dimensions of the spatiality of daily lives,
emphasizing how the ‘imagined geographies’ of
employers and workers combine with workplace
cultures and social networks to reinforce inter-
secting differences of gender, class and ethnicity
(also see Hanson and Pratt, 1995). Illustrating the
intense geographical constraints experienced by
many of those living in poor neighbourhoods
(who include first-generation immigrants who
have travelled large distances before arriving in
Worcester), they argue ‘that there is a stickiness
to identity that is grounded in the fact that many
women’s lives are lived locally’ (Pratt and
Hanson, 1994: 25), which often leads to a strong
sense of differences between women, and which
masks interdependencies and commonalities.
As many studies have shown, mobility, espe-
cially commuting times and distances, varies
markedly between different social groups (for a
review concerned with gender in relation to other
factors see Law, 1999). Put another way, the scale
at which daily lives are lived varies very signifi-
cantly; indeed the idea of ‘scale’ is itself a social
construct that generates and sustains interlocking
differences – gender, class, race and so on
(Marston, 2000; Ruddick, 1996; N. Smith, 1993).
Space and spatialities clearly fracture (and are
fractured by) gender and other categories and
therefore contribute to a proliferation of versions
of feminine and masculine subjectivity. In so far
as the politics of identity grounds political inter-
ests in particular experiences, this proliferation
has resulted in a splintering of feminist politics
and a recognition of multiple feminisms. Taken
to an extreme, this implies a conceptualization of
subjectivities as corresponding directly to posi-
tions people occupy within the frameworks pro-
ducing difference, including those of gender,
class, race and place. But people are not bound to
pre-given subject positions in this way: experi-
ences of gender, class, race, place and so on are
themselves social constructs, and influence
rather than determine people’s political affilia-
tions and actions. Moreover the notion of corre-
spondence between a particular form of
subjectivity and a particular structural position
‘buys into’ the notion of individuals as centred,
bounded, autonomous entities.
We would argue that while the imposition of
socially constructed and therefore ‘fictional’
conceptual labels (such as race, gender and so
on) can be used for politically dubious purposes,
and so must be questioned and never treated as
‘natural’ (Irigaray, 1993), they have a long
history of usage and cannot be simply abandoned
by emancipatory political projects. Moreover,such separations and categorizations can be used
strategically, for example to highlight those parts
of the webs of our identities (Griffiths, 1995)
around which changes most urgently need to be
made (in relation, for example, to experience of
racism, heterosexism, ageism and so on). But,
there are still potential problems linked with
feminist usage and understanding of the concept
‘woman’: being a woman can mean different
things at different times in different places to dif-
ferent people, and no one definition of woman-
hood would be appropriate across the broad board
of women’s experience. Why then would we wish
to argue for and present a ‘definition’ at all?
While it is clear that women in no sense con-
stitute a unitary group, a constructive and usable
understanding of the concept of womanhood is
still required for the purposes of feminist theory
and practice. (After all, for whom are femi-
nists concerned, if not ‘women’ as a group?)
Wittgenstein’s anti-essentialist philosophical
approach, and particularly his notion of ‘family
resemblance’, is potentially useful here. Following
Davidson and Smith’s (1999) feminist interpreta-
tion of this notion, we would contend that con-
cepts such as ‘woman’ need not be understood
and employed representationally, as referring to
certain essential features shared by all members
of the group ‘women’. Rather, much as
Wittgenstein argues that we identify members of
a family in terms of, for example, resemblances
in eye colour, build or patterns of speech, and not
through a particular feature or features common
to all, concepts such as ‘games’ or, according to
Davidson and Smith, ‘women’ can be understood
productively in terms of a ‘complicated network
of similarities overlapping and crisscrossing’
(1994: 74). That is to say, when we use the term
‘woman’ we need not use it in a way that speci-
fies and depends upon any particular shared
experiences or features for its sense. In fact, what
we understand and what we communicate by the
term will alter according to the context within
which it is used, and this shifting, contextualized
sense of how a concept operates in practice is
extremely, multiply, useful for feminist
theorists and cultural geographers.
There are powerful reasons why we should
resist the temptation to produce a precise defini-
tion that attempts to ‘pin down’ exactly what
being a woman entails. Such exact conceptual
boundaries are likely to be restrictive, exclusive
and counter-productive politically. In so far as
we do however require a working definition of
‘woman’, we would argue in favour of ‘strategic
imprecision’, that we have no need to settle or334 PLACING SUBJECTIVITIES3029-ch17.qxd 03-10-02 10:56 AM Page 334