suggest that the former is transformed by the latter. We talk, for example,
about ‘my body’, what ‘I’ think of ‘it’, what ‘I’ put ‘it’ through and what ‘I’ want
‘it’ to look like. On one level this linguistic habit resonates with our experience.
Our socially instituted capacity for reflexivity allows us to turn back upon and
objectify ourselves, effecting a distinction between what Mead (1967) referred to
as the ‘I’ and the ‘me’; the body as subject and the body as object. On this level
Body & Society © 2005 SAGE Publications (London, Thousand Oaks and New Delhi),
Vol. 11(1): 1–35
DOI: 10.1177/1357034X05049848
www.sagepublications.com
Downloaded from bod.sagepub.com at Ondokuz Mayis Universitesi on April 23, 2014
01 crossley (ds) 24/1/05 3:15 pm Page 2
we both are our bodies and we have a body (Crossley, 2001). However, it is
necessary to recognize this split as reflexive rather than substantial in nature. It
derives from our acquired capacity to assume the role of another and thereby to
achieve an outside perspective on ourselves, a process which generates a sense of
our being distinct from the qualities we identify with our self when assuming this
‘other’ role. It does not indicate a substantial distinction between mind/self and
body. It does not often even reflect the emergent stratification between the body
as a biochemical structure and the body as a sensuous, active agent. ‘My body’,
the body I ‘have’, is a moral, aesthetic, acting and sensuous being. I worry as
much about its appearance, performances and transgressions as I do about its
biological structure. The best work in the sociology of the body recognizes this.
It challenges dualism, insisting that ‘I’ am ‘my body’ and that body projects are
therefore reflexive projects (see esp. Crossley, 2001, 2004a; Entwistle, 2000;
Monaghan, 1999; Smith, 2001; Sweetman, 1999; Wacquant, 1995, 2004).
In this article I advance this idea through an exploration of what I call reflex-
ive body techniques (RBTs), a concept which builds upon Marcel Mauss’s (1979)
concept of body techniques (see also Crossley, 1995, 2004a, 2004c) and upon my
own earlier work on reflexive embodiment (Crossley, 2001). The concept of
RBTs, I will show, affords a powerful analytic purchase upon the embodied and
reflexive processes and practices involved in projects of body modification/main-
tenance and, indeed, upon the reflexive separation of the embodied I and me.
The concept of RBTs also frames my second theme: the social distribution and
diffusion of practices of modification. Specifically, I will demonstrate and seek to
explain the fact that the overall repertoire of RBTs belonging to any group can
always be differentiated into: (i) clusters which all members practise, (ii) clusters
which the majority or a large minority practise and (iii) clusters which only a
small minority practise. Furthermore, I will demonstrate and seek to explain the
fact that within the ‘zone’ of less widely practised RBTs we find clusters which
‘go together’ thematically and/or in the sense of being statistically associated.
Recognizing this pattern of distribution is important because it alerts us both to
the different meanings attaching to specific clusters and to their variable
conditions of diffusion and appropriation; their levels of accessibility and the
different balance of costs and rewards that attach to them.
This is an important observation in relationship to our broader focus. General
descriptions such as ‘body modification’ and ‘body maintenance’ can be mislead-
ing because they imply that we are dealing with a set of practices with a common
identity, purpose, accessibility, etc. They fail to distinguish between the social
logic of distinct sets of practices. This can lead to theoretical accounts which do
likewise. Giddens’ (1991) theory of ‘the body in late modernity’, to take one
F I
T A J
D G N
P Q
Figure 1 A Radex
more females3 – and involved representation from a variety of age,4 social class,5
ethnic6 and religious7 groups. I make no claim with respect to representativeness,
however. The sample was convenient and sufficient for my present, preliminary
investigations, but it is far from perfect. Nor do I suggest that the survey results
are particularly startling. I found what I expected to find. The point of the survey,
however, was to allow me to check that my assumptions were borne out amongst
a population whom I have no reason to believe are unusual,8 to consider how we
might move from assumptions to empirically verifiable models, and also to
explore the above-mentioned methodological innovations.
With this said, we can turn to the main arguments of the article. I begin with
a discussion of Mauss’s formulation of ‘body techniques’. My claim, to reiterate,
is that we can conceptualize and analyse practices of body modification/main-
tenance as particular types of body technique, namely reflexive body techniques,
and that there are advantages to doing so.
Body Techniques
Mauss (1979) arrived at the concept of body techniques after having observed
both that certain embodied practices (e.g. spitting, hunting techniques and eating
with a knife and fork) are specific to particular societies, and that others vary
considerably in style across societies and social groups. Women walk differently
to men, the bourgeoisie talk differently to the proletariat, the French military
march and dig differently to British troops and so on. Building on these obser-
vations, he defines body techniques as ‘ways in which from society to society
men [sic] know how to use their bodies’ (1979: 97). This definition is potentially
problematic as it can seem to suggest, in the fashion warned against above, that
‘men’ and ‘their bodies’ are different things. Given the way in which Mauss
pursues his point, however, it is reasonable to assume that this is not his inten-
tion. Indeed, the concept is used to effect a sophisticated, albeit often tacit inno-
vation in non-dualistic sociology. Mauss’s description of body techniques as
‘habitus’ is an important point of entry for grasping this innovation. ‘Habitus’,
he explains, is a Latin rendering of the Greek ‘hexis’ (or ‘exis’), a concept which
is central to Aristotle’s philosophy, wherein it denotes acquired and embodied
dispositions9 which constitute forms of practical reason or wisdom (see Aristotle,
1955). Habitus and thus body techniques have a double edge in this definition.
They are forms of embodied, pre-reflective understanding, knowledge or reason.
And they are social. They emerge and spread within a collective context, as the
result of interaction, such that they can be identified with specific social groups
or networks:
. . . I have had this notion of the social nature of ‘the habitus’ for many years. Please note that
I use the Latin word – it should be understood in France – habitus. The word translates infi-
nitely better than ‘habitude’ (habit or custom), the ‘exis’, the ‘acquired ability’ and ‘faculty’ of
Aristotle (who was a psychologist). It does not designate those metaphysical habitudes, that
mysterious ‘memory’, the subject of volumes or short and famous theses. These ‘habits’ do not
vary just with individuals and their imitations; they vary between societies, educations, propri-
eties and fashions, prestiges. In them we should see the techniques and work of collective and
individual practical reason rather than, in the ordinary way, merely the soul and its repetitive
faculties. (Mauss, 1979: 101)
I will return to the question of how body techniques distinguish and differenti-
ate social groups later. Here I am interested in the manner in which the concept
simultaneously holds together social, corporeal and cognitive elements. In doing
this ‘body techniques’ rejoins the Durkheimian tradition it derives from in two
respects. First, it rejoins the concept of ‘social facts’, integrating it with a
consideration of biological and psychological facts (see also Lévi-Strauss, 1987).
Body techniques are social facts. They vary across societies and social groups.
They pre-exist and will outlive the specific individuals who practise them at any
point in time. Mauss even seeks to show – albeit somewhat problematically
(Crossley, 1995, 2004a) – how they ‘constrain’ agents. At the same time, however,
they presuppose biological structures and embody knowledge, reason and
psychological properties. Styles of walking vary across social groups, for
example, indicating a social basis, but all of these different styles presuppose
bipedalism, not to mention the plasticity and intelligence that allow the organism
to develop and learn different ways of walking. They thus have biological
preconditions. Furthermore, styles of walking embody understanding and
knowledge. Switching to ‘tiptoes’ when silence is required, for example, indicates
a grasp of the conditions most conducive to minimizing noise, while walking a
tightrope, and indeed walking per se, requires a practical grasp of principles of
balance, force, etc. When we adjust our posture to steady ourselves we engage in
practical physics. Finally, certain styles of walking, such as a proud march or
arrogant strut, embody an emotional intention (in the phenomenological sense
of ‘intention’) and may even be employed as a means of generating such an inten-
tion (see Crossley, 2004c). As both Sartre (1993) and Merleau-Ponty (1962) note,
we can generate emotional intentions, putting ourselves into particular moods,
by acting out the mood; that is, by performing the body techniques (partly)
constitutive of it. This is a key function of body techniques within certain rituals
(Crossley, 2004c). Thus, body techniques have a psychological dimension too.
Second, ‘body techniques’ extends the Durkheimian notion of collective
representations, a notion that Mauss himself developed with Durkheim
(Durkheim and Mauss, 1963). It identifies collective forms of wisdom and
Hairdressing, massage, dental work and cosmetic surgery, for example, usually
entail that one ‘body’ is worked upon, physically, by another or by a team of
embodied agents. It might even be extended to include such distanciated and
mediated interactions as those that connect the manufacturers of pharmaceuticals
to those who distribute and use them. Pill-popping is a socially complex, dis-
tanciated and mediated RBT. Equally, it can entail a single ‘body’ acting upon
itself. This might involve one part of the body being used to modify/maintain
another part; for example, when I use my hand to brush my hair or clean my
teeth. It might, however, entail a total immersion of the body into a stream of
activity whose purpose is to modify or maintain that body as a whole. When I
jog, for example, I launch my whole body into action, in an effort to increase my
fitness, burn off fat/calories, tone up my lower body, etc.
Each society or social group has a repertoire of RBTs. This repertoire is one
element in the broader set of collective pre-representations of that society. And
a portion of our daily routine is taken up performing techniques from this reper-
toire. We wash, clean our teeth, brush our hair, dress ourselves, perhaps shave
and/or apply cosmetics. Other techniques from the repertoire are built into
weekly or monthly routines. We exercise, shave parts of our body, have our hair
cut, cut our finger nails, dress up for a night out, etc. And beyond our routines,
we periodically venture ‘one-off’ modifications, such as a piercing, tattoos or
cosmetic surgery, all of which involve bodily manipulation of the body; that is,
RBTs. RBTs are techniques of the body, performed by the body and involving a
form of knowledge and understanding that consists entirely in embodied
competence, below the threshold of language and consciousness; but they are
equally techniques for the body, techniques that modify and maintain the body
in particular ways.
It might seem peculiar to regard the more mundane of these techniques as
acquired aspects of a culture. As Mauss’s work shows, however, they do vary
across societies. And, as Goffman has argued, they only seem mundane to those
of us who have achieved sufficient temporal distance from the process of learning
them to have forgotten that we did have to learn them, sometimes with difficulty:
To walk, to cross a road, to utter a complete sentence, to wear long pants, to tie one’s shoes,
to add a column of figures – all these routines that allow the individual unthinking, competent
performance were attained through an acquisition process whose early stages were negotiated
in cold sweat. (Goffman, 1972: 293)
For practical purposes it also makes sense to refer to ‘ensembles’ of RBTs; that
is, sets of techniques which are practised together for a common purpose. ‘Exer-
cising’, ‘getting dressed’ and ‘putting on make-up’ are each examples of this. Each
refers not to a single body technique but to a set (an ensemble) of techniques. It
may be necessary, for purposes of certain types of research, to break these en-
sembles down into their constituent techniques. If we were studying health and
fitness, for example, we might want to break ‘exercising’ down into finer
categories such as gym work and swimming, or even ‘2 12 reps bench press’
and ‘20 lengths front crawl’. We may even need to tease out the peculiarities of
the way in which the technique is performed: e.g. bench press with arms parallel
and feet touching the ground. It is not always necessary to chase the description
down to this level, however. Identification of the basic ensemble may be
sufficient. My above-mentioned survey includes a mix of specific body tech-
niques and ensembles but for economy of expression I will use ‘RBT’ to refer to
both.
It is my contention that practices of body modification/maintenance are best
understood in terms of RBTs. I have three reasons for this. First, the concept of
RBTs entails that ‘bodies’ are maintained and modified by way of bodily effort
and embodied competence. We thus both avoid dualism and thematize reflex-
ivity. Second, the concept encourages us to identify the ‘mindful’ and social
aspects of embodied activity (e.g. know-how and understanding), not subordi-
nating those aspects to the symbolic meaning bestowed by representations,
discourse, consciousness, etc., and not reducing embodied activity to mere
mechanical behaviour. Third, the concept is sufficiently concrete to facilitate
empirical analysis and sufficiently rich for that to include both ethnographi-
cal/phenomenological investigation of the doing of RBTs, their lived dimension
(see Crossley, 2004a; Wacquant, 2004), and also more quantitative explorations
of them. As forms of practical understanding, RBTs need to be understood quali-
tatively, in a phenomenological manner. And they are also thematized within
projects and narratives which call for qualitative investigation. But qua social
techniques they are diffused and distributed through society. They can be
observed, categorized, enumerated and tested for statistical association both with
one another and with other social phenomena. They thus call for and admit
quantitative analysis also.
Before we can push ahead with this concept, however, we need to reflect upon
the role of ‘purpose’ in relation to RBTs. I have said that RBTs are body tech-
niques whose primary purpose is to act back upon the body so as to modify or
maintain it. At its most basic this entails that RBTs are generic body techniques
which an agent annexes, in a specific context, for the explicit purpose of (perhaps
amongst other things) modifying their body in a particular way: for example, in
an effort to lose weight they elect to take a walk once a week. In many cases,
however, reflexive purposes have generated either dedicated techniques or dedi-
cated variations upon generic techniques. ‘Jogging’, for example, is a form of
running adapted to serve the purpose of exercise. In contrast to a mad dash for
the bus, a jog entails that I pace myself (a temporal modification), adjust my
breathing and ‘settle into’ a comfortable and efficient posture and stride. Jogging
embodies a particular temporal projection (I will run for this long). It is oblivi-
ous to the urgency that animates the person running after the bus. And in these
respects it embodies the purpose of running for the benefit of running. When I
jog I relate to my environment in a different way. Perhaps I lengthen my step,
utilizing my body differently in order to utilize the ground differently; simul-
taneously utilizing the ground differently in order to utilize my body differently,
pushing it towards the ‘burn’ that I am seeking. Street-lights and other random
objects become stage markers, triggering alterations in pace and direction, for a
journey whose only goal is transformation of the ‘vehicle’ itself and whose
success is measured against this goal. In this way, both agent and environment
are instituted10 in a specific ‘jogging-like’ manner by way of a dedicated RBT.
Moreover, in contrast to the person running after the bus, whose fear of missing
the bus and looking silly often leads them to ‘disguise’ their sprint as much as
possible, I commit myself publicly to my jog. My action is accountable. I
embody a social type or role, ‘the jogger’. And I am able to do this because
jogging is not my private invention but rather a socially emergent and publicly
known technique that belongs to my society’s repertoire. I have selected and
learned from this repertoire and for this reason everybody knows what I am
doing when I stagger past them.
Purpose also enters into the analysis of reflexive body techniques in the
respect that the body can be modified for different reasons. One might modify
one’s body for reasons of health, beauty, sporting success, etc. Again, this might
involve significant permutations of an apparently singular technique. Body-
builders, powerlifters and individuals who want to ‘tone up’ and ‘trim down’
might each use dumbbells and barbells, for example, and might even do the same
exercises (bench press, squat, etc.). However, the way in which they do those
exercises will vary. The ‘toner’ will tend to do a high number of repetitions with
weights they can quite easily lift because this is good for toning; the powerlifter
will do relatively few reps with a weight that is very heavy for them because this
increases strength; the bodybuilder, who is concerned to increase muscle bulk but
also muscular definition and ‘rips’, will use a combination of the two. Further-
more, both of the latter two will tend to work out for much longer in any weekly
cycle. We need to be mindful of these differences when studying RBTs.
safe-because-same manner. They invest the flow of lived time with meaning by
punctuating it, but this meaning centres upon continuity and sameness rather
than transition. It is integral to grasp this balance of reproduction and trans-
formation in our understanding and analysis of RBTs, and indeed also the
different temporal configuration that specific techniques can assume. RBTs have
a spatio-temporality that is central to their meaning. This is reflected in the
linguistic duality of ‘body maintenance’ and ‘body modification’, which I have
employed hitherto in this paper. The former denotes techniques used repetitively,
for reproductive purposes, the latter denotes techniques used to effect a specific
transformation.
Why agents engage in this body work is a key question in sociology, about
which most of the major perspectives in the area have something to say (see
Bartky, 1993; Baudrillard, 1999; Bordo, 1993; Bourdieu, 1977, 1984; Foucault,
1980; Giddens, 1991; Shilling, 1993). I do not have the space to address this
question fully here. However, it is my contention that we must approach it in a
way that recognizes the great diversity of RBTs in the societal repertoire and the
very different social logics that can attach to their appropriation. ‘One size fits all’
explanations, such as we get from most of the above-mentioned theorists, are
deeply problematic because they fail to recognize this diversity. In what follows
I will explore this issue of diversity and social logics in more detail in an effort to
lay the basis for a more sophisticated approach to body work, centred upon RBTs.
Likewise, Bourdieu’s (1977, 1984) focus upon distinction and Elias’s (1984) upon
the civilizing process both draw out forms of bodily practices, specific to social
groups, which alter bodily appearance and thereby distinguish and mark out
those groups visually. Furthermore, as Durkheim also emphasized, techniques of
body modification are sometimes employed to mark out categorical distinctions
within a group, for example between males and females or adults and children.
Contemporary western societies differ considerably from the totemic clans
studied by Durkheim. Interestingly, however, my above-mentioned survey
revealed gender to be a key factor affecting appropriation of reflexive body tech-
niques. The survey found statistically significant and often very large differences
for gender in relation to 21 (out of 40) RBTs and 6 (out of 19) forms of
consultation (see Table 1). In particular, practices such as the shaving of armpits
and legs, the painting of toenails and fingernails, manicure and the use of cosmet-
ics (other than soap and shampoo) were sharply gender differentiated.
This finding bears out the claim of those who argue (i) that gender remains an
extremely strong locus of social division in contemporary societies, (ii) that the
body is a key site where this division is constructed and played out and (iii) that
gender identity is something which is ‘done’ or ‘practised’ (e.g. Bartky, 1993;
Connell, 1987). In addition, the arguments of these writers provide a strong and
important lead for the analysis of RBTs, which I will return to later on. Import-
antly, however, the concept of RBTs, as both a theoretical and an empirical
notion, allows us to further elaborate upon and explore these ideas about gender;
to go beyond general theoretical claims about transformations of the body
associated with, for example, ‘emphasised femininity’ (Connell, 1987) by speci-
fying just what transformations are effected, by what means and by what types
and proportions of women (relative to men). Furthermore, as I discuss in more
detail below, we can begin to explore the patterns of clustering of these various
techniques and, by means of this, differentiate varieties within emphasized femi-
ninity.
The low numbers of men practising certain techniques is as interesting as the
high number of women in this respect. Shaved armpits are as much an affront to
‘hegemonic masculinity’ (Connell, 1987), punishable amongst men, at least in the
absence of extenuating circumstances, as they are an expected standard of
‘emphasised femininity’. And a monitoring of the appropriation of these tech-
niques among men allows us to gauge changes in masculinity. We can speculate
upon all of this without recourse to the concept of RBTs, of course, but the latter
provides an important and workable means for empirically operationalizing such
speculation and recording change. Shifting rates of uptake for particular RBTs
among men and women respectively, allow us to track shifts in dominant models
of masculinity/femininity – although, of course, the meaning of changes is never
self-evident and must be deduced.
Consultation Practices
Read magazine/newspaper article on beauty tips 57.1 3.4 .000
Read magazine/newspaper article on health tips 50 10.8 .000
Read magazine/newspaper article on skin care tips 48.4 2.6 .000
Read magazine/newspaper article on exercise tips 36.1 16 .000
Read any of the above 69.6 24.2 .000
Read a health-dedicated magazine 21.7 5.8 .034
Consulted a beauty-dedicated website 3.8 0 .032
One might expect to find similar sharp distinctions in relation to the practices
of certain religious groups and perhaps subcultures of various kinds. I found no
statistically significant differences pertaining to class, however. And although
contemporary ethnographic studies of specific working-class communities have
Mapping Techniques
Consider, first, the frequency distribution for the techniques surveyed by my
questionnaire, as given in Figures 2 and 3. The range of this distribution stretches
from 100 percent, for having washed one’s hands at least once in the last seven
days, through to 0.3 percent, for having had a septum or tongue piercing, or
having ever used anabolic steroids for purposes of building one’s muscles. This
is a continuum and any attempt to demarcate definite lines of division along it
would inevitably be arbitrary. Moreover, we already know that certain of the
practices are heavily gendered, such that some scores represent a mean of high
No. Technique %
100
90
80
70
60
50
40
30
20
10
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 34 35 36 37 38 39 40 41 42
female and low male scores. I return to this latter point shortly. First, however,
I want to suggest that, although any exact cut-off points would be arbitrary, we
can divide this continuum into three overlapping zones. At one end of the
continuum we have a core zone consisting of RBTs that are statistically normal;
that is to say, which most people (e.g. 90 percent+) practise within a specifiable
time-frame. At the other end of the continuum we have a marginal zone consist-
ing of RBTs that are statistically deviant; that is, which most people (e.g. 95
percent+) do not and never have practised. Between these two zones we have
what we might think of as an intermediate zone, a broad continuum of RBTs
with rates of uptake that vary in the general population but which are neither so
high as to be normal and thus ‘core’, nor so low as to be statistically deviant and
thus ‘marginal’.
In what follows I will be using this concept of ‘zones’ to develop a differen-
tiated account of body work in late modern societies. The appropriation of RBTs
in these different zones requires different explanations, I will suggest. First,
however, I want to push the idea one step further by unpacking the idea of a
continuum. While useful as a point of departure, the image of the continuum is
problematic on account of its linearity. We know, for example, that some body
techniques are strongly gendered, such that they might qualify as ‘core’ for
women, while they are ‘marginal’ for men. Likewise, we can at least speculate
that other techniques in the intermediate and marginal zones will be further
differentiated from one another. Marginal techniques might belong to different
and competing subcultures, for example, such that placing them side by side in a
common ‘marginal’ category paints a misleadingly homogeneous picture. To do
justice to this we can attempt a two-, rather than a one-dimensional mapping of
RBTs, using multi-dimensional scaling (see Figure 4).
As with all forms of multivariate statistics, multi-dimensional scaling offers
contrasting methods for mapping data. As such we are forced to select the most
2 diet
cossurg
manicure
careful eat quick tan
dyehair
floss
bellybutton
paint f nails sunbed
1 paint tcosmetics
nails vitamin supp
leghair
pithair
braceletnecklace food supplement
steroid
bodybuilding supplem
0 breathfreshner 5-9hrs exercise
earring comb
wash sunbathe
after/perfume
wash face
bath/shower
brush
wash
anti/deo
hair
hands
hair
teeth
1-4 hrs exercise ring
–1
1-3 tattoos
eyebrow
tongue
>3 tattoos
–2 other pierce
–3
–2 –1 0 1 2 3
the ‘either/and’ techniques falling between, such that we can regard health and
appearance as distinct thematic segments. I do not have the space to explore this
any further here, however.
As we move further round we come to another more distinctive cluster,
comprising 5–9 hours exercise a week, use of bodybuilding supplements and use
of anabolic steroids for bodybuilding purposes. We might be tempted to deem
this a bodybuilding cluster. As with any cluster, however, we can push our
analysis further and differentiate within it. In this case, for purposes of illustrat-
ing MDS technique as much anything else, I have done this. In the first instance
I selected from my sample only those respondents who had done more than five
hours’ exercise in the previous seven days. I then reran the MDS analysis,13
selecting all variables with a direct relationship to bodybuilding: namely, working
out with weights in a gym, reading bodybuilding magazines, visiting bodybuild-
ing web-sites, taking bodybuilding food supplements and taking steroids (for
bodybuilding purposes). The results, displayed in Figure 5(A), suggest a polarity
between gym work and other bodybuilding practices. This indicates, as one
might guess, that working out with weights in a gym is by no means indicative
of ‘bodybuilding’. Many more people ‘work out’ than ‘bodybuild’.
Pushing this further still, I ran the MDS analysis again,14 using the same
respondents and variables, but this time mapping respondents (cases) rather than
the variables. Using MDS in this way allows us to conduct a form of cluster
analysis. We are able to see what sorts of ‘camps’ our respondents fall into when
clustered according to specific variables, and which respondents fall into which
camps.
The results for this analysis are given in Figure 5(B). In all, the 82 respon-
dents who exercised for more than five hours in the seven days prior to filling
in my questionnaire break down into five clear clusters which we can profile
in terms of their relationship to the basic variables we have used in our MDS
analysis (I double-checked this using hierarchical cluster analysis and achieved
the same results).15 Seventy-eight of the respondents fall into one of two
clusters (Cluster B and Cluster C). Cluster C contains all respondents who
exercised for more than five hours but did not work out in the gym or engage
in any bodybuilding-related practice (n = 55). Cluster B contains respondents
who did work out in the gym but didn’t do any other bodybuilding-related
activity (n = 23). Cluster A, to the left, contains two respondents, both of
whom worked out in the gym and used a bodybuilding supplement. Further
left again, in the bottom left corner, we have a single respondent (case 23) who
registered positively for all indicators of bodybuilding, including steroid use.
Finally, towards the bottom right of the plot we have ‘case 52’, a respondent
Cluster B
Cluster A n = 23
n =2
Cluster C
n = 55
who had exercised for over five hours in the previous seven days and had read
a bodybuilding magazine in the previous month, but hadn’t worked out in the
gym and hadn’t done anything else bodybuilding related. In effect then, the
association between five hours+ exercise in a week and the various ‘body-
building’ variables turn out to be more complex than Figure 4 suggests. Most
of the 82 respondents who did a lot of exercise didn’t do it in the gym, with
weights, and the majority of those who did, did not engage in any other
‘bodybuilding’ RBTs. Only three respondents out of the 82 really ‘profile’ like
bodybuilders on the closer inspection that this further analysis affords us. In
addition, ‘case 52’ indicates that a person may do a lot of exercise and read
about bodybuilding but not engage in the central RBT of the bodybuilding
world (i.e. working out in the gym).
Having identified our clusters in this way we could, in a more detailed
analysis, proceed to profile our clusters socially and, if we had the data, perhaps
biographically also. I am not going to push the analysis in this direction,
however. My purpose has simply been to show how we can use MDS to further
extend the analysis of Figure 4. At which point we can return to it.
At the bottom of Figure 4 we have a cluster of tattooing- and piercing-
related variables. These are widely spread, reflecting the fact that the ‘>3
tattoos’ and ‘1–3 tattoos’ variables are coded in a mutually exclusive fashion in
the data set and so could not be positively associated, but their very distinct
location, low on the plot, suggests a genuinely meaningful cluster based around
these less conventional (perhaps primitivist-inspired) modifications. Interest-
ingly, bellybutton piercing, which is also included on the plot, is located at
some distance from this cluster (in the ‘feminine’ segment), suggesting that this
particular type of piercing has a different location within the societal repertoire
of RBTs. It has migrated beyond the bounds of an ‘alternative’ subculture on
the margins into the mainstream (that is, the intermediate zone/feminine
segment).
As my discussion of the ‘heavy exercise’ cluster demonstrates, Figure 4 is just
the beginning of what could be a much more detailed analysis. Each cluster can
be further analysed and mapped. As it stands, however, Figure 4 and my analysis
of it are useful because they allow us to conceive of body modification/mainten-
ance as a structured and differentiated social space. Moreover, they locate
concrete practices in that space in accordance with real frequency distributions
and real statistical associations. As such they invite and facilitate explanations
that will advance our understanding of ‘body work’. In the final section of the
article I offer a preliminary explanatory framework.
within this zone is an effect of power balances which enforce their normalcy and
involves a forgetting of the historical struggles involved in the achievement of
that state of normalcy.
Like the core zone, the intermediate zone is historically variable. It is subject
to the movements of fashion, and contains practices which might previously have
belonged either to the core or the marginal zone. Indeed, it is not at all
uncommon for techniques from the marginal zone, a zone of innovation and
experimentation, to migrate into the intermediate zone after being appropriated
by the fashion and advertising industries. Likewise, when these techniques fall
out of fashion they will tend to fall back into the marginal zone. And migration
can occur between the core and intermediate zones also. Once fashionable tech-
niques, appropriated because they mark distinction, can, as noted above, be
appropriated by moralizing movements intent on imposing them upon society as
a whole, or alternatively can become so fashionable as to outgrow fashion (and
certainly their power of distinction) to become normal.
towards involuntary weight gain and obesity, which has triggered an increase in
dieting and working out, being a clear illustration of this (see Crossley, 2004b).
However, we have here a basic framework which invites further elaboration –
both qualitative and quantitative – and which takes our understanding of body
modification/maintenance forward.
Conclusion
In this article I have attempted to push the sociological analysis of body modifi-
cation/maintenance forward on substantive, theoretical and methodological
fronts. Practices of modification/maintenance, I have argued, can be understood
as reflexive body techniques; social techniques, collectively shared but individu-
ally rooted in the corporeal schemas of agents. The concept of RBTs has many
advantages, but specifically it is important because it emphasizes reflexivity,
refuses dualism and facilitates empirical investigation.
Each society, I have suggested, has a specific repertoire of such techniques but
techniques are not equally or evenly distributed and diffused throughout the
social body. Specifically I have argued (i) that RBTs fall into different frequency
zones (core, intermediate and marginal), and (ii) that further differentiation is
visible in the intermediate and marginal zones, as RBTs cluster in accordance
with specific subcultures, fields and movements (and the concerns associated
with these fields). Alerting ourselves to these distinctions is important, I have
suggested, because different clusters of RBTs have different socio-logics and need
to be explained in different ways. Where some RBTs are strongly encouraged, if
not made compulsory, for example, others are outlawed, and we must account
for this in our attempts to explain these practices.
Alongside this substantive engagement it has been my intention to explore, for
methodological purposes, the utility of multi-dimensional scaling. This method,
like any method, has limitations. What I have attempted to show in this article,
however, is the way in which we can use it to map out the social distribution of
RBTs, identifying their distinct patterns of clustering. Moreover, in relation to
my ‘heavy exercise’ cluster, I showed how we might move from general clusters
of RBTs through to a cluster of social agents themselves, a step which may, in
turn, facilitate a sophisticated form of social and biographical profiling of
particular practitioner communities – e.g. primitivist communities or bodybuild-
ing communities.
Many loose ends remain at the end of the article. Many avenues are yet to be
explored. This is a good thing. It invites and facilitates further analysis. One very
obvious way in which the analysis might be extended, however, is through a
more detailed exploration of the specific clusters marked out on my map (Figure
4). This is an exercise that might use statistical procedures such as MDS, but
perhaps in conjunction with the more ethnographical and qualitative forms of
analysis which the analysis of RBTs also affords.
Notes
1. Things are not quite so simple as this, as there are different techniques for determining
distances, and any mapping exercise involves a process of variable selection – not least because outliers
can throw the map off course. These processes of selection are part of the interpretative process and
do shape the map. Nevertheless, there are still limits built into the use of the procedure about what its
possible output can be.
2. E.g. concerning beauty, health, tattooing, bodybuilding.
3. The sample was 39.5 percent male and 60.5 percent female.
4. In terms of age, 14.1 percent of the sample were 16–19-year-olds, 25.7 percent were in their 20s,
27 percent in their 30s, 14.1 percent in their 40s, 15.1 percent in their 50s, 2.6 percent in their 60s, 1
percent in their 70s, 0.3 percent in their 80s.
5. In the sample, 29.3 percent were students, 4.6 percent retired, 1.3 percent unemployed, 4.3
percent unskilled manual workers, 5.9 percent semi-skilled manual workers, 6.9 percent skilled manual
workers, 14.8 percent clerical workers, 10.5 percent managerial grade workers, 19.1 percent
professionals, and 3.3 percent owned small businesses.
6. In the sample, 88 percent identified as White, 3.7 percent as Indian, 3.7 percent as Pakistani, 1.3
percent as black, 1 percent as Chinese, 1.3 percent as mixed race and 1 percent in other terms.
7. The religious breakdown was: 30.2 percent Protestant, 10.3 percent Catholic, 8.3 percent
unspecified Christian, 5.3 percent Muslim, 2.7 percent Hindu, 0.7 percent Jewish, 0.7 percent Sikh, 0.7
percent ‘other’ religious and 41.2 percent ‘not religious’.
8. Some of my respondents were recruited from the health club I attend and have been studying
ethnographically. They do represent a sampling bias in terms of a ‘body’ questionnaire. However, this
was a minority of the sample.
9. ‘Disposition’ is the usual English translation of ‘hexis’ and ‘habitus’. ‘Habit’ would have
worked but, as Camic (1986) notes, its meaning has been considerably changed and degraded in the
20th century, largely under the impact of psychological/physiological behaviourism.
10. I use Merleau-Ponty’s concept of ‘institution’ here, which is a modification of the phenom-
enological concept of ‘constitution’. ‘Constitution’, Merleau-Ponty argues, suggests that the agent
bestows meaning and order ex nihilo, where ‘institution’ suggests that the agent deploys socially
acquired schemas of meaning and order – ‘techniques’ in this case (Merleau-Ponty, 1979).
11. As Merleau-Ponty (1968) notes, to recognize their image in the mirror infants must first learn
to ‘derealize’ the image; that is, they must learn to see it as an image and not as another person. Then
they must learn to use the mirror to manipulate their own image, matching actions to their inverted
reflections. Much work in child development focuses upon this process whereby children learn to use
the mirror image to manipulate aspects of their own appearance (e.g. Amsterdam, 1971). And
Romanyshyn (1982) notes how this is extended in adolescent and adult life, where we play with and
in front of the mirror, rehearsing anticipated agentic ‘performances’ and fantasizing. Mirror play is a
complex and acquired technique.
12. One can use different measures of ‘distance’ in multi-dimensional scaling, each of which yields
different results. If frequency distribution is important and one is dealing with head counts then Phi-
squared is often a useful measure. I have used Phi-squared here.
13. On this occasion ‘Euclidean square distance’ proved a more useful measure of distance for the
MDS analysis.
14. Again distances were measured using Euclidean square distance. See notes 12 and 13.
15. Cluster membership is not clear from the diagram as cases overlap in the same space. It is
available as output for MDS on SPSS. I have not included it here because the existence and basic profile
of the clusters is more important than their actual composition, and because case numbers are arbitrary
and are only of use to us if we want to further explore the details of cluster members, which we do
not in this context.
References
Amsterdam, B. (1971) ‘Mirror Self-image Reactions Before Age Two’, Developmental Psychobiology
5(4): 297–305.
Aristotle (1955) The Ethics. Harmondsworth: Penguin.
Bartky, S. (1993) Gender and Domination. London: Routledge.
Baudrillard, J. (1999) The Consumer Society. London: Sage.
Bettleheim, B. (1986) The Informed Heart. Harmondsworth: Penguin.
Bordo, S. (1993) Unbearable Weight. Berkeley: University of California Press.
Bourdieu, P. (1977) ‘Remarques provisoires sur la perception sociale du corps’, Actes de la Recherche
en sciences sociales 14: 51–4.
Bourdieu, P. (1979) Outline of a Theory of Practice. Cambridge: Cambridge University Press.
Bourdieu, P. (1984) Distinction. London: Routledge and Kegan Paul.
Bourdieu, P. (1989) Practical Reason. Cambridge: Polity.
Camic, C. (1986) ‘The Matter of Habit’, American Journal of Sociology 91: 1039–87.
Canter, D. (ed.) (1985) Facet Theory. New York: Springer-Verlag.
Charlesworth, S. (2000) The Phenomenology of Working Class Experience. Cambridge: Cambridge
University Press.
Connell, R. (1987) Gender and Power. Cambridge: Polity.
Cresswell, M. (2003) ‘Psy-crats and Survivors’ (working paper, Department of Sociology, University
of Manchester).
Crossley, N. (1995) ‘Body Techniques, Agency and Intercorporeality: On Goffman’s Relations in
Public’, Sociology 29(1): 133–49.
Crossley, N. (1996) Intersubjectivity. London: Sage.
Crossley, N. (2001) The Social Body. London: Sage.
Crossley, N. (2002) Making Sense of Social Movements. Buckingham: Open University Press.
Crossley, N. (2003) ‘From Reproduction to Transformation: Social Movement Fields and the Radical
Habitus’, Theory, Culture & Society 20(6): 43–68.
Crossley, N. (2004a) ‘The Circuit Trainer’s Habitus: Reflexive Body Techniques and the Sociality of
the Workout’, Body & Society 10(1): 37–70.
Crossley, N. (2004b) ‘Fat is a Sociological Issue: Obesity in Late Modern, Body-conscious Societies’,
Health and Social Theory 2(3): 222–53.
Crossley, N. (2004c) ‘Ritual, Body Technique and (Inter)subjectivity’, pp. 31–51 in K. Schilbrack (ed.)
Thinking through Ritual: Philosophical Perspectives. London: Routledge.
Davis, K. (1995) Reshaping the Female Body. London: Routledge.
DeMello, M. (2000) Bodies of Inscription. London: Duke University Press.
Durkheim, E. (1915) The Elementary Forms of Religious Life. New York: Free Press.
Durkheim, E. and M. Mauss (1963) Primitive Classification. London: Cohen and West.
Elias, N. (1984) The Civilising Process. Oxford: Blackwell.
Nick Crossley is Reader in Sociology at the University of Manchester (UK). He has published widely
on the issue of human embodiment. His first book on the subject, The Social Body, was published by
Sage in 2001. He is currently writing a follow-on book, Reflexions in the Flesh: Embodiment in Late
Modernity, which will be published by McGraw-Hill/Open University Press.