Professional Documents
Culture Documents
Derek Gregory - Social Relations and Spatial Structures (1985) PDF
Derek Gregory - Social Relations and Spatial Structures (1985) PDF
Edited by
Derek Gregory
Professor of Geography , University of British Columbia
and
John Urry
Professor of Sociology , University of Lancaster
M
MACMILLAN
ISBN 978-0-333-35403-2 ISBN 978-1-349-27935-7 (eBook)
DOI 10.1007/978-1-349-27935-7
Editoria l matter and selection © Derek Gregory and John Urry 1985
Individual chapters © Philip Cooke, Anthony Giddens,
Derek Gregory, David Harvey, Doreen Massey, R. E. Pahl,
Allen Pred, Peter Saunders, Andrew Sayer , Edward W. Soja,
Nigel Thrift, John Urry, Richard Walker, Alan Warde 1985
Reprinted 1994
Contents
List ofFigures VB
Preface Vlll
List ofContributors x
1 Introduction
by Derek Gregory and John Urry 1
2 New Directions in Space
by Doreen Massey 9
3 Social Relations, Space and Time
by John Urry 20
4 The Difference that Space Makes
by Andrew Sayer 49
5 Space, the City and Urban Sociology
by Peter Saunders 67
6 The Spatiality of Social Life: Towards a Transformative
Retheorisation
by Edward W. Soja 90
7 The Geopolitics of Capitalism
by David Harvey 128
8 Class, Division of Labour and Employment in Space
by Richard Walker 164
9 Spatial Change, Politics and the Division of Labour
by Alan Warde 190
10 Class Practices as Regional Markers: A Contribution to
Labour Geography
by Philip Cooke 213
vi Contents
We are very grateful to the contributors to this book, not only for the
chapters they have produced but also for their other stimulating
contributions to the various debates considered here. Although the
volume is in no way a 'manifesto', it has been very much a
collaborative project with all sorts ofother arguments weaving in and
out of its main themes and contributing to a sense of intellectual
excitement at the contemporary cross-fertilisations between human
geography and sociology.
Derek Gregory is particularly grateful to Michael Dear, Felix
Driver , Tony Giddens, Peter Gould, Jack Langton, Chris Philo and
Nigel Thrift for numerous discussions of many of the themes addres-
sed in this volume, and John Urry is similarly grateful for discussions
with members of the CSE Regionalism Group and the Lancaster
Regionalism Group: especially Mike Savage, Dan Shapiro, Sylvia
Walby and Alan Warde ; and with other members of the Department
of Sociology at Lancaster.
We are both indebted to Steve Kennedy at Macmillan for his
encouragement and penetrating editorial criticism, to Elizabeth
Black for copy-editing a difficultmanuscript with both sensitivity and
skill, and to Christopher Philo for compiling the index.
DEREK GREGORY
JOHNURRY
section off and study. All it had - or all it appeared to have - was a
dimension: space, distance .
So what happened was that geography set itself up as 'the science of
the spatial'. There were spatial laws, spatial relationships, spatial
processes. There was a notion that there were certain principles of
spatial interaction which could be studied devoid of their social
content. At a less highflown, but socially more significant, level, there
was an obsession with the identification of spatial regularities and an
urge to explain them by spatial factors. The explanation of geogra-
phical patterns, it was argued, lay within the spatial. There was no
need to look further. As we have said, the inner city problem was
coming to prominence at that time, and it provides a good case in
point. For there was a clear spatial regularity. Geographers (and by
no means only geographers), assuming that behind every spatial
pattern lies a spatial cause, looked to the characteristics of the inner
cities (in other words to the characteristics of the location itself) to
explain their demise. This was probably the initial geographical
version of blaming the victims for their own problems. What ·had
happened in this combination of the rush for positivism and the need
for an identity in the institutionalised academic division oflabour was
that geographers and geography had made some astonishing claims-
that there was a world of the purely spatial, spatial laws devoid of
substance or content, and spatial processes it was possible to wrench
out of their social context.
The bulk of the 1970sand indeed until now, has been taken up with
arguing that this is an untenable position, that there are no such
things as purely spatial processes; there are only particular social
processes operating over space. It was now argued that what was
happening in the inner cities - the loss ofindustrialjobs - had more to
do with industry than with the cities, that the cities were at the sharp
end of what was to become a much more general process of de-
industrialisation. It was also argued that the fact that they were at the
sharp end had more to do with the character of their industry (that it
was less competitive, more labour-intensive, with lower productivity)
than with the cities' characteristics as locations. In other words, it was
argued - to such an extent that it became a theme tune of the times -
that the spatial is a social construct. It is a position which, in very
general terms, still holds good.
But, in its turn, it also entailed some serious problems. First - and
to talk only in terms of the academic division oflabour - it meant that
12 New Directions in Space
Within studies of industrial location the big advance of the 1970s was
to get production itself into the equation. Instead of the exclusive
concentration on 'location factors ' and on explaining one spatial
pattern by another, it was argued that it was necessary to examine
what was going on inside the factory or office, that 'location factors'
were themselves the result of changes in production, and that in order
to understand spatial change it was necessary to go behind 'spatial
factors ' and spatial pattern. The kind of substantive arguments
around which this debate developed in Britain were the inner city
problem (already discussed) and the evaluation of the impact of
regional policy. Formally it led to a whole series of studies in which
the dominant lines of causality were understood to run from causes of
changes in production, through those changes themselves (changes in
technology, reorganisations of the labour process, etc.) to an evalua-
tion of the impact of those changes on the locational requirements of
production, to the explanation of the (changes in) spatial pattern.
There were, of course, variants. There was a strong line of argument
that in many situations spatial considerations might be totally unim-
portant in the production of spatial outcomes. This might be true in
cases where decisions were being made about closures, for example,
or jobs were being lost through intensification of the labour process.
In such circumstances spatial patterns of job loss (whether resulting
from the operation of the market or from decisions within multi-
locational companies) might result entirely from internal characteris-
tics of individual plants rather than from any more obviously spatial
factors, such as the wider characteristics of the areas within which
plants were located. But in very general terms the explanations tended
to follow the stages indicated above. It was essentially a one-way,
linear sequence. What was crucial to it (and what was new) was that
spatial change was understood as being an outcome of production
change .
It was, as I have said, an important point to make at the time, and it
needed driving home, but it also now needs to be modified, for it is not
the whole of the story - or rather, the story is not quite so simple. The
sequence 'production change ~ spatial change ' ignores the crucial
impact of spatially-organised locational opportunities (or the lack of
them) and of the use of distance and spatial separation themselves.
Each of them can have an impact on what happens to production.
The use by capital oflocational change as part ofa wider strategy for
14 New Directions in Space
The fact that this was happening was recognised quite early. The
work of Hymer and ofWestaway was pre-eminent in pointing out the
performance of different levels and kinds of managerial functions at
different locations.' There was a hierarchy, social and spatial, from
the strategic to the routine. There are two points to be made about
this. First, what are really 'stretched out over space' are the relations
ofeconomic ownership and possession." It is a hierarchy ofparticular
places in these relations, related to the performance of particular
functions. And those members of the managerial structure occupying
different places within that hierarchy thereby have different places
also in relation to the overall antagonism, within the relations of
production, between capital and labour. Such differences are one
element only in the determination of place in social structure. Rough-
ly, the nearer to headquarters you are functionally, the higher your
social status, and since the hierarchies have a spatial form, social and
spatial differentiation are clearly related .
But second, it is the nature of this relationship (between social form
- or in this case more accurately, place in the relations of economic
ownership and possession within capitalist production - and spatial
form) which is interesting. It is usually characterised, at least implicit-
ly, as though a pre-existing managerial hierarchy is, in explanatory
terms, subsequently allocated across space, as though the process is
akin to dealing out between locations a pre-existing pack of cards.
There is an implicit, perhaps unnoticed, assumption that the extended
hierarchies of management and control were developed with the
aspatial concentration of capital and then were distributed between
plants as the large firms developed multi-Iocational patterns.
It did not, of course, happen like that. Becoming multi-Iocational
was part and parcel of the growth oflarge firms; it was also part of the
extension of the hierarchies of management and control. What is
important to the argument here is that that very fact was one of the
things which influenced the social divisions which developed within
managerial hierarchies . The fact of'multi-locationality itself will push
towards the adoption of particular kinds of divisions in terms of the
formal disaggregation offunctions within the overall social relations
of production. The hierarchy of managerial functions, and of man-
agerial personnel , which stretches, say, from the headquarters in the
metropolitan region, to a regional headquarters in a smaller city, to
an outlying branch plant, is in part moulded in its form by the fact of
spatial separation. So, too, is the hierarchy of social status and the
detail of social structure. These are not just social divisions of labour
16 New Directions in Space
distributed over space; the form of the social division is itself influen-
ced by the fact that it is distributed over space.
It is also more than that. Spatial distribution and contrasting
geographical contexts - spatial structure in a wider sense - will also
mould the social divisions which exist. Certainly there is evidence for
this in Britain today . At the lower end of management hierarchies, at
the level of a lowly local branch-plant, management can be reduced
to fairly routine administration and management can feel almost as
much on the receivingend of, and in the dark about, decisions handed
down from above as is the production workforce. That sense of
differentiation, and potential antagonism, can be reinforced both by
relative geographical isolation from other elements of 'their man-
agerial class' and, when branch plants are in far-flung, production-
only regions, by their very different social context. When managerial
hierarchies within a country tend to take on fairly consistent in-
terregional forms the social divisions which might anyway exist along
them may be moulded and accentuated by the very fact of spatial
division. The growing managerial strata in the UK would certainly
merit further study along these lines.
Thus the fact that social processes take place over space and in a
geographically-differentiated world affects their operation. Many of
the most significant changes in social structure which have taken
place in recent years in the UK could not really even be sensibly
conceptualised outside a spatial geographically variegated form .
From a point in the early 1960swhere shortage oflabour seemed for a
moment to be a block on the hoped-for expansion of production,
what has actually emerged is that the restructuring of British capital
has taken place in a context of decelerating job-growth and sub-
sequent decline. At the beginning of the period what was necessary
was an expanded supply of labour. Later the requirement became
much more simply cheaper labour. In both cases spatial expansion,
geographical movement , was essential to capital's way out of its
problems. In the first place the peripheral regions were needed as
extensions to the supply of labour. Later their labour was used to
undercut the strength of labour elsewhere. Since the 1960s,what new
jobs have been created in the economy have, on the whole, gone
elsewhere. In the early 1960s in economic terms labour in Britain
was in a relatively strong position. It is no longer so, and not just
because of the 'world' recession and the particularly spectacular
collapse of employment in Britain. Spatial reorganisation has also
been an important weapon in the changing balance of class forces.
Doreen Massey 17
That would have been true even had there been no recession, as
industry expanded into new sources oflabour, which in so many cases
also meant new places, and that in tum functioned not only to
expand, but to lower the price of labour below what it would other-
wise have been.
Understanding the spatial organisation of society, then, is crucial.
It is central to our understanding of the way in which social processes
work out, possibly to our conceptualisation of some of those proces-
ses in the first place, and certainly to our ability to act on them
politically.
Thequestions ahead
the social, can the social be theorised autonomously from its spatial
form , requirements and implications? What, in other words, is the
effective import for research of this widening recognition of the
significance of the spatial and the geographical?
A number of options appear to be available and to be being
pursued. First, if we really mean that it is impossible to conceptualise
social processes and structures outside their spatial form and spatial
implications, then the latter must also be incorporated into our initial
formulations and definitions, into our basic concepts - of capital, for
instance, to point to the most obvious. That is no easy task. There are
some clear possibilities - in the spatial implications of the self-expan-
sion of capital, for instance. Yet so often the results seem either
incredibly limited and banal (reinforcing the difficulties already being
experienced in elaborating a wide and full range of generative
mechanisms and causal properties) or we get involved in rather
arbitrary series of multiple possibilities for how social processes will
operate in a range of different and individually and formally-specified
'spatial contexts'. Neither seems on the face of it a very attractive
possibility.
Yet necessary relations are not aspatial. How can 'necessary' rela-
tions be necessary if they depend, for example, on contiguity - unless
that contiguity is specified? Inherent causal properties may depend on
spatial form . What then happens to the argument that 'the spatial' is
necessarily contingent? As a number of commentators have re-
marked, the whole realm of politics and culture, like the spatial, often
also seems to be relegated to the arena of the contingent. The two
seem to be closely related. But to accept that position would leave the
range of possible necessary relations extremely limited, and limited
primarily to the economic. Part of what is fundamentally at issue here
is the reassessment of our definition of necessary relations within the
social sciences.
It may indeed be argued that 'geography' only comes on to the
scene at a later stage of analysis - that it is inherently contingent. The
question then is what becomes of the injunction that social processes
are constituted spatially? Williams has made a trenchant argument on
the parallel relations between the social and the natural; that they are
impossible to conceptualise separately and then 'put together'.sAnd
what ofUrry's important point that the operation of social laws over
space may change the nature of those laws themselves?" Equating the
geographical with the concrete seems to be a very minimalist position.
Research developments at the moment, however, seem to tend in
Doreen Massey 19
the latter direction. This, I think, reflects in part the extreme difficulty
and problematical nature of the first project. But it also reflects wider
theoretical and political shifts. The unique is back on the agenda. The
recognition and understanding of particularity is theoretically the
mirror image of, and politically the equally-necessary obverse of,
pointing to the generality and necessity of underlying mechanisms.
And much ofthis particularity is spatially structured. In that sense the
widening recognition of the importance of the geography of social
structures, and of the fact that social structures are in the concrete
world constituted geographically, is in tune with a widening and more
general appreciation of specificity and of internal form. It is certainly
important, both politically and theoretically, to developments within
the social sciences. Even while some of the more fundamental
theoretical questions remain unresolved (or, as they may , remain
forever in dispute), it is vital that we accept, and continue to insist
upon, the importance of space and spatial variations in concrete
analysis.
Notes
Introduction
4. The view of the social world being sustained here - that it compr-
ises a number ofinterdependent, mutually modifying, four-dimen-
sional, space-time entities - means that we are confronted by a
peculiarly complex 'open system', as Bhaskar puts it. He says:
It is characteristic of open systems that two or more mechan-
isms, perhaps of radically different kinds, combine to produce
effects; so that because we do not know ex ante which mechan -
isms will actually be at work (and perhaps have no knowledge of
their mode of articulation) events are not deductively predicta-
ble. Most events in open systems must thus be regarded as
'conjunctures'P"
Two methodological points followfrom this. First, as Sayer points
out , what is involved is a form of retroductive inference Ii la
Hanson." That is to say, science involves the postulation of
theories in order to explain empirical phenomena - these theories
are built up 'in reverse', retroductively, and they are developed in
John Urry 29
Theanalysis of space
I have so far talked generally of the relations between time, space and
the social world. I have made relatively little reference as to how we
might analyse such relations apart from noting the importance of the
distinction between enduring entities and empirical events. Sayer has
recently tried to discuss some of these issues - in the case of studies of
industrial location - by distinguishing between external , contingent
relations, on the one hand, and internal, necessary relations, on the
other." The former consist of those relations where the objects in
question do not stand in any necessary relationship with each other,
and when they can exist independently of each other and of the
relationship between them . The latter, by contrast, consist of those
relations which are necessary for the very objects to exist, so that they
cannot exist without such relations. An example of the former would
be the decision of a particular company to employ workers from one
town rather than another; examples of the latter would be the rela-
tions between capital and wage-labour, or landlord and tenant. Sayer
points out that many relationships will in fact consist of some com-
bination of the two forms of relations so that specificempirical events
should not be viewed as simply resulting from either external and
contingent or internal and necessary relations. For example, he says:
John Urry 37
3. the degree to which more generally the social relations within civil
society are based on the 'local community' rather than on either
commodity relations or on the state. Where this is the case then this
involves sets of social relationships which are multiplex (neigh-
bours who are workmates who are leisure-time companions, etc.)
where 'everyone knows everyone else' and where these sets are
organised into a locally structured and delimited system." The
consequence of such a community structuring of the local civil
society is to produce communion so that one's neighbourhood has
an emotional meaning derived from the people who live there and
from the mutually supporting ties of trust , friendship and re-
ciprocity within that local civilsociety, even if this is a 'mutuality of
the oppressed'." Market transactions within such a place-bound
community are indelibly suffused by considerations of long-term
reciprocity and community. Interactions occur within a given
physical setting, and one's living space is necessarily personalised,
John Urry 41
There are two other crucial considerations here which I have not
considered : first, the spatial structurings of other social entities,
especially the state;82 and second, the analysis of how the complex
spatial/temporal interdependence between these diverse entities with
varying causal powers in fact produce specific sets of empirical
events." Many of the other chapters relate to these issues and I will
not consider them further except with regard to one issue, namely, the
conditions under which 'collective organisations' may develop out of
social classes or out of other social groupings. It is obvious that the
formation of such organisations is an absolutely crucial issue with
regard to the realisation of the causal powers of such classes or
groupings . But it is also an issue which relates very directly to the
complex interrelations between capitalist relations of production and
civil society.
The reason for the last claim stems from the classic analysis of such
issues by Olson.v' He maintains that it is not in fact correct to argue
that the absence of the kind of class conflicts that Marx expected
shows that Marx overestimated the strength of the rational choices
that workers will make. On the contrary, says Olson, it is precisely
because workers do act in terms of rational self-interest that class
actions and organisations are not more marked. This is because the
achievement of 'socialism' is a public good whose benefits accrue to
all workers whether or not they participate in the class struggle that
produced it. Class actions are thus a specific manifestation of the
general logic of collective action . Such action is very difficult to
produce and especially to sustain since the gains to be derived from it
are in the nature of public rather than private benefits. How do these
restrictions on the collective actions of social classes relate to the
issues already discussed?
Marx seems to have believed that individual selfishness and cal-
culation were the principal human characteristics generated by
capitalist production, and that as the production became more and
more widespread so rational self-interest and utilitarian calculation
would also become more common. But this makes it increasingly
John Urry 43
Conclusion
Notes
1. Within British sociology see Giddens (1981b) and Urry (1981c)as wellas
some of the literature relating to the 1982 British Sociological Associa-
tion Conference on 'Core and Periphery'; and see the chapters by Saun-
ders and Sayer in this volume.
John Urry 43
such through their relation: one cannot exist without the other. There
are also necessary conditions for the existence and reproduction of
objects and relations; the institution of marriage is a necessary condi-
tion for the occurrence of weddings. On the other hand patriarchy can
exist independently of capitalist social relations and hence their rela-
tion is said to be contingent, although some particular forms of either
may become interdependent. When necessity in the world is dis-
covered, we can make strong theoretical claims about it, by using
abstractions to isolate necessity from the web of contingencies in
which it usually exists. Where relations or conditions are contingent
(that is, neither necessary nor impossible) abstract theory must
remain agnostic about their form on any particular occasion or at
least make claims about what might happen in them conditional upon
an assumption about their existence. For example, it is contingent
where industrial capital is located and it can involve itself in the
production of an enormous variety of commodities, but if it happens
(contingently) to be involved in producing a fixed resource such as
coal (conditional statement) then obviously its location is con-
strained.
A third aspect of the realist approach adopted here concerns
causation. Objects are understood to possess causal powers and
liabilities to do or suffer certain things by virtue of their structure and
composition, but whether these powers or liabilities are activated
depends on contingently related conditions. For example, labour
power is a causal power necessarily possessed by people by virtue of
their physical and mental structure, but whether and with what effects
this power is exercised depends on contingently-related conditions
such as the availability of employment. It follows from the previous
point that abstract theory can recognise the powers or liabilities as
necessary properties of their objects but can only remain agnostic or
make conditional statements about what actually happens in con-
crete circumstances. To explain the latter, concrete research is needed
which discovers the actual (contingent) conditions in which the causal
mechanisms of interest are located .
These concepts are essential for understanding the difference that
space makes .
effect: that things must be in the right place if we are to use them or be
affected by them is commonsense. Yet , geographers apart, few social
scientists have paid any attention to space and the difference is
supposedlymakes. Recent work on space by human geographers has
reached the conclusion that space can only be understood in terms of
the objects and processes that constitute it, with the implication that
the study of space must be rooted in social theory. In the case of
Marxist theory, there have been re-readings or plunderings of Marx's
works in search of insights about space ." These have generated the
inaccurate impression that Marx had a great deal to say about space.
In fact , as Saunders has observed, neither Marx, Durkheim nor
Weber paid more than passing attention to space ." I want to argue
that while space makes a difference, these and other social theorists
have been largely justified in giving it so little attention in their
abstract theoretical work, although 'the position with regard to con-
crete research is quite different. This may seem a surprising combina-
tion of ideas, at least until we look more closely at the meaning of
'space' .
Much has been written on this in geography and although a virtual
consensus about concepts of space has emerged it will be argued that
it is not yet a satisfactory one." The clarification of its meaning is a
metaphysical problem. Like many such problems, it involves a term
which is both thoroughly familiar and apparently unproblematic,
and yet mysterious: explications of the term are necessary to distin-
guish legitimate from confused uses, but they will inevitably produce
a possibly infuriating mixture of insights and leaden banalities.
Whatever the popular hostility to this kind of discussion, the real
prisoners of metaphysics are those who think they do not need it!
The usual discussions that have taken place in geography distin-
guish between an absolute concept in which space is empty and a
relative concept in which space only exists where it is constituted by
matter. The absolute view is implicit in commonsense, but is inco-
herent because 'what is empty is nothing, and what is nothing cannot
be'," and because nothingness is implied to be capable of having an
effect - as when we speak of the 'friction of distance' or the 'effect of
space'. Terms such as these are really shorthand for frictions between
particular substances which constitute space, and as we know from
natural science, the coefficient of friction varies from substance to
substance." '
However, what is often overlooked by advocates of the relative
concept is that while space is constituted by objects it is not reducible
to them. Following Harre," the subtleties of the relative concept can
52 The Difference that Space Makes
related objects or practices) and with the causal powers and liabilities
necessarily possessed by objects in virtue of their nature. As it is
contingent whether and how these are exercised (since it is affected by
contingent conditions) concrete empirical research is needed to deter-
mine the actual effects of the causal powers .
Abstract social theory need only consider space insofar as neces-
sary properties of objects are involved, and this does not amount to
very much . It must acknowledge that all matter has spatial extension
and hence that processes do not take place on the head of a pin and
that no two objects can occupy the same (relative) place at the same
time. So, for example, in an abstract discussion of rent we can note
that even if people do not need land for its soil or minerals, they need
it as room for their spatially-extended activities and that perhaps also
they need it to be accessible to certain other objects if they are to do
certain things. Things then might be taken as necessary conditions
and therefore as material for abstract theory. From this many seem to
have concluded that social theory can and should be spatial.
But note just how little this says about actual spatial forms . To
observe that capital must be accessible to property-less workers says
very little about the actual spatial form of labour markets, which of
course is not surprising, given the contingent nature of this form .
Even where objects necessarily have certain spatial forms - as in the
linearity of many types of communication medium - this still leaves
room for enormous contingent variety, and the spatial form of the
conditions in which such objects exist, for example, settlement pat-
terns, is also contingent. Hence, while it is important for abstract
theory to be aware of the existence of space, the claims that can be
made about it are inevitably rather indifferent ones, such as Anthony
Giddens' 'distanciation' and 'time-space edges' .'? That the coales-
cence and integration of communities and economies which were
formerly independent produces an increased 'distanciation' of actors
and institutions from one another is an important feature of capitalist
societies, as we shall see later. But the actual nature of the difference
that this makes can again only be specified further by empirical
research on concrete instances . In other words, concepts such as
'distanciation' have a useful theoretical or meta-theoretical role but
they do not and should not be expected to say much about concrete
spatial forms .
Even where social theory deals with objects whose spatial form has
been deliberately arranged so as to take advantage of certain proper-
ties of the objects, for example, new towns, the variety of possible
Andrew Sayer 55
idea that space makes a difference, for the principle concerns the
position of mechanisms with respect to conditions with which they do
not happen to interact. If there is a major spatial variation in a
condition which does affect the outcome of the experiment - for
example, a strong anomaly in the earth's gravitational field in the
pendulum experiment - then space will make a difference in the sense
outlined above. Otherwise, the position of the experiment within
some arbitrary system of co-ordinates is irrelevant.) Social systems
are open because of the capacity of people to change their own nature
and their environment; hence closed system experiments are impossi-
ble. Theorists therefore have to rely upon 'thought-experiments'
using models whose assumptions ensure that the hypothetical system
is closed, for example, equilibrium assumptions in economic and
isotropic plains in geography.
In many cases the models are both spaceless and timeless. Insofar
as they therefore deny necessary features of behaviour, they can only
be of interest as fictions. The difference that spatial form makes can
be assessed using models only by assuming that certain (contingent)
spatial relations have a particular configuration - for example, the
uniform distribution of purchasing power in Loschian models of the
space economy.P Such models provide striking demonstrations of
the difference that space can make; for example, the aspatial perfect
competition model becomes a model of spatial monopolies. To make
such models dynamic one must again incorporate assumptions about
the form of sequences of contingently related events, such as a
particular rate of growth . Estimating the difference that space and
time make in such thought-experiments therefore requires prior
assumptions about the space-time patterns of causal powers." In
other words, the modeller has to determine in advance precisely those
circumstances which are contingent and relevant to the processes of
interest.
A less formal and more approximate way of speculating on possi-
ble effects of mechanisms in concrete systems as they are mediated by
spatial form , involves imagining a number of possible configurations
and making conditional statements about results. For example, it
could be hypothesised that in a competitive capitalist system, firms
which are located in regions having cheaper than average labour
power will be able to delay automation longer than those in regions
with scarce and expensive labour power. David Harvey's book, The
Limits to Capital." contains many such conditional statements.
Given the difficulty of anticipating the range of possible concrete
Andrew Sayer 57
Misconceptions of space
those constituents, only this time the constituents are further reduced
to just those which are socially produced, arbitrarily ignoring its non-
social constituents. The spatial is partly constituted by the social, but
it is reducible neither to natural nor social constituents. Certainly
spatial structures may have 'roots in production and class relations'
as Soja puts it, but there are plenty of other possible constituents -
natural and social - of space." (Note that the fact that natural
phenomena are only intelligible to us through social media has noth-
ing to do with the question of their production, which is independent
of social phenomena.)
The reduction of space to its constituents can be a fertile source of
confusion. Let us take two examples. The first is from Castells' The
Urban Question:
Let us examine this difficult passage more closely. To use the term
'space' is to use a contentless abstraction, and hence this description
of space as a material product [1] or a real object [2] is confusing
especially where as in [1] and [3] it is treated as equivalent to elements
such as 'men'. There are no structural laws 'that govern . . .
[the] . . . existence and transformation' of space as such, only the
elements that constitute space. The last sentence of the quotation [4]
can be taken in two ways: either that a general social theory should be
able to specify, in advance, the form (including the spatial form) of
concrete systems; or merely that in referring to objects which have
spatial extension, any general social theory is implicitly spatial. Only
the latter interpretation is tenable; as already noted, actual concrete
forms cannot be anticipated purely by reference to the implicit spatial
60 The Difference that Space Makes
Note how space appears to mean 'territory', and once this is one-
sidedly reduced to its political and ideological significance, we begin
to forget space in the sense of form or configuration altogether, and
the difference between this kind of analysis and unashamedly aspatial
theory diminishes.
As Sack has shown, we continually separate and recombine space
and substance in our thinking, but errors are often introduced
through an unsuccessful recombination. For example, it is popular to
discuss relations between developed and underdeveloped regions in
terms of 'centre and periphery'. Developed and underdeveloped re-
gions must have some spatial form but it does not have to be one of
centre and periphery, even loosely interpreted. Once we start to think
of the former in terms of the latter contingent spatial forms and hence
use them as surrogates for economic relations, some of which may be
necessary, confusion and error enter, for it becomes difficult to see
that the spatially peripheral is not necessarily functionally peripheral,
nor the spatially central, functionally central. Space and process in
this case cannot be successfully recombined at this level.
Crude separations and recombinations of space and substance are
even more common in everyday thinking as when people from the
North of England characterise the South-east as middle class and
refuse to believe that a town like Tunbridge Wells could have any
working-class people. The appropriate spatial unit for conceptualis-
ing class is people, not towns or regions.
In the preceding sections I have argued that the spatial specificity of
abstract theory differs markedly from that of research on concrete
systems and that insofar as space makes a difference to the latter,
concrete research must take spatial form into account even if it is not
directly interested in it. While I have been concerned to illustrate
some common errors in the treatment of space I do not want to
underestimate the genuine difficulties of conducting concrete (and
hence spatial) research even where such errors are avoided . Too often
Andrew Sayer 61
One source of difficulty in concrete research arises from the fact that
it is often simply not feasible to achieve a perfect recombination of
space and substance. Consider the example of research on labour
markets. We cannot expect to make much sense of the pattern of
events in such systems if we abstract from the spatial relations bet-
ween job-seekers and job-vacancies, and aggregate up all the local
area statistics into a national total as if there were a single, homogen-
eous, national market, located on the head of a pin. In a system as
open as this (in the sense defined earlier) there are likely to be few
intelligible links between aggregate quantitative relations and actual
causal interactions; hence the usual mismatches between numbers of
vacancies and unemployed. However, it would hardly be feasible to
take into account spatial form as it concerns each pair ofjob-seekers
and relevant vacancies. The usual compromise solution is to break
down the national data into fairly discrete labour market areas, but
this is only an approximate way of handling the effect of the spatial
form of the phenomena of interest. Given that social systems are so
heterogenous and open, it is small wonder that enduring empirical
regularities are not found , despite the hopes of positivist approaches.
Empirical studies which abstract from space can hardly be called
concrete because they conceal the actual forms of combination of
system elements. To those habituated to this kind of research , spatial
variations are little more than deviations from the norm and only of
interest as such; some even treat the deviations as somehow 'less real'
than the norm. Yet it is of course the norm or average which is the
statistical fiction; the individual voter, for example, however unusual,
is certainly no less real than the average voter .
But if there is a problem of explanation for single, concrete, open
social systems such as labour markets, the problem is all the greater in
studies which attempt to synthesise the concrete effects of the interac-
tion of many systems, such as the history of the industrial revolution,
or the nature of regional change. Researchers in such fields are faced
with a choice of relying upon aggregates and averages which describe
the whole system but occlude causality by ignoring spatial form or
62 The Difference that Space Makes
Conclusion
The recent interest in space and social theory should not allow us to
overlook the questions of how social theory has managed to pay
space scant attention without too much trouble and how those theor-
ists who have been preoccupied with space have not been able to say
very much about it. However, in the sphere of concrete studies, both
the difficulty of such research and its poor record in developing
explanations owes a great deal to the failure to consider spatial form.
Notes
1. Sayer (1981).
2. This interpretation of the concrete is derived from Marx's oft-quoted
statement 'The concrete concept is concrete because it is a synthesis of
many determinations, thus representing the unity of diverse aspects', in
the Grundrisse: Marx (1973) p. 101.
3. e.g. Harvey (1982) is in part such are-reading.
4. Saunders (1981).
5. Perhaps the best of this literature are Blaut (1961) and Sack (1980)
pp. 55-78, although the latter is marred by a failure to present a consis-
tent use of realistic philisophy.
6. Blaut (1961).
7. Sack (1973).
8. Harre (1971); Sack (1980).
9. Sack (1980).
10. Giddens (1981b).
II. See Bourdieu (1977).
12. Bhaskar (1975).
13. Harre (1970).
14. Losch (1954).
15. Cf. Sack's discussions of non-congruent or contiguous laws in Sack
(1980) pp.66ff.
16. Harvey (1982).
17. Sayer (1984).
18. Castells (1977).
19. Sayer (1979).
20. Lefebvre (1976a).
21. See Sayer (1984).
22. Saunders (1981).
23. Soja (1980).
24. Castells (1977) p. 115. I do not know what Castells means by 'conjunc-
turallaws'.
25. Lefebvre (1976b).
66 The Difference that Space Makes
variations in time and space lies at the very heart of the sociological
enterprise.
This argument has the further implication that, if sociology is to be
centrally concerned with the spatial dimension of human activity,
then its core substantive concern will be on spatially-defined units of
social organisation. Put another way, if space is central to social
theory, then the study of different socio-spatial arrangements -
villages, towns, cities - would appear to be fundamental to social
analysis. This is an implication which Giddens at least is happy to
follow, for he argues that the city constitutes the core object of
concern for sociology and that urban sociology lies at the very heart
of the discipline:
been four main attempts over the past seventy years to identify a
particular type ofsocial relationship or social process as distinctive to
social life in cities. All four attempts have failed, although all four
have bequeathed some interesting and useful legacies in the process.
The fact that they have failed suggests that the attempt to relate
particular social processes to particular spatial forms , and thereby to
establish a rationale for an 'urban' sociology concerned with the
study of cities, may be futile. This futility stems from the fact that the
city has ceased to represent a significant social unit of organisation in
advanced industrial societies. Just as a sociology of the tribe is likely
to appear oddly anachronistic in modern industrial societies where
tribes have long since been transcended as units of social organisa-
tion, so too does a sociology of the city.
Given that cities, however else we may define them, are clearly large
and complex (socially differentiated) human aggregates, it is not
surprising that 'urban' theorists have attempted to identify specific
patterns of interaction which are associated with cities and explicable
72 Space. the City and Urban Sociology
Castells did not, of course, deny that processes other than consump-
tion also take place lwithin an urban context, for all levels and
elements of the CMP are present within the urban system. Rather, he
argued that the only process which was specific to the urban system
was that of consumption (reproduction of labour-power), and that
'urban' problems , 'urban' crises, 'urban' politics and so on were thus
to be understood through an analysis of consumption processes
within the CMP.
Of all the criticisms which have been made against Castells's for-
mulation, two are pertinent for our present concerns . The first is that
the identification of consumption with the spatial unit of the city is
arbitrary and unconvincing since processes of consumption, no less
than processes of production, of politics or of ideology, are not
territorially bounded. As Mingione observes 'The consumption
process itself is not definable in a purely territorial context, it does not
correspond to any "urban question " but is rather an important part
of the general social question'."
The second is that any attempt to theorise the city in terms of its
functions within the CMP cannot ignore the question of production,
for as Lefebvre, Harvey and others have argued, capital investment in
the 'urban system' may represent a crucial 'circuit' within overall
investment flows, while the physical infrastructure of the city may
provide an effective subsidy to the production costs of private
capital."
What these two lines of criticism indicate is that, like human
ecology, the Marxist urban sociology of the 1970swas effectivelyset
upon two irreconcilable stools. On the one hand Castells, like Park,
was concerned with a particular process in society (for Castells,
consumption/reproduction; for Park, competition/adaptation). On
the other, he was also, again like Park, concerned to retain an
empirical focus on the city even though this necessarily entailed much
wider processes (for Castells, production as well as consumption; for
Park, the cultural as well as the biotic).
Just as human ecology eventually split between those who retained
the theoretical focus but abandoned the city as their specificobject of
enquiry and those who retained the city but lost the theoretical
rationale for studying it, so too does Marxist urban sociology seem
destined to split in much the same way. On the one hand, we are
76 Space, the City and Urban Sociology
For Marx, this was because the antithesis between town and coun-
try was at this time, and at this time only, the phenomenal expression
of the deeper and more basic class division between the rising indus-
trial bourgeoisie and the declining landed feudal nobility. The town-
country division was thus the manifest form of the underlying con-
tradiction between two competing modes of production; a capitalist
mode which became established in the new industrial towns away
from the feudal restrictions of rural serfdom and the medieval guilds,
and a declining feudal mode based mainly on agriculture.
Unlike the period of Antiquity, when cities such as Rome were
simply administrative centres for a predominantly agricultural
society based on a slave mode of production, and unlike the contem-
porary capitalist period, where capitalist social relations have
pervaded both town and countryside alike, Marx therefore saw the
city in the late feudal period as an 'historical subject' for the first and
only time."
Weber likewise focused his attention on Western cities in the
medieval period. His theoretical concern was with different types of
human association and particularly with the origins of calculative
rationality in social relationships. He argued that in the ancient
world , the basis of human association was the family or the clan
irrespective of whether the place of residence of particular family
units was in town or countryside. In the medieval period, however,
the traditionalism of clan or family authority was eroded by the
growth of city guilds and corporations in which individuals were
constituted as citizens with rights and duties legally defined through
membership of the city as the unit of legal and political organisa-
tions .>'
The occidental city in the medieval period was thus the basic unit of
social organisation legally, politically and economically, and the
break which it represented with traditional bonds represented the
necessary condition of the later development of economic rationality
and political democracy in the West. Today, however, the city as the
basic unit of social organisation has been replaced by the nation-
state, for the city is no longer the locus of autonomous markets,
autonomous law, autonomous military organisation and so on . Like
Marx, therefore, Weber suggests that it is only in the medieval period
that the city constituted a sociologically and historically significant
entity.
Durkheim's analysis is similar . His concern is with the changing
source ofsocial cohesion over time consequent upon the development
78 Space, the City and Urban Sociology
Like Weber, then, Durkheim sees the basis of social organisation and
solidarity today as nothing less than the nation-state (indeed, at one
point he likens the nation-state to 'one big city'). The implication of
this is that, in advanced capitalist societies, an attempt to develop a
sociology of the city will inevitably broaden out into a sociology of
the society as a whole.
Two main conclusions follow from all this. The first is that we need to
modify Giddens's assertion that a concern with the city lies at the very
core of social theory . The second is that any explanation of contem-
porary social phenomena must nevertheless remain alert to the possi-
ble implications of spatial organisation as a factor inhibiting or
facilitating the development of certain processes in society as a whole.
Summarising these two themes, we may say that the first points to the
development of a 'non-spatial' sociology while the second points to
the inadequacy of an 'aspatial' sociology.
By a 'non-spatial' sociology I mean that there can be no social
theory of the city in contemporary industrial capitalist societies. This
is the clear lesson to be drawn both from the work of the 'classical'
social theorists and from the failure of all attempts over the past
sevety years to develop a specifically urban social theory. Giddens's
80 Space , the City and Urban Sociology
It seems clear from this that his earlier emphasis on the crucial
theoretical significance of the city can only refer to pre-capitalist (or
what he terms 'class-divided' as opposed to 'class') societies, for a
social form which has been 'eradicated' under capitalism can hardly
remain central to its analysis.
Giddens's argument concern ing the role of the city in pre-capitalist
societies is in many ways reminiscent of that of Weber. For Weber,
the city was an autonomous political, legal and military unit, while
for Giddens it was a 'crucible of power' or a 'power-container'
through which the absolutist state exerted social domination by
virtue of the concentration of authoritative resources (that is, infor-
mation which was necessary for surveillance). And just as Weber saw
that in the modern period , the city has been replaced by the nation-
state as the territorial unit which claims a monopoly over the
legitimate use of force, so too Giddens recognises that, with the
advent of capitalism from the late eighteenth century onwards, the
city gave way to the nation-state as the crucial 'power-container' in
society:
which lies at the heart of this argument, Urry's basic theme - that
certain capacities for action will only be released or triggered off in
certain temporal or spatial contexts - is an important one. The point
may usefully be illustrated with reference to the riots which occurred
in many British cities in the summer of 1981.
These riots, which involved both white and black youth but which
tended to break out mainly in areas of high black concentration (for
example, Brixton and Southall in London or Toxteth in Liverpool)
can probably only be explained in terms of a variety of sociological
factors.v Unemployment rates among young people - especially
young West Indians - certainly constituted one factor, but we would
also need to take into account police practices over a number of years
which were (or were experienced by many as) racist, provocative
political activities by Fascist and white supremacist movements such
as the National Front, mobilisation by left fringe parties such as the
SWP, and so on.
Nor, of course, could any analysis of these events afford to ignore
the fact that most riots occurred in inner city areas (and, for that
matter, that they occurred during the summer time). However, it
would clearly be a mistake to try to find the causes of the riots in either
the spatial (in this case, high-density decaying inner-city areas) or the
temporal (long, warm summer evenings) dimensions, for their ex-
planation lies in the development of certain tendencies in British
society as a whole. As we saw earlier in the discussion of Rex's work
on 'housing classes', conflicts which take place in cities are generated
by factors which are not specifically 'urban' in origin. Racism in
British society, for example, is a product ofa long history of imperial-
ist expansion, government action and inaction, scientific and pseudo-
scientific speculation, and so on, none of which is 'urban'. Similarly,
unemployment is the product of an international recession, dramatic
restructuring in British industry, government monetarist policies,
etc., and again, none of these are 'urban' factors . The spatial dimen-
sion of the riots cannot be overlooked, but it was clearly only a
secondary factor acting to facilitate them in some areas (for example,
where disadvantaged and frustrated sections of the population were
concentrated and were able to come together as a mass, albeit disor-
ganised, force on the streets) and to inhibit them in others (for
example, areas where such people were dispersed).
It also follows from this argument that interventions by govern-
ments or other agencies which seek to bring about social change
through specifically 'urban' policies are fundamentally flawed, for if
Peter Saunders 83
We have shown that the very strong spatial bias in much recent
thinking about the deprived may have been overdone . The
evidence relating to employment, for example, tends to show that
. .. for any given type of workers, the level of unemployment in
inner areas is not higher than in other parts of the conurbation. The
evidence on poverty does not suggest that it is highly spatially
concentrated but instead strongly focused on the bId and spread
throughout many parts of the conurbations . . . All of this might
push us in the direction of solutions which attack deprivation in
terms of support for particular types of client group wherever they
are located."
Whether we also agree with Engels that the only solution to such
problems liesin the transcendence of capitalism is a matter for debate,
but where we must surely concur is in his diagnosis that 'urban'
problems can only have 'societal' solutions.
84 Space. the City and Urban Sociology
This brings us on to the third point. We have seen that although the
spatial unit of the city is not a sociological unit in capitalist industrial
societies, spatial factors may nevertheless inhibit or facilitate the
development of processes originating in society as a whole. It follows
from this that governments may be able to adopt spatially-specific
programmes which, while not solving the problems to which they are
addressed, nevertheless influence where those problems manifest
themselves. British regional policy since the war provides one exam-
ple, for the battery of regional incentives and penalties which have
been adopted by central government and local authorities has not
solved the problem of unemployment and under-investment but has
had some effect in shifting it around from one area to another.
Similarly, the introduction of urban 'enterprise zones', development
corporations, 'freeports' and so on may well have some effect in
stimulating private enterprise in the designated areas , but this is likely
to be at the expense of adjacent areas which cannot offer the same
advantages to existing businesses. Spatially-specific policies and
programmes may therefore have some effect in determining where
problems arise, but they have had precious little effect in preventing
such problems from occurring or recurring elsewhere. Such policies
may be justified on grounds of spatial equity , but they cannot be
justified on grounds of social effectivity.
We saw in the first section of this chapter that although all four
attempts this century to establish a theoretical foundation for a
sociology of the city have failed, each attempt has left a fruitful legacy
in the form of particular questions and issues which remain pertinent
once they are divorced from a concern with territorial arrangements.
From human ecology we have been left a theoretical concern with
the problem of how human groups manage to adapt to their physical
environment. This is an important and legitimate question for
sociologists to pose, and in the post-war years it has been taken up
and incorporated within mainstream functionalist sociology. Talcott
Parsons, for example, has identified adaptation as one of the four
'functional prerequisities' of social organisation."
From the 'cultural' tradition of writers such as Simmel and Wirth
we have been left two distinctive concerns . The first is the question of
the sociological significance of number. The second is a concern with
Peter Saunders 85
Notes
centre and periphery, the national and the local, was already present in
Pahl's earlier work, however - notably in Pahl (1968) - and in other
influential urban sociological writings during the 1960s such as Stacey
(1969). It is a theme which has resurfaced in academic debates in the
1980s by virtue of the clashes between central government and radical
local authorities.
38. Saunders (1982).
39. In addition to the above, seeSaunders (1981), ch, 8; Saunders (1983); and
Cawson and Saunders (1983).
6
The Spatiality of Social Life: Towards
a Transformative Retheorisation .
EDWARD W. SOJA
tion, each of which is used and incorporated into the social construc-
tion of spatiality but cannot be conceptualised as its equivalent.
Within certain limits (which are frequently forgotten), physical
nature and human cognition can be theorised independently with
regard to their particular spatial dimensions or attributes. The clas-
sical debates over 'absolute' and 'relative' qualities of physical space
exemplify the former, while such questions as whether we construct
mental maps which are geometrically Euclidean, Minkowskian,
Riemannian, or otherwise, illustrate the latter. This possibility of
independent theorisation, however, does not mean rigid separation
or unassailable autonomy, for the three spaces interconnect and
overlap . Furthermore, not only are the spaces of nature and cognition
used and incorporated into the social production of spatiality, they
are significantly transformed in the process. This social incorporation
and transformation sets the primary limitation upon the independent
theorisation of physical and mental space, for each is to a significant
degree socially produced and must therefore be theorised and under-
stood as part of the spatiality of social life.
Conversely, the social production of spatiality cannot be complete-
ly separated from physical and cognitive space. Natural and
biological processes affect society, and social life is never entirely free
from the physical friction of distance . The impress of this 'first
nature', however, is not naively and independently given. It is always
socially mediated and recast as part of the 'second nature' that arises
from the organised and cumulative application of human labour and
human knowledge. The space of physical nature is thus appropriated
in the social production of spatiality - it is literally made social.'
Two major conclusions can be drawn from these observations.
First, the social production of spatiality encompasses and incor-
porates the social production of nature, the transformation of
physical space in the creation of a second nature which concretely
manifests its essentially social origins. Following from this, it can be
argued that the results of an independent theorisation of physical
space cannot be directly applied to a materialist interpretation of
spatiality. To do so would obscure the social origins of spatiality and
its socialtransformability in the causal logic of unalloyed naturalism
and physical mechanics.
A parallel argument can be made with respect to cognitive space.
The 'presentation' of concrete spatiality is always wrapped in the
complex and diverse 'representations' of human perception and cog-
nition, without any necessity of direct and determined correspon-
94 The Spatiality of Social Life
This does not just mean that social theory must be historically and
geographically specific. More importantly, social theory must be
about the time-space constitution of social structure right from the
start . 5
The direct producers are freed from the soil only to become trapped
in a grid - one that includes not only the modern factory, but also
the modern family, school, army, the prison system, the city and
national territory.!?
Men make their own history, but they do not make it just as they
please; they do not make it under circumstances chosen by them-
selves, but under circumstances directly encountered, given and
transmitted from the past .'?
The importation of the term 'ecology' into the social sciences has
done little to help matters, since this tends both to encourage the
confusion of the spatial with other characteristics of the physical
world that might influence social life, and to reinforce the tendency
to treat spatial characteristics as in the 'environment' of social
activity, rather than as integral to its occurrenceP
For its part, Marxist research has up to now ... considered that
transformations of space and time essentially concern ways of
thinking: it assigns a marginal role to such changes on the grounds
that they belong to the ideological-cultural domain - to the man-
ner in which societies or classes represent space and time. In reality,
however, transformations of the spatio-temporal matrices refer to
the materiality of the social division of labour, of the structure of
the State, and of the practices and techniques of capitalist econ-
106 The Spatiality of Social Life
omic, political and ideological power; they are the real substratum
of mythical, religious, philosophical or 'experiential' representa-
tions of space-time."
Capitalism has found itself able to attenuate (if not resolve) its
internal contradictions for a century, and consequently, in the
hundred years since the writing of Capital, it has succeeded in
achieving 'growth'. We cannot calculate at what price, but we
know the means: by occupy ing space, by producing a space r"
Notes
1. I refer here, in particular, to current work on the theory of the state, the
spatiality of capitalist crisis, and geographically uneven development.
For example, see Carney, Hudson, and Lewis (1980)especially the essay
by Lipietz; Harvey (1982); and Soja (1981).
2. I have deliberately chosen the term spatiality to refer specifically to
socially-produced space, the created forms and relations of a broadly-
defined human geography . All space is not socially produced, but all
spatiality (as it is used here) is. It should be noted that this definition
allows for arguments and analysis which confer causal influence to
particular spatial configurations and spatial relations (as in the interrela-
tion of core and peripheral regions or the effect of a specific built
environment, behaviourally-defined territories, or particular locational
patterns). This spatial contingency of social life, however, arises not from
124 The Spatiality of Social Life
LP
M - C ... P ... C' - M + 11m, etc.
{
MP
from buying cheap and selling dear, but in so doing their gain is
another's loss. Redistributions of social power through unequal
exchange may be important to the rise and subsequent reorgan-
isations of capitalism (for example, the initial concentration of
wealth through merchant trading and the subsequent centralisa-
tion of capital in giant corporations). But redistribution is no
adequate basis for the continuous circulation of capital. A heal-
thy capitalist economy is one in which all capitalists earn positive
profits. And this requires that real value be added in production.
Living labour (as opposed to the 'dead labour' embodied and
paid for in other commodities) is, then, the exclusive source of
real value added in production.
3. Profit has its origin in the exploitation ofliving labour in produc-
tion. ' Exploitation' can here be stripped of its more emotive
connotations. It denotes a moral condition in which living labour
is treated as a reified 'factor' of production and a technical
condition in which it is possible for labour to create more in
production than it gets through the exchange of its labour power
as a commodity. It does not follow that the labourer receives as
little as possible. Situations arise in which the labourer gets more
at the same time as the gap between what labour gets and what it
creates in production also increases. Put another way, a rising
material standard of living for the labourer is not necessarily
incompatible with a rising rate of exploitation.
4. The circulation of capital, it follows, is predicated on a class
relation. 'Class' is also a loaded term. But I can here give it a
restricted and very simple meaning. The circulation of capital
entails the buying and selling oflabour power as a commodity.
The separation between buyers and sellers opens up a class
relation between them. Those who buy rights to labour power in
order to gain a profit (capitalists) and those who sell rights to
labour power in order to live (labourers) exist on opposite sides of
this buyer-seller divide. The division of class roles that this
implies is not exhaustive of all possible or even important class
relations under capitalism. Nor is the buying and selling of
labour power exclusively restricted to the domain of capital
circulation. But without the capital-labour relation expressed
through the buying and selling oflabour power, there could be no
exploitation, no profit and no circulation of capital. Since all the
latter are fundamental to commodity production and social re-
production, so the class relation between capital and labour is
David Harvey 131
Either capital or labour power are devalued, but not both. To the
degree that the dominant power relation favours capital, so the most
likely persistent condition will be that of capital shortage and labour
power surpluses with all the attendant social devastation that at-
taches to the devaluation of labour power.
The condition that interests me most, however, is that in which
unemployed surpluses of capital and labour power exist side by side.
This is the condition ofcrises into which capitalism is periodically and
inevitably plunged because its technological dynamic undermines its
capacity to sustain growth. Both capital and labour power are then
devalued. Is there no way to avoid such an unmitigated social,
economic and perhaps even political disaster? To pose that question
is in effect to ask : are there ways to absorb the surpluses productively
through opening up new conduits and paths for the circulation of
capital? In what follows I shall argue that spatial and temporal
displacements offer ample opportunities to absorb the surpluses with,
however, dramatic consequences for the dynamics ofaccumulation. I
shall then go on to show that neither stratagem offers a permanent
resolution to the inner contradictions of capitalism but that resort to
either (or both) fundamentally alters the way in which crises are
expressed.
evitably slow and this itself can check the persistence of any spiral
form of development.
The reallocation of surplus capital likewise poses problems. The
surplus can exist as money, commodities or productive capacity. Ifit
exists as particular use values (shoes and shirts) or productive
capacity (lathes and lasts) it cannot be converted directly into a
railroad or a new educational service. It must first be converted into
money . This presents the first barrier to be overcome because overac-
cumulation defines a state in which the smooth conversion of capital
from one form to another, and into money in particular, has become
impossible. Credit can surmount this barrier but it does not do so
unambiguously. Why would that credit not be used to create even
more surpluses of the same sort? Why would it be attracted into long-
term investment when every capitalist is deeply conscious of the need
to conform to the dictates of socially-necessary turnover time? Here is
the second barrier to be overcome. The answer to it is as old as
capitalism itself. It lies in the creation of 'fictitious capital' - bonds ,
mortgages, stocks and shares, government debt, etc." What fictitious
capital does is to convert a long-drawn-out circulation process (the
capital embedded in a railroad, for example) into an annualised rate
of return. It does this by facilitating the daily buying and selling of
rights and claims to a share in the product offuture labour. The rate is
sometimes fixed (bonds) or variable accord ing to what labour actual-
ly produces year by year (shares). But it is measured in terms exactly
comparable to the rate of profit over socially-necessary turnover time
in current production.
Through the use of credit and fictitious capital, surplus capital can
flow from one sphere to another. When, for example, the annualised
yield on some fictitious capital (railroad shares, government debt,
etc.) exceeds the rate of profit in current production, then there is an
incentive for capital to switch from present to future uses. The switch
is unlikely to be smooth, however, because of the 'lumpiness' of many
of the investments involved (railroads, hospitals, etc.) and the dif-
ferent working periods required to make a project operational. Fur-
thermore, the nature of many of the investments - public uses and the
difficulty ofcharging directly for their use - often precludes action on
the part of individual capitalists, so that new special organisational
forms (joint stock companies, state and quasi-public enterprises)
have to be created if railroads, ports, universities, scientific and
educational centres, etc. are to be created. Capital markets must also
be well organised if clear market signals as to differentials in the
138 The Geopolitics of Capitalism
7.1). At some point the debts must be paid. Exactly when depends
upon the turnover time of the capital deployed in particular physical
and social infrastructures. But accelerating fictitious capital forma -
tion - the true heart of the spiral of development - means that more
and more living labour in current production has to be given over to
working off past obligations.
45
40
'oE" 3 5
o
s
eQ.3 0
'"
~25
*~ 20
C
'~15
"
'"0.
'"
~ 10
'"
C
~ 5
Two possibilities for crisis formation then arise. Under the first,
overaccumulated capital stored in physical and social infrastructures
is realised through an active growth in current production (facilitated
in part through improved infrastructures). But then overaccumulated
capital flows back from storage to combine with excess capital in
current production to create ever-greater pools of surplus capita1.
The capacity for further fictitious capital formation is blocked either
by labour or resource constraints or by the circulation of capital in the
existing infrastructures which cannot be disturbed before their life is
out without devaluation. A general crisis ensues with surpluses every-
where subject to devaluation. Under the second, the capital stored in
physical and social infrastructures is not realised and is devalued . The
crisis now appears to be provoked through the lag of productivity
(and perhaps through shortages of both capital and labour) in current
140 The Geopolitics of Capitalism
velopment under capitalism that can, among other things, explain the
significance and evolution of state functions (local, regional,
national, and supra-national), uneven geographical development,
interregional inequalities, imperialism , the progress and forms of
urbanisation and the like. Only in this way can we understand how
territorial configurations and class alliances are shaped and reshaped,
how territories lose or gain in economic, political, and military power,
what the external limits are to internal state autonomy (including the
trasition to socialism), or how state power, once constituted, can itself
become a barrier to the unencumbered accumulation of capital, or a
strategic centre from which class struggle or inter-imperialist strug-
gles can be waged.
The historical geography of capitalism has to be the object of our
theorising, historico-geographical materialism the method of en-
quiry. Bravely said, but hard to do. To begin with, we encounter an
incredible variety of physical and biotic environments across the
surface of the earth, many of which have been substantially modified
by centuries of human action. The diversity of that action has
produced a variegated geographical landscape in which cultural and
socio-structural differentiations have taken deep root. Such par-
ticularistic geographical differentiation may be encompassed but by
no means totally crushed under the homogenising heel of the circula-
tion ofcapital. Viewed abstractly, space also possesses more complex
and particularistic properties than time. It is possible to reverse field
and move in many different directions in space whereas time simply
passes and is irreversible. The metric for space is also less easily
standardised. Time or cost of movement over space do not necessarily
match each other and both yield different metrics to simple physical
distance . Compared with this, the chronometer and the calendar are
wondrously simple. Geographical space is always the realm of the
concrete and the particular. Is it possible to construct a theory of the
concrete and the particular in the context of the universal and abs-
tract determinations of Marx's theory of capitalist accumulation?
This is the fundamental question to be resolved.
Marx was not necessarily wrong to prioritise time over space. The aim
and objective of those engaged in the circulation of capital must be,
after all, to command surplus labour time and convert it into profit
David Harvey 145
The labourers, for their part, face a similar dilemma. If they cannot
escape from the wages system altogether they will presumably move
in order to improve their real wages, work conditions, etc. The irony
here is that the capitalist development process relies on exactly such
behaviour to co-ordinate the demand and supply oflabour power in
space. But on the other hand, workers can also improve their lot if
they stay in place, collectively organise , and fight for a better life. To
this end they may build their own social and physical infrastructures
(or co-opt those promoted by the bourgeoisie), struggle to control the
state apparatus and thereby enhance their power to improve their
lives. And to the extent they succeed, they too may support measures
that constrain the free geographical mobility of labour power (im-
migration in particular). The tension between free geographical
mobility and organised reproduction processes within a confined
territory exists for both capitalists and labourers alike. And how that
tension is resolved for either depends crucially on the state of class
struggle between them . Capital flight (and the consequent undermin-
ing of territorial coherence and state power) is as typical a response to
working-class victories within a territory as is individual worker
mobility to escape the more vicious forms of capitalist exploitation.
Again, I will lay these tensions aside for the moment in order to get to
my immediate point: none of this can occur oblivious of the immobile
social and physical infrastructures necessary to ensure the reproduc-
tion of labour power of a certain quantity and quality.
We can now draw a fundamental conclusion . The ability of both
capital and labour power to move at short order and low cost from
place to place depends upon the creation of fixed, secure, and largely
immobile social and physical infrastructures. The ability to overcome
space is predicated on the production of space. But the required
infrastructures absorb capital and labour power in their production
and maintenance (see Section III) . We here approach the heart of the
paradox. A portion of the total capital and labour power has to be
immobilised in space, frozen in place, in order to facilitate greater
liberty of movement for the remainder. But the argument now can be
brought full circle because the viability of the capital and labour
committed to the production and maintenance of such infrastruc-
tures can be assured only if the remaining capital circulates down
spatial paths and over a time-span consistent with the geographical
pattern and duration of such commitments. If this condition is not
met - for example, if insufficient traffic; is generated to make the
railroad profitable, or expansion of production does not follow on
massive investment in education - then the capital and labour com-
150 The Geopolitics of Capitalism
through the territory under their control. The struggle for commun-
ity, regional or national solidarity as the ideology behind the alliance,
may support, reconstitute or in some cases (as I believe can be shown
for the United States) actively create local and regional cultures and
traditions. The conclusion is inescapable: if regional structures and
class alliances did not already exist, then the processes at work under
capitalism would necessarily create them.
I advance this proposition independently of any appeal to the
concept of the state . I do so because I want to stress that the drive
towards state formation and dissolution under capitalism has to be
understood in the context of forces making for the formation and
dissolution of regional class alliances. The state is different, however,
from other agents in a variety of respects. First, territory and the
integrity of territory is the objective of its personnel to a degree
uncharacteristic of other agents. Second, by virtue of its authority, it
can givefirmer shape and cohesion to regional class alliances through
the institutions of law, governance, political participation and
negotiation, repression and military might. Third, it can impose
relatively firm boundaries on otherwise porous and unstable geogra-
phical edges. Finally, by virtue of its powers to tax and to control
fiscal and monetary policy, it can actively promote and sustain that
structured regional coherence to production and consumption to
which capitalism in any case tends and undertake infrastructural
investments that individual capitalists could not tackle. It can also
become a central agent for the promotion of nationalist ideology. For
all these reasons the state becomes the key to the expression of the
tendency to form regional class alliances and adds its own specific
rationale to this fundamental underlying process.
The upshot is a regional class alliance that typically builds upon the
apparatus of state power, engages in community boosterism and
strives for community or national solidarity as the means to promote
and defend an amalgam of various class and factional interests within
a territory. Spatial competition between localities, cities, regions and
nations takes on new meaning as each regional alliance seeks to
capture and contain benefits in competition with others . Global
processes of class struggle appear to dissolve before our eyes into a
variety of interterritorial conflicts. Lenin is vindicated.
But the stability of any regional class alliance is undermined by
exactly the processes that Marx described so well. Accumulation and
overaccumulation, class struggle and technological change disrupt
and transform regional alliances in much the same way that they
David Harvey 153
affect all fixed spatial configurations (see Section V). Even the most
solid partners in a regional alliance may be tempted to move even at
the best of times and at the worst of times individual behaviour
becomes very unpredictable. Competition forces all economic agents
to be on the alert for the main chance to make a geographical move
that gives them an advantage over their rivals. The instability in part
results simply because individuals do not have the luxury of knowing
exactly what their rivals will do. Similar problems arise in the realms
of class struggle. While capital and labour may move into an alliance
on some issues (barriers to cheap imports, for example) and com-
promise on others (collective bargaining procedures, for example) the
antagonism between them can never totally disappear. And when
class struggle sharpens, the alliance becomes more and more fragile.
Factions of capital may be tempted to flee the region altogether or
strike back at the power of labour by threats to move or threats to
open the floodgates to cheap imports or low-wage immigrant labour.
Such threats can antagonise other factions of capital who cannot so
easily escape local commitments. Financiers, producers, merchants,
landlords, etc. do not necessarily see eye to eye. And labour, which
once adopted conciliatory policies to consolidate its position within
the class alliance, may be tempted to resuscitate more revolutionary
demands. The conditions for breakdown and disintegration of the
regional alliance are ever present. The dynamic of capitalism in the
end tends to break apart the very alliances it initially promotes. The
stresses become particularly fierce under conditions of crisis. It then
seems that the only way to keep the alliance intact is to seek an
external resolution to the region's problems.
tion, the search for a spatial fixconverts the threat of devaluation into
a struggle between unstable regional alliances over who is to bear the
brunt of the crisis. Faced with the prospect of splintering into a
thousand warring pieces, a regional alliance may consolidate itself
and turn its destructive tendencies towards the outside. The export of
unemployment, inflation and idle productive capacity become the
stakes in an ugly game. Trade wars, dumping, tariffs and quotas,
restrictions on capital flow and foreign exchange, interest-rate wars,
immigration policies, colonial conquest, the subjugation and domin-
ation of tributary economies, the forced reorganisation of the
territorial division of labour within economic (even corporate)
empires and , finally, the physical destruction and forced devaluation
achieved through military confrontation and war, can all be caught
up as part and parcel of the processes of crisis formation and resolu-
tion . The search for a 'spatial fix' takes a viciously competitive and
perhaps even violent turn.
The theoretical argument I have here set out is, I hold, as fundamental
to the elucidation of our present plight as it is to the interpretation of
the historical geography ofcapitalism. If! am correct, and I hasten to
add that I hope I am grossly in error and that history or others will
quickly prove me so to be, then the perpetuation of capitalism in the
twentieth century has been purchased at the cost of the death, havoc
and destruction wreaked in two world wars. But each war has been
waged with ever more sophisticated weapons of destruction. The
bourgeois era has certainly witnessed a growth in destructive force
that more than matches the growth of productive force so essential to
the survival of capitalism. That the latter should also require the use
of that destructive force appears insane . Yet the ideologists ofcapital-
ism shed no tears but, like Schumpeter, sing paeans of praise to what
they term the 'creative destruction' through which capitalists so
dramatically transform the world . But our present plight must surely
give us pause . As temporal and geographical solutions to the inner
dialectic ofoveraccumulation run out, the crisis tendencies of capital-
ism once more run amok, inter-imperialist rivalries sharpen and the
threat of autarky within closed trading empires looms : The struggle
to export devaluation within a disintegrating world order comes to
the fore and belligerence dominates the tone of political discourse.
With this comes the renewed threat of global war, this time waged
with weapons of such immense and insane destructive power that not
even the fittest stand to survive. The message which Marx long ago
jotted down in that notebook that became the Grundrisse impresses
upon us more urgently than ever:
Capitalism did not invent war any more than it invented writing,
knowledge, science or art. Not all wars, even in the contemporary era,
David Harvey 163
can be truly regarded as capitalist wars. And war will not necessarily
disappear from the human scene with the demise of capitalism. But
what our theory strongly urges is that we see the replacement of the
capitalist mode of production, that expansionary and technologically
dynamic process of circulation that we began by examining, as a
necessary condition for human survival. And that is a task beyond the
prerogative of any single class or community. It is, I submit, a task
that should be the immediate focus of every atom of our collective
attention.
Notes
1. Harvey (1982).
2. Harvey (1982) ch.3.
3. Harvey (1982) ch. 6.
4. Harvey (1982) ch.9 and 10.
5. Harvey (1975); (1982).
6. Marx (1973) p. 539.
7. Marx (1973) pp. 533-4; Cohen, 1978.
8. Aydalot (1976).
9. Harvey (1982) ch. 12.
10. Harvey (1978); (1982) ch. 13.
11. Kaldor (1978).
12. Marx (1973) pp . 749-50.
8
Class, Division of Labour and
Employment in Space
RICHARD A.WALKER
Class is arguably the single most crucial axis on which human life
turns in the modem world, yet is at the same time one of the most
difficult of social facts to grasp . Marx provided a powerful concep-
tion, rooted in the mode of production, with which to understand the
class character of capitalist societies. This theory remains subject to
doubt, however, because of persistent failure of vision in the face of a
social reality that does not conform to tidy conceptual systems. Space
is another fundamental dimension of human life, yet the geographic
element in the social sciences has atrophied for want ofa way in which
spatial relations might comfortably be integrated into social theory.
Fortunately, the recent revival of philosophical discourse in the social
sciences, which seeks to conjoin 'theoretical realism' with so-called
'structuration theory', has rekindled the spirit ofinquiry once known
as dialectical materialism, while contributing much in the way of
clarity and systematic development of basic tenets. This mode of
thought offers a way out of the dead-ends to which class analysis and
geography have come, although the refined tools of philosophy need
to be fitted into the powerful machine of Marx's theory of capitalism
in order for the analytic workto proceed.
Class analysis has been persistently stymied by four problems in
social theory which can be fruitfully addressed by a dialectical
materialist, or Marxist 'structurationist' approach. The first is that
posed as the question of structure and agency, which is addressed in
section I.
The second problem confronting class analysis is the stratification
of the social world . Social systems consist of several layers of nested
Richard A. Walker 165
I Thestructuring of class
The 'problem' of class begins with the inability to see classes revealed
in a self-evident way through empirical scrutiny of the everyday social
world . There have been various attempts by Marxist theorists to
wrestle with the unhappy correspondence of class theory to apparent
reality, and I shall indicate these in summary fashion.
The first is to treat class as a strictly empirical category. Classes
then appear as boxes into which individuals may be sorted on the
basis of their measurable characteristics. This empiricist fallacy,
which goes back at least to Kautsky, usually leads to reading non-
conforming groups out of the class structure ofcapitalism altogether,
herding them into sizeable middle classes (new petty bourgeoisie, new
class, etc) which come to ingest so much of the populace as to render
the bourgeoisie and proletariat mere shadows of their Marxian
selves.'
One answer to this dilemma is to push all the problems of noncon-
formity between theory and reality into the realm of consciousness.
That is, objectively the proletariat includes the vast mass of people,
but they do not all know it or act on it. This is Lukacs' dualism of
class-in-itself and class-for-itself (conscious of itself). 2 E. P. Thomp-
son 's influential break with tradition was to make a virtue of adver-
sity, and to focus on the everyday, subjective experience of people as
the heart of the process of class formation, rather than an afterth-
ought." History, ambiguity and struggle were thus reinserted into
what had become an academic debate. Life was breathed into the
history of classes and class struggle. But Thompson's grasp on the
dialectics of agency and structure has subsequently been shown to be
slippery , and the ties between the subjective , creative aspects of
experience and their objective base in economic relations frequently
becomes so attenuated in this approach as to bear little resemblance
to Marx's analysis ." Agency overwhelms structure and the problem
of empiricism remains. History stays a great flood with many eddies
but no central currents. Przeworski's effort to grasp the dialectics of
structure and agency for class analysis has admirably recaptured the
spirit of Thompson, but remains subject to the same limitations: the
economic relations virtually disappear from the discussion. 5
Recently there have been several attempts to restore the 'objective'
character of Marxist class analysis while recognising certain complex-
ities of correspondence between class categories and empirical posi-
tions. Poulantzas introduces structural theory into the study of class,
Richard A. Walker 167
the less, lest all this seem a return to relativism and historicism, capital
does still accumulate, the working class rears its head with the spread
of capitalist relations of production around the world, and the social
scientist can, after all, slice through the uncertainty to reveal some-
thing of the shape of underlying mechanisms at work: although
capturing the essentials of social history is perhaps harder work and
less precise business than generations ofoptimistic scientific socialists
and positivist social scientists have been willing to acknowledge .
It is not surprising, then, that people have found class to be an
elusive phenomenon. But we can understand something of both the
dilemma and the opportunity opened up by a structuration approach
once we begin to enrich class analysis through the inclusion of the
division of labour, employment and location of workplaces as inter-
secting 'levels' of capitalism.
was to spiral up and out from a 'bare bones' definition of class, not
just adding elements to the analysis, but recasting the initial
categories as they are seen from new angles. In other words , Marx's
analysis of the dimensions of class is part of his entire study of the
dimensions of capital.P (The 'secret' of the mysterious unfinished
chapter of Capital on class is that there are unlikely to be any surprises
there to anyone who had read the rest of the book.)
As Marx shows, once industrial capital stood up on its own feet, it
began to revolutionise economy and society in ways that soon left the
issue of enclosures or the hand-loom weavers, that is, the mere fact of
wage-labour, far behind. New areas of economic importance such as
the engineering and sales functions have been opened up which have
become the cutting-edge of class change and as much a part of the
'essence' of class as the formal condition of working for wages, which
has lost some of its force in class formation today. This does not in
any way reduce the historical significance of classic proletarianisa-
tion, but wage-work now establishes capitalist class relations only in
the weak sense. Class is a relation ofpower which must continually be
maintained, extended and recreated in the face of changing condi -
tions. It is therefore inextricably entwined with the development of
capital(ism) as a system of production, circulation and exploitation.
Carchedi, despite other problems with his analysis, is thus right to
insist on lodging the definition of class within a framework of the
evolution of capitalism.s?
In short, not only does the course of class struggle, or the everyday
clash of people in a class-structured context, 'determine' the strength
of the class relation and the clarity of class formation, but that
struggle must be placed on the shifting ground of an unfolding
economy, or system of production as a whole. The latter cannot be
adequately treated, nor the problem of class formation tackled,
however, without delving into another fundamental level of political
.economy - the division of labour.
between the detail and the social division oflabour, space enters into
the discussion as a matter of course. The spatial division between
discrete workplaces is usually the most practical basis for distinguish-
ing between the two. It does not follow, however, that division of
labour is merely an empirical concept, to be contrasted with the
'structural' force of class." One has to extract the underlying struc-
ture of work organisation and production systems from the caco-
phonous reality of corporations, factories, departments and jobs -
and the spatial division of labour.
Both the detail and social division of labour are cross-cut by a
vertical, or hierarchical, division of labour. This introduces further
confusion, which is compounded by the intersection with such con-
cepts as management and corporate hierarchy. Within the detail
division oflabour, for instance, one finds various levels of managers,
or directors, of the immediate labour process. It is not satisfactory to
treat such distinctions only in terms of 'authority structures', or
systems of class control, separate from the detail division oflabour, as
Giddens does." The organisation of all large labour processes re-
quires direction, independent of the class character of society. It also
requires various kinds of labour skills, with more or less command
over the techniques of production and different levels ofsocial status.
The confusion arises because under capitalism administration and
command of techniques readily become the province of capital,
infused with a heavy dose of class relations." The vertical element in
the social division oflabour, not only within firms but between what
appear to be higher and lower functions of capital, is even tougher to
handle. Regardless of capitalist class relations, command over large
organisations, research and marketing activities, or the circulation of
money implies very different levels of command over social produc-
tion as a whole. This hierarchical differentiation is compounded by
distinctions of skill and scarcity of certain kinds of labour.
Thus relations fraught with power differentials emerge from the
division oflabour and the practical carrying out of social production.
That is, the division of labour provides a material axis around which
people develop capabilities, knowledge, associations - and power.
These power relations are independent from those of class, in the first
instance, although they may easily become overlapped or even absor-
bed into the nexus of class power. All discussions of class trip over this
tangle of social relations. Critics of Marx have asked repeatedly how
ownership and exploitation can be the only sources and objects of
176 Class, Division of Labour and Employment in Space
power. The answer is they cannot. The problem, then, is to mesh the
division of labour and class without eliminating the distinctive
character of either.
their input and output linkages and labour demands will be dif-
ferent .t?
Second, workers require a certain fixity for their reproduction.
This refers both to fixed capital in the consumption fund, such as
houses, schools and parks, and the time necessary for secure social
relations to develop, in the sense of habituated practices. Without
these, life ordinarily becomes both materially and socially impoveri-
shed. The element of fixity in the formation of working-class com-
munities itself introduces a degree of social divergence among labour
forces. People do not randomly mix across space. These differences
are compounded by diverse roots in pre-capitalist social formations
and the industrial histories of particular communities. 51 The sphere
of consumption and labour personnel reproduction thus cannot be
treated as if it were wholly unrelated to work, as is so often done .52
Finally, the employment relation requires some time and stability
for a workable social order of the workplace to arise. Conversely
employers must recreate the relations of class by repeatedly recon-
quering the workplace over time.
None of these things is absolute, of course. Yet immobility is a
persistent feature ofhuman life and capitalist production. The issue is
actually the relative mobility of capital and labour, of one kind of
production process versus another. The mobility of capital is highly
developed today, thanks to such innovations as electronic money
transfer, telephone communication and the multinational corpora-
tion .P None the less, capital has not freed itself of the necessity of
taking a material form as instruments of production. Similarly,
capital is not normally invested in wholly new plants , but is added to
or drained from existing facilities. There are other barriers to capital
mobility, as well, such as small firms with little investment search
capacity, poorly developed information or financial linkages, and
personal allegiances to place. Industries also vary considerably in
their material fixity, from the vagabond character of construction
and transport activities to the long life of some mines or large steel
complexes.54 Labour migrates, too. While this is often triggered by,
or directly mobilised by, movements of capital," labour migrations
have a dynamic of their own once a stream is established, which may
come to bear little relation to job prospects .
Further, industries differ in the degree of stability needed for a
successfulemployment relation. Some employers, such as automobile
companies or machining firms, desire a stable, low-turnover work-
force on which they can rely for predictable, knowledgeable long-
184 Class, Division of Labour and Employment in Space
I have tried here to advance the idea ofclass structuration a short step
by adding three additional layers of structure - division of labour,
employment, and space - to class analysis. The focus of this effort has
been to portray the employment relation as an open system in which
class relations are actively constituted through the practical activities,
interactions and power struggles of people on the job. A second
purpose has been to argue that given the inherent geographic element
in production, class manoeuvre and the reproduction of capital and
labour, classes are necessarily constituted in and through the use of
space.
This way of looking at things, it should be added, grew out of the
practical exigencies of a research effort that was seeking to compre-
hend the logic ofindustrial location for a group of specificindustries,
with specific differences in technology, labour relations, and
locational patterns. At the same time, the construction of this model
in no way entails rejection of any basic tenets of the Marxist analysis
of capital. Indeed , each level of structuration in a sense grows out of
the previous ones: division of labour out of the capitalist drive to
accumulate, the employment relation out of the class encounter
within the division oflabour, and the spatial division oflabour out of
the employment relations in various industries.
The conclusions that flow from this view of class formation are
likely to be controversial. It appears, at first glance, to introduce a
strong element of class fragmentation along the lines ofthe division of
labour and the regional unevenness of employment. Indeed, it implies
a somewhat weaker concept of class and process of reproduction of
class structure than most Marxist approaches. But it accords much
Richard A. Walker 187
The issue, then, is not whether or not classes are fragmented but
whether the fragments of everyday life provide a basis for weaker or
stronger formation of class-oriented institutions, practices and in-
sights. Do people encounter the capital-labour relation in ways that
emphasise or hide the class dimension of their existence? And do they
encounter it in a manner that makes it possible to build more strongly
class-oriented organisations, activities or politics, and hence to re-
produce class structure in a progressively stronger fashion? Certain
simple and habitual acts, such as going to work for profit-making
firms, create the reality of class structure in the weak sense, but no
more. The making of a coherent working class or a polarised class
society is a cumulative historical process which depends on both
existing material sources of commonality and the creation of further
resources from those at hand. Hence, crucial defeats of efforts to
build up class organisation, class politics or class culture actually
weaken the force of class structuration.s- This approach does not
deny that class polarisation can take place very fast, given the right
circumstances. But it suggests that, in human as in physical geogra-
phy, it is often the slower, less dramatic processes of building up and
tearing down that have the greatest effect in the long run.
Notes
admit my own previous conception of class has been too static. See
Walker and Greenberg (1982).
21. For a critique of functionalism, see the work of Giddens and his foll-
owers, for example, Giddens (1979) and (1981b).
22. Wright (1982).
23. cf. Cohen (1978).
24. Marx (1967) ch. 14.
25. Marglin (1974).
26. Wright (1980) p. 354.
27. Giddens (1981a).
28. Marx (1967) ch. 14 and 15. Braverman (1974).
29. Wright and Perrone (1977).
30. For example, Poulantzas (1975). See discussion in MacKenzie (1982);
Wright (1978) and (1980).
31. Weber (1978).
32. This is a position I once favoured: Walker and Greenberg (1982). MacK-
enzie (1982)and Giddens (1981a) fall into the same trap in trying to use
labour market segmentation to explain class inequality.
33. Wright (1976).
34. Wright (1980).
35. Thompson (1978).
36. Carchedi (1977).
37. Walker (1979).
38. The following discussion is based on Storper and Walker (1983) and
(1984).
39. Marx (1967) ch.24 .
40. Nichols and Beynon (1977).
41. Rosenberg (1976).
42. Edwards (1979).
43. Aronowitz (1978).
44. Giddens (1979).
45. Burawoy (1979).
46. See Storper (1982).
47. In emphasising capital reproduction over labour reproduction, I am
open to all the criticisms made by McDowell (1983).
48. Storper (1982); Storper and Walker (1984).
49. Storper and Walker (1983) and (1984).
50. Walker and Storper (1981).
51. Bleitrach and Chenu (1981); Massey (1978a).
52. For example, Giddens (198Ia).
53. Bluestone and Harrison (1982); Walker and Storper (1981).
54. Walker and Storper (1981).
55. Clark and Gertler (1983); Piore (1979).
56. Storper (1982).
57. Massey (1978a).
58. See for example, Bluestone and Harrison (1982).
59. For example, Gregory (1982b); Walkowitz (1978).
60. See again Sayer (1982).
61. Giddens (1981a) p. 1l0.
62. Davis (1980).
9
Spatial Change, Politics and the
Division of Labour
ALANWARDE
Babel
Although by now many people may be tiring of the lament that social
scientists ignore space and time, it can hardly be denied that there is
no corroborated theory concerning historical trends in spatial patter-
ning. No agreement exists as to the mechanisms which distribute
significant social artefacts across space over time. Nor is there even
consensus about how to describe the spatial patterns. Questions
about the causes and the directions of spatial change are answered
with considerable assurance but little consistency.
Consider, on the one hand, Lojkine's account of the tendencies in
spatial change in contemporary society:
This view, that as time passes so too do local and regional cultural
differences, is inherent in classical sociology's central distinction
between Gemeinschaft and Gesellschaft and enshrined in recent
theories of modernisation, mass society, mass media studies, decline
Alan Warde 193
of community, and many more . Yet it is not clear that this trend is
either so preponderant or so advanced as is often imagined. Rose and
Unwin, for example, analysing regional support for political parties
in capitalist states in 1945 and 1970 - a study parallel in intent and
scale to that of Williamson 's comparative international study of
economies - showed only a marginal tendency for regional differen-
ces to diminish, regional parties survived, and regional voting pat-
terns persisted .8 The evidence for the United Kingdom has been most
closely examined by Hechter in his attempt to explain the phases in
the development of Celtic nationalism.9 He found tendencies towards
a decline in spatial differences in England, but considerable distinc-
tiveness in Wales, Scotland and Ireland. Having standardised for
industrial structure, he discerned a large 'cultural residual', as he
called it, in the Celtic regions which prevented the explanation of
political differences in terms of an economic division oflabour. There
are some difficulties with Hechter's analysis, not least the fact that
neither of the two theories he was testing, a diffusion thesis and the
internal colonialism thesis, remotely accounted for his data. But if
nothing else he casts serious doubt on the existence of trends towards
homogenisation.
The existence of fundamental inconsistencies in the descriptions
and explanations of economic and political trends is incontrovertible.
One source of this chronic confusion derives from the difficulty of
demarcating meaningful and comparable spatial areas . Whether one
detects specialisation or diversification depends to a degree upon
which level of spatial disaggregation one chooses to use. Analyses
variously refer to putative nations, regions, sub-regions, counties,
urban systems, localities and neighbourhoods. Some of these are
meaningful economic units, others are meaningful political units; but
rarely is a single unit amenable to both political and economic
analysis for any length of time. Nor is it necessarily the case that
trends at each level move in the same direction: it has been suggested,
for example, that statistical aggregation makes it appear that regions
are converging, whilst component sub-regions remain as diverse as
ever.!? Unsurprisingly it remains a matter of controversy among
spatial theorists at what levelthe most important spatial effectsoccur.
A second set of problems arises from the understandable fervour of
social scientists who have painstakingly collected time-series data to
detect trends, even if the evidence be ambivalent. In the case of British
economic development, the empirical indicators regularly in use tend
to be inconclusive. Lee's figures for the regional distribution of
194 Spatial Change, Politics and the Division of Labour
Thebasic metaphor
terms of the quality of its labour force and the local balance of class
forces, there arises a problem of defining what might be called class
combination rules. Decisions to locate, expand or cease operations at
a given site will depend upon the combination of characteristics like
skill profiles, historical experiences of class conflict, etc. It is then
necessary to determine the rules of combination between various
class residues in any place, since different combinations will produce
different political effects. At present we know about the political
propensities of some highly distinctive locales: places with a high
proportion of manual workers in mining have different political
cultures from agrarian areas or retirement zones on the south coast.
But the political outcomes of a much wider range oflocal or regional
class formations need to be delineated. In this respect Massey's
formulation is highly indeterminate since she tells us little about what
typical combinations produce what effects.
Third, it is necessary to qualify Massey's assumption that the only
relevant local effects are class effects. The basic metaphor presents a
very truncated view of what comprises a spatial division of labour.
No contemporary account of the division of labour can afford to
ignore the broad range of social relationships which derive from the
process of the (re-)production oflabour power .v' There is significant
disagreement among Marxist accounts on the importance of state
intervention in the process of the reproduction oflabour power, and
on how effects of that intervention should be conceptualised - hence
the debate over collective consumption. But even those who consider
reproduction important take almost no notice of variations among
social institutions of civil society which produce labour power -
households, families, neighbourhoods, etc. Yet there are considerable
temporal and spatial variations in the conditions under which labour
power is produced, and different sets of social arrangements appear
to have pertinent political effects." These effects are germane to the
formation of a local surface.
The recognition of this aspect of local differences prompts further
attention to the question of local specificities. In Massey's metaphor
space is denied any autonomous properties. The critique of orthodox
geography's 'fetishisation of space' is tightly embraced in the meta-
phor. It seems correct to accept that space has no causal powers: as
Sack put it, only social entities of 'substance', which occur in space,
have effects.t" Yet different kinds of spatial structures remain
necessary general conditions of certain kinds of action. Minimally, it
is hardly controversial to say that the dispersal of agricultural work-
200 Spatial Change, Politics and the Division of Labour
because they tap a fresh, unorganised labour force, but because they
can be designed with industry widely dispersed and mingled with
equally dispersed residential quarters for labour. By contrast, the old
industrial cities built in the phase of industralisation are unfit to be
restructured. The rise of the sunbelt cities is thus a response to the
search for more effective, qualitative control by capital over labour.
Gordon's schema is promising . It discriminates between periods in
history , it contains a dual spatial reference (to regional economies
and urban ecologies) and it contains a dynamic mechanism - indus-
trial class struggle. To what extent it could be applied to Britain is a
matter of speculation. There seems to have been some similar correla-
tion between industrial militancy and city-size in nineteenth-century
Britain.r' British industry did move toward suburban locations,
though at a considerably later date than in the USA, though that may
well be because Britain was much slower to transform the labour
process, scientific management techniques not having been in-
troduced until the 1930s.4 5 Similarly, residential segregation in the
British city seems to have occurred at a different stage of industrial
development and not to have changed much over the past hundred
years." Also, some authors have considered that the visibility of the
bourgeoisie in the British towns was conducive to conflict, though
others have attributed the same outcome to the absence of a bour-
geoisie.f?
Regarding theoretical issues some reservations should be regis-
tered . First, Gordon conflates contradictions and struggles: con-
tradiction and conflict are logically and historically separable, con-
tradictions not necessarily leading directly to struggle." Second, the
historicity of the scheme is dubious. Gordon's scheme is rather
unilinear, accounting for uneven development by a very even logic. It
would seem to imply, for instance, a strong tendency towards un-
iform, nationwide spatial differentiation within each phase . Third,
spatial effects are specified with almost exclusive reference to labour
militancy and capital 's search for control, to the neglect of other
processes. Finally, and incidentally, the explanation of the rise of the
sunbelt is not wholly convincing, as it is unclear why just any green-
field would not have been satisfactory, orwhy the sunbelt emerges as
a part of the emergence of the corporate era of accumulation.
Labour markets
Although both Massey's and Gordon's accounts implicitly refer to
labour markets as a result offocussing on the quality oflabour power
204 Spatial Change, Politics and the Division of Labour
Collective consumption
The work of Castells is sufficiently appreciated to obviate detailed
exposition. It contains some clues for the discovery of a more com-
plex metaphor. Castells is primarily concerned with the determina-
tion of social practices within urban systems, which he delimits in a
fashion similar to the British geographers' notion of a metropolitan
labour area, constructed 'on the basis of a map of commutings'.53 The
distinctive characteristic of urban systems in advanced capitalism is
that they are structured in accordance with a logic of collective
consumption. Cities are the loci for the provision of services which
socialise the process of the reproduction of labour power, thus min-
imising the costs to capital of transport, health-care, education and
housing, such necessary services being wholly or partly financed and
organised by the state . It is unclear whether this has been the principal
role of cities throughout the period of the dominance of the capitalist
mode of production - explicitly it was not so in pre-capitalist
modes .54 Castells postulates a qualitative change in the period since
1945. He implies that the requirements of production previously
determined urban form , but that currently the spatial horizon of the
productive enterprise is the region. Sites are sought in an appropriate
region, industry being indifferent to locality." Hence, like other
French Marxists," Castells describes a tendency toward regional
specialisation in economic production. Meanwhile, the urban system
becomes primarily a repository of labour power wherein the state
supervises its effective reproduction.
206 Spatial Change, Politics and the Division of Labour
Castells should not have neglected the role of private and household
consumption when addressing the problem of the effects of social
arrangements for the production of labour power. The spatial im-
plications of variations in women's economic activity are considera-
ble. Though it has reduced markedly since the Second World War,
local and regional variations in the extent to which women, and
especially married women, were engaged in the formal economy were
very great. For example, within Lancashire in 1911 the proportion of
208 Spatial Change, Politics and the Division of Labour
single women economically active varied from 30 per cent (in Widnes)
to 80 per cent (in Rishton); and the proportions of economically
active married women ranged from 2 per cent (in Haydock) to 48 per
cent (in Great Harwood)." Furthermore it is known that the size of
labour markets for women is much smaller than those for men, that
is, women's activities are more spatially concentrated.v' There are
also considerable spatial variations in fertility rates , illegitimacy
ratios , etc., which suggest a weakness in spatial analyses which ignore
gender divisions and household practices." Moreover, recent in-
quiries into de-industrialisation have posited important local dif-
ferences in access to the informal economy which, it is predicted, will
play an increasingly important part in the production of labour
power in the near future." One account which draws together aspects
of de-industrialisation, household consumption, local class structure
and local politics is that of John Urry.
Urry employed a strategy similar to that of Castelis in disaggregat-
ing some elements of the total process of the reproduction of capital-
ist societies in order to generate an historical periodisation in which
different mechanisms generate different spatial effects at different
times. In The Anatomy of Capitalist Societies he distinguished three
distinct spheres, the economy , civil society and the state. Civil society
was denoted the sphere of circulation, reproduction and struggle; and
it was hypothesised that over time there was a movement within civil
society from the domination of the sphere of circulation to the
domination of the sphere of reproduction. On this basis, Urry iden-
tified a movement in types of political struggle. In the first period
struggles revolved around the establishment of formally equal ex-
change and , typically, generated class struggles. In the second era ,
struggles over the distribution of rewards and the consumption of
commodities led to a decline in the importance of class: other
movements grew in importance, movements of 'women, youth,
blacks' , and class struggle became less intense as the state intervened
to help socialise the costs of the reproduction of labour power."? A
general tendency was diagnosed for the emergence of struggles con-
cerned with the consumption of use-values and the manner of their
provision, whether commodified or not.
In subsequent essays concerned with de-industrialisation and
politics Urry has attempted to specify the spatial consequences of this
scheme." Adopting a view of spatial change in production owing
much to Massey, Urry generates a viewof increasing local differentia-
tion of political conflicts. The new patterns of industrial location have
A/an Warde 209
Conclusion
Notes
13. See the criticisms of Keeble (1977) by Massey and Meegan (1978) about
the validity of alternative economic indicators.
14. For example, Saunders, in this volume.
15. For example, Sayer (1982).
16. See Union of Radical Political Economics (1978) pp. 1-12; Lovering
(1978); and the summary by Massey (1978a).
17. See the comprehensive survey of Lebas (1982).
18. Lojkine (1976). But note that his position seems to have modified con-
siderably of late (see Lojkine, 1981).
19. Lipietz (1980a).
20. Massey (1973; 1979).
21. Massey and Meegan (1978; 1979); Massey (1978a).
22. Massey (1978b; 1979); Massey and Meegan (1982).
23. Massey and Meegan (1978);also the conclusions of Massey and Meegan
(1982).
24. For example, Walker and Storper (1981) who argue a strong case for
labour as the basic source of competitive advantage, but who stress that
the conditions of its reproduction are critical in respect of location
strategies.
25. Such propositions are made on the authority of the influential work of
Frobel et al (1980).
26. Massey and Meegan (1979; 1982); Massey (1978b).
27. Massey and Meegan (1982).
28. It may thus be possible to explain the relatively strong urban-rural shift
in employment in Britain, most recently and comprehensively documen-
ted by Fothergill and Gudgin (1982).
29. See Massey (1979) pp.234-5 for the clearest description of her
'schematic way of approaching this ... historical process'.
30. Massey (1980).
31. Massey (1978a).
32. Massey (1980) p. 1.
33. Seefor example some of the work of Lovering (1978); Cooke (1981); Day
(1982).
34. See Gittins (1982).
35. See for example Liddington and Norris (1978); Macintyre (1980) and
Gittins (1982) for some preliminary suggestions of how domestic arran-
gements and political effects vary over space and time.
36. Sack (1980) passim.
37. Taylor and Johnston (1979) pp.221-69.
38. Nagel (1955).
39. Gregory (1978a) p.46.
40. Giddens (198Ib) pp . 39-40.
41. For example, Johnson (1979); Homans (1969).
42. Lojkine (1981) pp.97-9.
43. Gordon (1978a); or see, alternatively his similar argument (1978b).
44. Lees (1982).
45. Littler (1982).
46. Ward (1980); Shaw (1979).
47. See Anderson (1976) note 59.
212 Spatial Change, Politics and the Division of Labour
Introduction
collectively, or not, as the case may be. The probability that such
motivations have some basis in production-related issues, including
wage-struggles, is at least worthy of some investigation. Such an
approach might have enabled Gramsci to pin down some part of an
explanation of the apparently motiveless 'turbulence' to which the
South was so prone.
The question of when and for what reasons apparently firm struc-
tural constraints upon social practices cease to operate, thereby
opening up substantially the scope for individual action, is by no
means well understood. However, the recent work of Giddens has
helped somewhat to formulate the nature of this core relationship of
social theory.' In elaborating his theory of structuration he gives a
central position to the concept of social practices. These are both
constrained and facilitated by structure which, in turn, is reproduced
by the actions of practising individuals and collectivities. The struc-
tures, as for example those of dominance and subordination of social
classes, cease to be reproduced when practices no longer conform to
them. Interestingly Giddens incorporates a spatial (and temporal)
dimension into this theory of social action arguing that social systems
may be more or less tightly controlled according to the level of
development of integrative institutions - such as the law or the state.
Where there is a relatively less well-developed system of structural
control there may be instances of spatially distinctive social practices
that are also temporally uneven. Cities are seen by Giddens as having
been crucial to the degree of'distanciation' present in a given social
system. This was due to their powers of control through surveillance
and information-gathering, functions since taken up by the state in
contemporary societies.
Unfortunately, Giddens' account does not get us very much closer
to those forms of structuration and disassociation from structural
constraint that have operated beyond the purview of specifically city
or state administration, in the factory or the landed estate. Moreover,
in the last two kinds of setting it is quite conceivable that levels of
surveillance, or at least knowledge about the practices of workers
inside work, were and remain quite high. Yet in both settings there are
plenty of examples of counter-structural practices. Bycontrast, where
counter-structural practices were able to take root and form the basis
for quite socially widespread though often spatially-contained
ideologies and actions , the level of control exercised by the initiators
of such practices was often woefully weak. It is clear that there are
means by which powerful ideas can be formed and transmitted, with
Philip Cooke 219
work for direct consumption. These enterprises are small and locally-
owned and though workers may be conscious of their exploited
condition they are so exposed to unemployment that they can do little
about it. The third category is the traditional craftwork sector where
relatively autonomous skilled workers have both a tradition of
cohesiveness and a reasonable income enabling them to secure a
degree of control over working and living conditions not shared by
the other labour sub-markets. For example, Bleitrach and Chenu
speak of the pauperisation by space and time imposed by the long
journeys to work of steel and chemical workers to their free-standing
growth-pole, and by marginalised workers to their small town and
rural workplaces where reproduction costs are low.
Clearly, this approach is not unproblematic: it works with a too
rigid theory of the state and of the labour market to make its points
thoroughly convincing. 8 The presentation of the state acting
smoothly to further the direct interests of French monopoly capital
may be less unbelievable than it would for other developed econ-
omies, but it still comes over as rather too slick an account of the
growth pole programme. The extent to which Gaullism, with its
intensely patriotic strand, was simply being expressed in economic
modernisation, or, as well, in a quest to unify a politically diverse
society by responding to pressure for development 'from below', is
unexplored here. Similarly, one wonders how far the formalism of the
triple labour market model captures the agency effects of organised
labour in addition to those of the different kinds of capital which
seem so neatly to have sewn up the regional economy." Nevertheless,
these formalisms do not entirely obfuscate some of the useful findings
of the research. The methodology employed has usefully sought to
link together production relations and certain relations of civil
society - family, locality, cultural and ideological practices - in
explaining the nature of a specific socio-spatial formation and
elements of its deformation.
In 'conclusion, therefore, each of these contributions helps to
establish the importance of the spatial dimension to the understand-
ing of differentiation in the characteristic social practices conducted
by particular class-groupings . The spatial edges of these social prac-
tices can be quite marked as combinations of the structural factors
which have been discussed take on a different emphasis. It is impor-
tant to see such changes as clear ones rather than in terms of an
imperceptible merging of practices along some kind of continuum.
This is because - as will be seen in the section which follows - even
222 Class Practices as Regional Markers
Theproductivebase
ly reject women's work of the kind described, but this simply was not
the case.26
A deduction that can be drawn from the absence of exclusion of
women from heavy and dangerous work in a part ofindustrial South
Wales where unionisation was relatively weak, paternalism was
strong and labour supply was not over-abundant, is that the control
of the female labour supply is a source of male bargaining power in a
context of antagonistic capital-labour relations. The basis for this
power is the fact that labour power reproduction is not, in the final
analysis, something over which capital has direct control. Rather,
labour power is produced in the domestic sphere and, as such, is
subject to controls other than the mere demands of capital. In a
context where capital is capable of impoverishing a workforce gen-
erally through depressing the average wage rate by employing
women, it makes sense for men to use the defensive institution of the
trade union to exclude the source of wage-rate reduction and then
argue for a family wage. Of course, had patriarchy not been the
condition upon which the convention was established that income-
disparity by gender was an appropriate practice, organised labour
would not have been so vulnerable to this problem in the first place.
This seems, in conclusion, to cast considerable doubt on the notion
that the domestic division of labour is somehow essential to the
development of capitalist relations of production.s? If anything, the
domestication of women, at least as far as South Wales is concerned,
was an outcome secured by organised male labour to fight income
erosion . It was then sustained by its capacity to enforce patriarchal
relations in the home, in part through the institutionalisation of the
family wage.
A final point concerns the role of women as supporters of militant
trade union practices despite their confinement to the domestic
sphere. Their assertiveness outside the domestic sphere leads to a
questioning of their acquiescence within it, and Macintyre's allusion
to the power of patriarchy as expressed through the medium of male
violence towards women seems a plausible, if partial, explanation."
There are three other factors which deserve special attention in
considering the determination of specific social relations in the con-
text of industrial South Wales. The first of these is historical and
concerns the nature of the transition across Giddens' 'time-space
edges' or between different modes of social relations, expressions of
distinctive productive bases under the same general relations of
market exchange and production for profit. A key marker of this
Philip Cooke 233
The main financial returns came from the sale of store cattle and of
casked salted butter. But no holding had calves to rear for sale or
milk for butter-making without the services of a bull. Bulls
however are expensive animals to keep, and only on holdings of
some 65 acres (or more) could they be kept. Five holdings out of
every six were smaller than this, so that every five holdings depen-
ded on the sixth for a bull's services. And payment for these services
was not in cash but in labour at the hay harvest ... local
idiom . .. indicates that people saw the society in which they lived in
terms of the three main divisions of'gwyr mawr' (gentry), farmers
and 'pobol tai bach' (cottagers) . .. but in certain contexts they
(farmers and cottagers) were both ofthem 'pobl gyffredin ' (ordinary
people) vis avis 'y gwyr mawr' and the threefold division of society
seems to become twofold."
[Welsh] was the language of the common people in the sense that it
was incorporated into their own institutions. The chapels had
originated as their institutions: it was they had provided the elite,
and for nearly a century there had been harmonious understanding
and an implicit realisation that they - the common people - could
exert some kind of control. After all, this was why they were
nonconformists and not Anglicans."
Institutional specificities
Out of this turbulence emerged institutions which sustained impor-
tant elements of a prefigurative proletarian culture, in the sense of a
coherent set of intellectual, cultural and ideological practices. These
were widely understood and reproduced during a period when the
dominant culture and ideology of imperial Britain was militaristic,
nationalistic and individualistic."
Only three of these can be noted here; trade unions, health associa-
tions, and educational institutions, although others such as re-
creational and leisure practices, popular entertainment, but above all,
political practices, are clearly important. One reason for playing
down political practices is that through the development of Labour-
ism from its earlier strands in the Independent Labour Party, the
chapels and early trade unions, South Wales is not especially distinct
from other heavy industry regions in Britain. However, the mass
support which the British Labour Party gained from industrial South
Wales was constructed on a far more thoroughgoing basis of resis-
tance to ruling ideas and practices than was normal in such regions."
Trade unionism, as we have seen, was initially the province of craft-
workers who had effectively lost control in the coal-mining industry
by1900, but retained it until the 1960sin the steel mills. Initially trade
unionism in South Wales was fragmentary (organisation in coal being
centred on eight small, local unions , five of which possessed no
funds). But with the massive growth in numbers of unskilled workers
flooding to the coal, transport, steel, construction and retailing indus-
tries as coalfield production took off at the end of the nineteenth
century they gained in stability. A key factor in this process was the
integrating influence provided by trades councils located in the larger
urban centres . Interestingly, it was often the railway workers who
took a leading role in defence of the right to strike and in demands for
industrial unionism at this time."
However, so large was the mining workforce (270 000 in 1920) that
it exerted an overwhelming dominance over the political and social
life of the region . This is exemplified in the ways in which, once
localism had been overcome and the South Wales miners had joined
the Mining Federation of Great Britain and affiliated to the Labour
Philip Cooke 237
Party, the union itself became the major cohesive force in the regional
formation:
From the chapels to the free libraries, from the Institutes to the
sports teams , the control was a popular and a democratic one. The
organisation ofpolitics in the localities had a similar intent (diverse
ideologies notwithstanding), and in all this the primary organiza-
tion was the Union . . . the SWMF was, literally, the fount of
control in other spheres as well as that of the industry itself? "
Concluding remarks
This study has been premised on the notion that uneven spatial
development is not merely an economic phenomenon which dis-
tributes industrial sectors around the space-economy in heterogen-
eous fashion but with relative homogenising effects upon the class
relations which are formed in the diverse parts of the system. On the
contrary, it assumes there to be important socio-spatial variations
that are contingent upon the uneven character of capitalist develop-
ment. These variations arise from a combination of particular,
spatially-distinctive conditions of production, on the one hand, and
spatially-delimited social practices on the other. These practices ex-
press the relative domination and subordination of classes which
form the most necessary condition of production for profit.
Rather than provide merely a reprise of the theoretical, analytical
and empirical argument which has been offered in justification of that
conclusion, I will draw attention to the three features which, I think ,
Philip Cooke 239
Notes
tices which were once the coping resources of the poor - for example,
making jam or keeping chickens - have now been adopted by more
affluent strata seeking a more 'natural' life-style, but using advanced
technology where appropriate. The very forces which would seem to
encourage the development of the electronic home have produced a
reaction which stresses the value of craft production. A bourgeois
reaction against mass-production mirrors in a curious way necessary
activity for the needy. For the working class, the stripping of a car
engine or the production of a giant marrow may serve to support a
masculine identity in a way which perhaps waged labour no longer
can . So deskilling in the formal sector is matched by re-skilling in
informal work .
Furthermore, the effect of economic recession may be to encourage
people who have the time and resources to use their own facilities to
produce more goods and services for themselves as publicly provided
services and facilities decline. It is important to recognise, therefore,
that the same overall tendency has been seen by some commentators
as a coping response of an increasing proportion of the poor," while
others see it as a reflection of greater affluence, more free time and
greater technological skill and resources."
__.. . --_-1-
DIVISIO N OF LA8D UR ~
related with the policies and practices of both local employers and
local and national government. Other, unrecorded, work in the
household and between households should also be considered and
opportunities for doing this involve further considerations such as the
age and tenure of the housing stock, access to alternative sources of
food and raw materials and so on. These other elements of a local
milieu have generally been disregarded by geographers and regional
economists over the past thirty years. A previous generation was
perhaps more sensitive to the nuances of the local, particularly those
influenced by Vidal de la Blache, Jean Brunhes, H . J. Fleure and
others. It might not be too fanciful to return to a concept coming
directly from de la Blache, namely the personality of place. This
emphasis on cultural diversity is a useful corrective to an approach
which, perhaps over-deterministically, attempts to relate local prac-
tices of survival and 'getting by' to shifts in the global process of
capital accumulation. Certainly, one would want to resist the sugges-
tion that a given set of material conditions produces a direct response
through specific household work practices. Such would be a naive
and over-deterministic position. Nevertheless, one must equally resist
a completely voluntaristic and culturally-specific response, implying
that people have more control over their life chances than can possi-
bly be the case. The stance which is being adopted here attempts to
254 Capital, Political Economy and Household Work
hold a middle position between these two extremes, not out of some
pragmatic desire for compromise but rather because such a position
carries greater theoretical persuasion. It is claimed that household
work strategies are not a direct response to material conditions,
simply because these very material conditions are partly created by
the practices of previous generations. Indeed, cultural lag is an ex-
tremely important concept for understanding the practices of
everyday life. Working-class culture is a slippery concept, to be sure,
but , however hard to handle, that is no excuse for denying its reality.
Local cultural practices and traditions do vary from one milieu to
another and help to create the conditions which encourage new, and
distinctive, forms of capitalist penetration. Patterns of deference or of
militancy, traditions of married women's employment, distinctive
strategies ofgetting by, forms ofcollectivecommunal action, whether
instrumental or expressive, and many other such practices and tradi-
tions have distinctive geographical distributions. Massey's notion of
'layers' of capital is similar to the notion of a political generation as a
means of understanding contemporary configurations. A given
milieu needs to be understood, therefore, in terms of these layers of
capital investment, but also of micro-cultural styles, political genera-
tions, waves of house-building, immigration, local and national
government policies and much else besides. Geographers will, per-
haps, be more ready to accept such a stratigraphical analogy , the
erosion of some strata and the stubborn resistance of others to the
attacks of eroding agencies. Perhaps 'social geomorphology' would
be a better basis for analogy than sociography: at least the former
carries with it implications of historical process.
o 30m ,In
0- -50 km
there has been substantial investment over the past twenty years in the
establishment of a deep-water port at Sheerness, which is now one of
the main ports of entry for imported Japanese cars, and the develop-
ment of plants for Sheerness Steel and Abbotts Laboratories (making
pharmaceutical products) financed by Canadian and American
capital. Such foreign investment means that half of the main indus-
trial employment is controlled from outside Britain . The unem-
ployment rate is now just under 20 per cent. Opportunities for
employment for school-leavers are very limited as there is hardly any
development of service industries and married women compete for
semi-skilled factory work, increasingly on a part-time basis. Men
made redundant from the older, declining industries generally find it
impossible to get employment on the Island. The highly-paid
stevedores in the docks are mostly young men who have effectively
secured their own jobs, but, with mechanisation, fewer are needed
and natural wastage is not replaced. Similarly, the steel mill is reduc-
ing rather than recruiting labour: early in 1983 there were fears about
its future viability. Sheppey is not part of the newly-created Enter-
prise Zone which includes Gravesend but not Chatham. The holiday
industry is also in decline.
Away from the main industrial areas, there are wide open pastures
and saltings with geese, ducks and rabbits; the sea can provide fish
and pleasure in summer and there is a large new private housing
256 Capital, Political Economy and Household Work
she claimed was their good sense. Unhappily the loyalty of both
management and workers was not rewarded: the British government
refused to provide assistance when the company began to suffer from
the limitations of the EEC restrictions on steel production, and the
men who had been promised job security began to be laid off.
This unhappy experience, first of the dockyard workers and later of
those in the steel mill, was paralleled in smaller ways when other
companies which had been encouraged on to the island in the 1960s
with subsidies and cheap premises began to collapse in the late 1960s.
Some were taken over by larger corporations and then closed under a
rationalisation policy; others had cash-flow problems or simply bad
management. The history of employment and unemployment over
the past twenty years was, until the late 1970s, an erratic pattern of
rises and falls as firms came and went. More recently unemployment
has simply continued to rise.
Realistically, the people of Sheppey know that there is absolutely
nothing that they can do to increase their employment prospects.
British managers of multinational organisations have to explain to
their workers that if they object to what is proposed - redundancies,
rationalisation or whatever - the plant will simply close altogether or
move elsewhere. A long history of such closures easily persuades the
workers that the threat is not idle. The island workforce is geogra-
phically imprisoned. Women accept 60 pence an hour to do simple
component assembly work in their homes. There is little alternative
and there are as many on the waiting list to do such work as there are
already employed . Almost anything is accepted . Hundreds of acres of
land behind one of the housing estates are covered with Japanese cars:
the people were offered £300 toward their local community centre by
the company concerned. Even if the economy picks up, Sheppey
people know that the most efficient employers will continue to sub-
stitute machines for labour: they see as clearly as any academic
analyst the logic of their own disaster.
What, perhaps, they do not see yet are the cleavages based on
gender which are becoming more pronounced. Thus , women, clas-
sified as semi-skilled workers, are cheaper and more skilled than men
who are classified as skilled, and many of the manufacturing compan-
ies prefer to have a more easily controlled workforce. Rarely is it
necessary to make a woman redundant; natural wastage provides a
painless way of adjusting labour supply and demand. This may be
related to the domestic cycle, but some women resign not because
they want to but because the taxation system very seriously depletes
258 Capital, Political Economy and Household Work
any income they earn if their husbands are unemployed. Many firms
and their women workers prefer the flexibility of part-time work,
which carries with it the assumption that another member of the
household is the chief earner.
Although the pattern of employment is gripped in a downward
spiral , not all households are equally penalised . One of the paradoxes
is that the wives of those men in employment may find it easier to be
employed . For example, stevedores earn very high wages - certainly
in the island's terms - and by and large their work is more boring than
tiring, involving sitting in a fork-lift truck when there is a ship to
unload. However, they have the time and very often the skills to do
most of the home decoration, maintenance and improvement that is
required. Hence their homes are likely to be in good order, well-
equipped with consumer durables and thus very easy to keep clean.
Their wives are then free to enter employment once their children
enter full-time education. Such households are, comparatively, ex-
tremely well-off, and it is likely that these divisions will intensify:
increasingly those in employment speak for other members of their
household when they hear of vacancies. A polarisation is developing
between those households with multiple earners and those with none.
time, fruit, meat and Japanese radios were imported through the
docks and I quickly gathered a host of anecdotes in the pubs and on
the doorsteps. A respectable old lady walking by the walls of the
docks told of sheep flying through the air in front of her, and in some
cafes, it was said, steak and chips were cheaper than egg and chips.
Everyone seemed to know someone in the docks and people had
deep-freezes in sheds and garages stuffed with meat. The honest ones
would boast that they only had the 'odd crate of oranges' as if that did
not count. Certainly free bananas never counted as pilfering. At the
same time there were many painters, window-cleaners and home car-
tuning experts who advertised in the local paper. Materials were
relatively easy to come by. People laughed at a case reported in the
paper of a man who appeared in Court for receiving stolen property.
His defence was that he had done no more than ask in the local pub
where he could buy 'cheap bricks' . His mistake was that he gave
someone his address . He woke up one morning to find a load ofbricks
outside his house. Or so he said. The joke was the feebleness of his
excuse.
At that time the pubs were full most nights of the week - and with
over 140 licensed premises on the island that meant a lot of drinking.
The local brewers in Faversham saw Sheppey as one of their best
markets. People were busy doing up their houses, making extensions
to nineteenth-century cottages and doing up and selling old cars .
Sometimes this work was for money (shadow wage-labour),
sometimes it was simply to do a favour for a friend (communal work).
Many working-class people went away for a holiday for the first time
in their lives. They came back with Spanish dolls and guitar-shaped
wall decorations and they make a cocktail bar in their front room like
the one they saw in the hotel.
Those who were unemployed at that time were in a favoured
position. There was still a reasonable amount of fish to be caught in
the local waters, there were duck or rabbits to be shot and people had
money to pay for odd jobs being done . 'Cheap' goods were available
to circulate round the island, holiday-makers still brought a lot of
money to spend in Leysdown each summer, and the owner who lives
above the amusement arcade and is said to be a millionaire, hired
youngsters straight out of school every summer. The ordinary school-
leaver from Sheppey Comprehensive School traditionally never con-
sidered employment until the autumn - leaving at Easter he or she
would 'go down Leysdown' in May and either clean chalets or take
the money for machines and fair ground 'attractions'. People I met
260 Capital, Political Economy and Household Work
Conclusions
Acknowledgements
Notes
I. Cox (1978); Dear and Scott (1981); Harvey (1981); Massey (1978a; 1979;
1982a); Massey and Meegan (1982); Mingione (1981); Urry (1981c).
2. Szelenyi (1981).
3. Castells (1977); Harloe (ed.) (1977; 1981); Harvey (1973); Lojkine (1977);
Pickvance (1976); Saunders (1979; 1981).
4. Castells (1977).
5. Lebas (1981; 1982).
6. Frobel et al; (1980); see also Elson and Pearson (1981).
7. Szelenyi (1981).
8. Massey (1978a; 1979; 1982); Massey and Meegan (1979a).
9. Massey (1982).
10. Urry (1981c); Rees and Lambert (1981); Cooke (1982a).
II. Blackburn and Mann (1979).
12. Molyneux (1979); Sokoloff (1980).
13. Gershuny and Thomas (1980); Gershuny (1982; 1983); Meissner (1975).
264 Capital, Political Economy and Household Work
Time-Geography
(AI (B ) --,----------::-i
, ", ' I
Sp ace
Co-Iocanon In tim e- spa ce Th ree-dim ensronat time-space
The diagrams in Figure 12.1 show this in its simplest guise. Two
individuals live, say, a mile apart in a neighbourhood; their time-
space paths during the course of the day bring them into contact in
an encounter of short duration, for example, a restaurant, after which
their activities again diverge. If the daily activities of a specific in-
dividual are recorded, it is easy to build up a gross characterisation of
his or her routine activities, insofar as these comprise trajectories in
time and space. As a portrayal of a life-path, this would involve
generalised patterns of time-space movement within the 'life-cycle'.
A person may livein the house ofhis or her parents, for example, until
268 Time, Space and Regiona/isation
Critical comments
The interest of time-geography to the theory ofstructuration is surely
evident. l l Time-geography is concerned with the infrastructural con-
straints that shape the routines of day-to-day life, and shares with
structuration theory an emphasis upon the significance of the prac-
tical character of daily activities, in circumstances where individuals
are co-present with one another, for the constitution ofsocial conduct.
Hagerstrand's concentration upon everyday social practices is very
270 Time, Space and Regionalisation
Modesof regionalisation
'Regionalisation' should not be understood merely as localisation in
space, but as referring to the zoning of time-space in relation to
routinised social practices . Thus a private house is a locale which is a
'station' for a large cluster of interactions in the course of a typical
day. Houses in contemporary societies are regionalised into floors,
halls and rooms. But the various rooms of the house are zoned
differently in time as well as space. The rooms downstairs are charac-
teristically most used in daylight hours, while bedrooms are rooms to
which individuals 'retire' at night. The division between day and night
in all societies used to be perhaps the most fundamental zoning
demarcation between the intensity of social life and its relaxation -
ordered also, obviously, by the need of the human organism for
regular periods of sleep. Night-time was a 'frontier' of social activity
as marked as any spatial frontiers have ever been. It remains a
frontier, as it were, that is only sparsely settled. But the invention of
powerful, regularised modes of artificial lighting has dramatically
expanded the potentialities of interaction settings in night-hours. As
one observer has remarked:
the day. There is more multiple shift factory work, more police
coverage, more use of the telephone at all hours. There are more
hospitals, pharmacies, aeroplane flights, hostels, always-open res-
taurants, car rental and gasoline and auto repair stations, bowling
alleys, and radio stations, always active. There are more emergency
services such as auto-towing, locksmiths, bail bondsmen, drug and
poison and suicide, gambling 'hot lines' available incessantly . Alth-
ough different individuals participate in these events in shifts, the
organisations involved are continually active."
Disclosure
I was working on one side of the car and the boot lid dropped. It
just grazed the head of the fella working opposite me. I can see it
now. He stopped working , had a look round to see if anyone was
watching. I was pretending not to look at him - and then he held his
head. He'd had enough like. You could see him thinking 'I'm
getting out of this for a bit'. He staggered; I could see him looking
round. You know what it was like in there . Paint everywhere. He
wasn't going to fall in the paint . . . so he staggered about ten yards
and fell down with a moan on some pallets. It was bloody funny.
One of the lads saw him there and stopped the line. The supervisor
came chasing across. 'Start the line .. start the line . . .' . He started
the line and we had to work. We were working one short as well. It
took them ages to get him out of there . They couldn't get the
stretcher in. It must have been half an hour before they got him.
Him lying there y'know with his one eye occasionally opening for a
quick look round: 'What's happeningr'P
Regionalisation as generic
The differentiations between enclosure, disclosure, back and front
regions, apply across large spans of time-space, not only in the
contexts of co-presence. These are, of course, unlikely to be as directly
monitored reflexivelyby those whom they affect, although such may
be the case. Regionalisation within urban areas in contemporary
societies have been much studied since the famous early work of the
Chicago sociologists Park and Burgess. In most Western societies, the
zoning of cities into neighbourhoods with markedly different social
characteristics is strongly influenced by the operation of housing
markets, and by separations between individually-owned homes and
state-operated housing sectors. Neighbourhoods may not be as sym-
metrically zoned as some of the 'ecological' urban analysts suggested,
but their distribution certainly has the consequences of creating
various sorts of front /back contrasts. Industrial areas in northern
towns and cities in England were once the most visible features of the
built environment - factories and mills, as it were, proudly displayed.
But the tendency in urban planning in recent years has been to treat
such areas as unsightly, as back regions to be hidden away in enclosed
enclaves, or transferred to the edge of town. Examples can easily be
multiplied. The access of those in more affluent sectors of housing
markets to relatively easy transfer of property underlies the 'flight to
the suburbs' that has helped to transform city centres from regions of
frontal display to back regions of urban decay, which the 'respectable
classes' avoid. Ghetto areas may be rendered 'invisible' by their
regional enclosure in neighbourhoods having both very low rates of
property transfer and of daily mobility from those neighbourhoods
into other parts of the city. As always, various types of time-series
phenomena underlie such spatial regionalisation.
Regionalisation across long spans of time-space has been
analysed by many writers in terms of familiar notions such as 'uneven
development' and distinctions between 'centre' (or 'core') and 'peri-
phery'. These notions however, can be applied across the whole range
of the settings of locales, from large to small. Rather than discussing
the theme of uneven development here, however, I shall develop the
differentiation of centre and periphery by relating it to 'embedded-
282 Time, Space and Regionalisation
ness in time' . If the world economy has its centres, and cities have
their centres, so too do the daily trajectories of individual actors. In
modern societies, for the majority of males at least, the home and
workplace form the two main centres in which the day's activities
tend to be concentrated. Locales also tend to be centred regionally.
For example, some rooms in a house, such as spare bedrooms, may be
only used 'peripherally'.
Distinctions between centre and periphery tend frequently to be
associated with endurance over time.P Those who occupy centres
'establish' themselves as having control over resources which allow
them to maintain differentiations between themselves and those in
peripheral regions. The established may employ a variety offorms of
social closure to sustain distance from others who are effectively
treated as inferiors or outsiders." The 'established' industrial nations
Established
Let me at this point offer a summary of the main points in this chapter
so far. The discussion has been concerned with the contex tuality of
social life and social institutions. All social life occurs in, and is
constituted by, intersections of presence and absence in the 'fading
away ' of time and the 'shading off' of space. The physical properties
of the bod y, and the milieux in which it moves, inevitably give social
life a serial character , and limit modes of access to 'absent' others
across space. Time-geography pro vides an important mode of nota-
tion of the intersection of time -space trajectories in day-to-day
activity. But it has to be inserted within a more adequate theorisation
both of the agent and of the organisation of the settings of interac-
tion . In proposing the ideas ofiocale, and of regionalisat ion , I wish to
begin to formulate a scheme of concepts which help to categorise
contextuality as inherently involved in the connection between social
systems of smaller and larger scope." The graphic techn iques de-
In tersection of locales
(A) (8 )
Spa ce
Space Space
Bell for 9.0 a.m. goes, about half class in, mostly reading books.
Teacher enters breezily - 'Morning - ah, that's good, getting those
books out'. Teacher sits at desk, tidies up, gets register out. Mean-
while most of the other children have come into the classroom. The
later arrivals talk, swap some football cards, occasionally glance at
the teacher .
Teacher: Right , let's do the register, then , hurry up and sit down
you football maniacs - I see that Manchester United lost
again .
Manchester United Supporters: Oh yeah, well they're still better than
Liverpool.
Teacher: Gokey sarcasm in voice): Really? It must be all the spin-
ach they don't eat. Now then . . . Martin ... Doreen . . .
Alan . . . Mark (calls register and children answer) .
A child comes in late, looking sheepish, and walks to his seat. Other
children point and laugh .
children, these lie in some part along the narrow temporal boundaries
between classes, whether or not they involve physical movement from
one classroom to another. Although the weight ofdiscipline normally
bears down most on the children, it is often felt more oppressively by
teachers. Teachers usually have a back region to which they can
retreat - the staff room - which children ordinarily do not enter. The
staff room is no doubt a place for unwinding and relaxation. But it is
also somewhere in which tactics ofcoping with teaching tend endless-
ly to be discussed, formulated and reformulated.
It is in the nature of disciplinary organisations that the intensity of
surveillance inside inhibits direct control from outside. This is a
phenomenon which can be seen both in the internal regionalisation of
the school and its situation as a locale within other locales. Inside the
school, the concentration of disciplinary authority in separately-par-
titioned classrooms is the condition of the high level of control over
bodily positioning and activity which can be achieved . But this
circumstance also inhibits the direct supervision of the supervisor.
The head is 'in authority' over the teaching staff, but such authority
cannot be exercised in the same way as teachers endeavour to control
the conduct of children in their classes. Schools therefore tend to have
a rather sharply opposed 'double line' of authority. The control
which teachers seek to exercise over their pupils is immediate, involv-
ing the teacher's continuous face-to-face presence with the children.
Supervision of the activity of teachers , however, is necessarily in-
direct , and proceeds by other means. One might hazard a guess that it
is only in organisations in which a considerable amount of autonomy
from direct supervision is given in which a graduated line ofauthority
can be achieved . The enclosed nature of the school, and its clear
separation in time and space from what goes on in surrounding
locales, also inhibits supervisory control from the outside. Thus
inspectors may visit schools regularly to check upon their operation;
boards of governors and parents' associations may make their power
felt in influencing policies that help shape the life of the school. But it
is intrinsic to disciplinary power that what goes on in the 'power
container' ofthe school has a significant degree ofautonomy from the
very outside agencies whose ethos it expresses.
social systems. I do not employ the more familiar terms, 'micro-' and
'macrosociological' study for two reasons . One is that these two are
not infrequently set off against one another, with the implication that
we have to choose between them, regard ing one as in some way more
fundamental than the other. In Goffman's studied refusal to be
concerned with issues of large-scale social organisation and history,
for example, there seems to lurk more than a hint that in what he
sometimes calls microsociology is to be found the essential reality of
social life. On the other hand, advocates of macro sociological
approaches are prone to regard studies of day-to-day social activity
as being concerned with trivia - the most significant issues are those
of broader scope. But this sort of confrontation is surely a phoney
war if ever there was one. At any rate, I do not think that there can be
any question of either having priority over the other. A second reason
why the micro/macro division tends to conjure up unfortunate
associations is that, where there is no conflict between the two pers-
pectives, an unhappy division oflabour has tended to come into being
between then . Microsociology is taken to be concerned with the
activities of the 'free agent', which can safely be left to theoretical
standpoints such as those of symbolic interactionism or ethnometh-
odology to elucidate ; while the province of macrosociology is that of
analysing the structural constraints which set limits to free activity .
Such a division oflabour leads to consequences that are at best highly
misleading .
Why should the issue of the relation between 'micro-' and
'macrosociological' study be seen as so problematic by many writers?
The conceptual division of labour to which we have just referred is
undoubtedly the most significant cause. Reinforced by a philoso-
phical dualism, it cannot be escaped except by a more thoroughgoing
reformulation of social theory than most authors are able or willing
to contemplate. It will help to develop this point by looking briefly at
one of the more interesting recent discussions of the issue, offered by
Collins." Collins points out that the schism between micro- and
macrosociological approaches, as these terms are ordinarily under-
stood, has become accentuated over the past decade or so. While
social theory was dominated by functionalism and Marxism, or some
combination of the two, social relations in situations of co-presence
were typically regarded as substantially determined by broader
'structural' factors . However, as led especially by ethnomethodology,
microsociology has become a burgeoning field of interest , and one in
which the presumptions of these approaches have been taken to task
in a fairly radical fashion . In Collins's view:
Anthony Giddens 291
Such cities, we could say, do not yet exist in commodified time and
space." The buying and selling oftime - as labour-time - is surely one
of the most distinctive features of modern capitalism. The origins of
the precise temporal regulation of the day may perhaps be found in
the chime of the monastery bell; but it is in the sphere of labour that
its influence became embedded in such a way as to spread throughout
society as a whole. The commodification of time, geared to the
mechanisms of industrial production, breaks down the differentia-
tion of city and countryside characteristic of class-divided societies.
Modern industry is accompanied by the spread of urbanism, but its
operation is not necessarily fixed in any particular type of area.
Traditional cities, on the other hand, are both the main locus of
disciplinary power in class-divided societies, and as such set off from
the countryside - very often , physically and symbolically, by the city
walls. Together with the transformation of time, the commodifica-
tion of space establishes a 'created environment' of a very distinctive
Anthony Giddens 295
Notes
1. See Hiigerstrand (1970; 1973; 1975); also Gregory (1978; 1982a) and this
volume; Carlstein (1980); Pred (1977); Parkes and Thrift (1980); Thrift
(1983).
2. Hagerstrand (1975); cr. Parkes and Thrift (1980) pp .247-8.
3. Pred (1978).
4. Hiigerstrand (1967)p. 332; cf. Hawley (1950)ch. 13-15; Ericksen (1980).
5. After Parkes and Thrift (1980) p.245.
6. Janelle (1969).
7. Forer (1978).
8. Palm and Pred (1978).
9. Hiigerstrand (1978) p. 123.
10. Carlstein (1978) p. 159; Carlstein (1980).
11. cf. Carl stein (1981).
12. Hiigerstrand (1970) p.8.
13. Giddens (1981b).
14. Giddens (1981b) pp.161 ff.
15. Melbin (1978) p. 100.
16. Zerubavel (1979) p. 22; cr. Clark (1982)andZerubavel (1981). One might
point out that while the 'year' , 'month' and 'day ' have links to natural
events, the 'week' does not: see Colson (1926).
17. Aries (1973); Elias (1978b).
18. Hall (1966) p. 98.
19. Goffman (1959).
20. Goffman (1959) ch. 3.
21. Giddens (1981b) p. 169.
22. Beynon (1973) p. 76.
23. Elias (1978).
24. Goffman (1959) p. 128.
25. cr. Elias and Scotson (1965).
26. Weber (1978) pp. 341-4.
27. Giddens (1979) ch. 9.
28. Giddens (1981b) ch. 5 and passim.
29. Bourdieu (1977) pp. 143-52.
30. Bourdieu (1977) p. 153.
31. Pollard (1980).
32. Collins (1981b); see also Collins (1981a).
33. Collins (1981b) p. 82.
34. Collins (1981b) p. 99.
35. Rykwert (1976) p.202.
36. Giddens (1981b) ch. 5.
13
Suspended Animation: The Stasis of
Diffusion Theory
DEREK GREGORY
Introduction
bonds, even if they have rarely been able to provide a wholly satisfac-
tory explication of them . Giddens is no exception. Although the
primary focus of his theory of structuration is the topography of
time-space routines , he also discusses what he terms 'time-space
distanciation' and connects this to combined and uneven develop-
ment and to the administrative space of the nation-state.' But his
writings seem to me to marginalise the importance of locational
structures, and in seizingupon Hagerstrand's time-geography he has
seized upon the least 'geographical' of all geographies: or so I shall
claim.(') This is not to say that an unreconstructed location theory can
offer much guidance, however, and contemporary theorisations of
locational structure are embedded in much more inclusive theorems
about the production and reproduction of society than the 'separat-
ist' formations of conventional spatial scienceusually allowed." Even
so, these modern endeavours remain strategically incomplete in so far
as they continue to confine human agency within a theoretical grid
which represents the changing configurations of capitalism as
moments in the unfolding of an immanent 'logic'. There are naturally
some exceptions, since these enclosures are far from necessary ones,
but in large measure the contingencies of 'struggle' seem to be subor-
dinated to the structural compulsions of 'capital' in ways which
tremble on the very edges of the functionalism so strenuously re-
pudiated by Giddens and others. 8
The integration of these two problematics is thus an urgent task,
but it cannot be tackled on a meta-theoretical level alone." Partly for
this reason , the present discussion is confined to a critique of dif-
fusion theory and centres on Hagerstrand's seminal contributions
over the last thirty years. There are, of course, older traditions of
diffusion research, both inside and outside human geography. Some
of them, like the classical investigations in rural sociology during the
1920s and 1930s,are open to many of the same objections as Hager-
strand's early writings;'? others, and particularly Sauer's much less
formal reconstructions of evolving cultural morphologies, derive
from an anthropology whose discriminations between 'different
levels of reality' are at least partially congruent with the rejection of
methodological individualism entailed by a relocation of diffusion
theory within the philosophy of realism which I seek to explore in this
essay.'! Although I shall not discuss these other traditions in any
direct way, therefore, they are not completely- absent from the
argument: if such a circumscription is limiting, then it is not, I think,
disabling. Indeed, it facilitates what is sometimes called 'de-construe-
Derek Gregory 299
~- ~~I ~
~-
j
~ ~.. , . , ~J 1 i~
.
~': " : " :,~.'~1 l .. .. . :,.".-. •
,\
~
'~" '.'
-....,.r
j ~·'lr: .: ·
-........x:•.
......
., '1. .
., . -~
.~.~
. '- l
060402 OT
to close the gap between detailed story -telling and simplified statis-
tical model-building. Each side has to give up some of its present
ambitions . .. But still further adjustments are needed. In order to
locate the full picture of an innovational process in the broader
context of prerequisite conditions and offollowing consequences we
need a way of mapping the whole chess-board on which the game
takes place."
TheMellstock Quire
Time-space paths
Hagerstrand's diffusion theory had its origins in an investigation of
what he termed the 'population archaeology' of Asby. 'My attention',
so he recalled, 'was not to count anonymous numbers but to excavate
the individual biographies of everybody who had breathed the air of
Asby from 1840 and onwards up to 1940' .30 It was then that he
conceived'of depicting a biography as a time-space path', but he had
306 The Stasis of Diffusion Theory
In his derivation of the mean information field, which was the mains-
pring of his diffusion theory , Hagerstrand suggested that:
Path
c: ~
~.:-; Domain
Stat ion
.,
E
t=
Spa ce
Naturalism
Hagerstrand's diffusion theory was imbricated in a physicalism
which ultimately derived from Carnap and Neurath, and which was
part of the Vienna Circle's attempt to demarcate science from meta-
physics. In their view, scientific discourse was to be conducted
through 'one unified language which spoke about physical things and
their movements in space and time' and which took modern physics
as its model. Physics and psychology were seen as opposite poles:
'Physics is, altogether, practically free from metaphysics, thanks to
the efforts of Mach , Poincare and Einstein; in psychology, efforts to
make it a science free from metaphysics have hardly begun .' Accord-
ing to Popper, a fierce opponent of the physicalist programme:
Everything was to be expressible in this language, or translatable
into it, especially psychology in so far as it was scientific. Psy-
chology was to become radically behaviouristic; every meaningful
statement of psychology, whether human or animal , was to be
translatable into a statement about the spatio-temporal movement
of physical bodies."
310 The Stasis of Diffusion Theory
PROJE CT
- - Trans,ler
----- - St c raqe
Spa ce
and spaces in the fabric of[their] period and place' and are 'slotted or
fitted into certain parts of more or less cyclical activity systems.' This
means that they transform coupling periodicities, and through these
'displacement effects' set in motion a round of repacking activities in
space and time. In this way, the conditions and consequences of
innovation diffusion are written into what Hagerstrand calls the
space-time 'scores' of co-existing and competing projects."
These continuities set important limits to Hagerstrand's own
project, but I now want to draw out three contrasts between diffusion
theory and time-geography. As I hope will become clear, these chan-
ges of emphasis are closely interlocked and provide a series of keys
which, when filed down and polished, eventually open the doors to a
more rigorous social theory.
Notice that it is now order in time rather than order in space which
gives shape and substance to social life. Some commentators have
objected to this reading: to Parkes and Thrift, for example, it is a
'fundamental misconception' of the nature of time-geography. 54 But
here is Hagerstrand himself:
Whenever this spatial grid - the weft, so to speak, rather than the
warp - is reduced to a graph, however, it becomes exceptionally
difficult to explicate its developing historical geography. And it is
vitally important to do so because the casting and recasting ofwebs of
interaction is clearly not independent of the production and re-
production of the locational structures which contain them . These are
Derek Gregory 315
not constants, and Carlstein notes that even in the short term
'elements in the settlement system would change in size, some would
cease to exist, while others would be added to the landscape'<' The
locations of these stations - the 'spatial bases for activity' - are
supposed to be 'strongly determined by the sequence of events in
time' and in particular by the 'conditions of connectivity' which exist
within and between projects.58 In so far as these coupling periodicities
are transformed by technical change (which Hagerstrand regards as
'the prime mover');" then spatial configurations are likely to be
inherently unstable: according to Carlstein 'when innovations are
adopted, the settlement system is almost invariably rearranged to
accommodate them'.60 But these arrangements and rearrangements
need to be accorded a much more central place within time-geogra-
phy, because without a proper theorisation of the instability of these
locational structures - what Harvey sees as the chronic tension
between 'fixity and motion' within the 'structured coherence' of
spatial configurations" - space-time paths become moments in the
continuous, routinised reproduction of an 'enormous maze' : con-
fined within a 'categorial' framework, their webs can only spin in an
endless stasis."
This is patently not what Hagerstrand intends, but the fact remains
that to 'give prominence to the time dimension' does not necessarily
mean that 'studies of change and of development trends . . . take
precedence over examinations of equilibria and steady states'.63 If, as
Giddens has shown, it is an elementary error to identify the elapse of
time with transformation, then it is no less misleading to identify
space with stability." On the contrary, I want to say that models of
transformation require theorisations of spatial structure, not as flat,
frozen lattices - 'the most solid of all geometries' - but as hierarchical-
ly ordered arenas of social practice." This in turn demands a rework-
ing of conventional location theory and its incorporation within a
wider set of theorems about the production and reproduction of
society. In many ways, of course, Hagerstrand's various discussions
of the concept of domain, of 'survival and arena', speak directly to
such a project, since they connect through to an understanding of
domination in terms not of 'the relations between actors' but rather of
'the conditions for action imposed by arrangements on the stage'.66
These notes are programmatic rather than definitive, clearly, and I
shall return to them shortly; but given a proper 'prominence' they
seem to me to gesture towards the constitution of what Foucault calls
the societe disciplinaire, in which, in the first instance, 'discipline
316 The Stasis of Diffusion Theory
Space was treated as the dead, the fixed, the undialectical, the
immobile. Time, on the contrary, was richness, fecundity, life,
dialectic .. . The use of spatial terms seems to have the air of an
anti-history. If one started to talk in terms of space that meant one
was hostile to time. It meant, as the fools say, that one 'denied
history' They didn 't understand that [these spatial configura-
tions] meant the throwing into relief of processes - historical
ones, needless to say - of power.??
The ruling class (which broadly controls the institutions of the state
to serve its own interests), particularly in social formations where
the level of development of productive forces is low, employs a
hierarchical, space-leaping and relatively efficient system of chan-
nels of communication, as opposed to [the] exploited classes who
are forced to use a slow, contagious structure of communication,
which is unreliable for transmitting and receiving complex mes-
sages.
This ought to be put both more strongly and more subtly : all social
life goes forward within structural determinations, but these cannot
be understood in terms of a simple micro/macro distinction. Alth-
ough Yapa is thus correct to call for a model of biassedinnovation -
one in which non-adoption is not a passive state where the 'friction of
distance ' applies a brake to innovation, so to speak, but rather 'an
active state arising out of the structural arrangements of society'?" -
these ' arrangements' ought to be conceived as a hierarchical ensemble
of structures whose topographies mark out differential access to the
means of production and reproduction. This should not be taken to
imply an Althusserian trilogy of 'levels' and I have already rejected
the essentialism which this entails . Neither should it be assumed that
these ensembles are independent of human agency, since I take the
'causal powers' possessed by their constitutive structures to be exer-
cised through social struggles whose outcomes are never fully deter-
mined. If social structures are both conditions and consequences of
the social practices which are sustained through them, therefore, their
symmetries are always contingent. It may well be that, as Yapa notes,
'initial income inequalities usually are exacerbated by technological
innovation, leading to a circular and cumulative transfer of a dis-
proportionate share of the economic surplus from one class to
another': but the explication of these 'conditions and consequences'
does not involve the substitution of an ontology of structures for an
ontology of events.??
Class notations in any form are foreign to Hagerstrand's diffusion
theory, of course, but in so far as his time-geography is supposed to
incorporate a 'resource dimension ' it ought to be capable of respond-
ing to some of these criticisms. Certainly, Hagerstrand continues to
insist that 'one of the most problematical aspects of any project is to
get access to suitable information exactly when and where it is
needed'; but he also concedes that 'to see this process [of innovation]
as predominantly a question of communication' is to adopt a 'restric-
ted perspective' .80
In looking to the 'time-space mechanics of constraints', therefore,
Hagerstrand seems to be recognising the importance of differential
access - according to Blaikie, 'central' to any reformulation of dif-
fusion theory" - and rejecting his earlier framework in which 'in-
dicators [innovations] were avoided where it was obvious that econ-
omic or technical factors could impede their [diffusion] to a con-
siderable degree'. 82 But matters are not so straightforward. As I must
320 The Stasis of Diffusion Theory
Each project would seem to try , between its beginning and end, to
accommodate its parts, be they tangible or not, in the surrounding
maze offree paths and open space-times left over by other projects
or gained through competition with them . . . The interlocking of
projects ofdifferent life-span, up and down the hierarchy and between
hierarchies, is the central problem for analysis."
Conclusion
REALISM
E X PERI EN CE S
--8
--- --8 EMPIRICISM
SYS TEMS OF
SOC IA L
PRA CTI CES
ST RUCTUR ES OF
SOCIAL
RELATI ONS
Acknowledgements
Notes
49. For the statistical basis, see Mollison (1977); a convenient summary is
Cliff (1979).
50. Hagerstrand (1973) p. 70.
51. Gould (1981) pp. 1-3; cf. Hagerstrand (1976) pp. 332-3.
52. Williams, R. (1973).
53. Hagerstrand (1973) p. 82.
54. Parkes and Thrift (1980) p. 276.
55. Hagerstrand (1978) p. 124.
56. Hagerstrand (1970) pp . 10 and 19.
57. Carlstein (1982) p.44.
58. Hagerstrand (1973) pp. 80-1.
59. Hagerstrand (1970) p. 18.
60. Carlstein (1978a); cf. Carlstein (1982) p. 44.
61. Harvey, this volume.
62. For an account of the categorical paradigm, see Albrow (1974). Its
hallmark is the representation of change as a combinatorial game -
what Levi-Strauss once described as 'an endless play of combination
and recombination' of 'the same fundamental elements' - and Hager-
strand clearly endorses this viewthrough his own claim that 'the goal of
science' is 'to find the terms by which we can see change as a constant
permutation of a restricted number of invariants': see Gregory (1978b)
p.97; Hagerstrand (1974a) p. 52.
63. Hagerstrand (1970) p. 10.
64. Giddens (1981b) p.17.
65. Gregory (1982a) p. 194.
66. Hagerstrand (1973) p.85 ; Hagerstrand (1978).
67. Foucault (1980) p. 149. See also his 'Questions on Geography' passim,
pp. 63-77, in which 'domain' is presented as a 'juridico-political no-
tion', whereas Hagerstrand argues that 'the terrestrial domain can be
seen as a human elaboration of the animal territory': Hagerstrand
(1978) p. 143. There are, I think, definite continuities between Hager-
strand's proposals and Ratzel's much earlier discussion of the concept
of Lebensraum: see Gregory (forthcoming b).
68. Mayfield and Yapa, cited in Blaikie (1978) p.275.
69. Blaut (1977) p. 344.
70. Blaut (1977) p. 344; Blaut (1980) p. 28.
71. Cox (198Ia) p. 262; Boudon (1981)p. 86 also emphasises Hagerstrand's
methodological individualism.
72. Blaut (1980)p. 28. In an essay entitled 'From the Chicago School to the
Frankfurt School' Gouldner sketches out a series of connective themes
which seem to have escaped the notice ofmodem human geography: see
Gouldner (1976) pp. 118-37.
73. These three emphases are, of course, Giddens's characterisations of the
structural dimensions 'that are combined in differing ways in social
practices', Giddens (1979) pp . 81-2.
74. Braudel (1972) pp. 355-74; see especially Figures 28-30.
75. Giddens (1981b) pp. 169-81.
76. Blaikie (1978) p. 283.
77. Blaut (1977) pp . 347-8.
78. Yapa (1977) p. 359.
334 The Stasis of Diffusion Theory
79. Yapa (1977) p. 351. On the relations between 'an ontology of structures'
and 'an ontology of events' - and of the 'integrity' of these domains as
'differentiated features of social reality' - see Layder (1981).
80. Hagerstrand (1974) pp. II and 3.
81. Blaikie (1978) p. 274.
82. Hagerstrand (1967) p. II.
83. Hagerstrand (1975) pp. 10-11 ; Hagerstrand (1973) p. 86.
84. Hagerstrand (1970) p. II.
85. Hiigerstrand (1975) p. 5; van Paassen (1976).
86. Hagerstrand, cited in Carlstein (1982) p. 56.
87. Hagerstrand (1976) p.332; but Layder's discussion of 'interaction
structures' speaks directly to Hagerstrand's concerns: see Layder
(1981); Thrift (1983).
88. Hagerstrand (1974a) p. 53.
89. Carlstein (1978) p. 149.
90. Harvey (1973) pp. 22-49; Giddens (1979); Giddens (198Ib) . The rela-
tions between time-geography and the theory of structuration are com-
plex. On the one side, Pred and Carlstein are clearly sympathetic to
Giddens's project , but insist that this 'emerging consensus in social
theory' has, for the most part, failed to identify the 'cement' which
'binds' space and time into the production and reproduction of social
life (Pred, 1981a,p. 9), and that 'what has been done in time-geography
so far is absolutely central to any structuration theory of the kind
proposed by Giddens' precisely because it enables us 'to see its material
logic' (Carlstein, 1981 , p.48). On the other side, Giddens is plainly
drawn to the web model, whose 'general importance is that it emphas-
ises the co-ordination of movement in time and space in social activity,
as the coupling of paths or trajectories'; but he also claims that 'the
interest of this conception of social activity' does not depend upon
' Hagerstrand' s particular formulation of it, to which various objections
can be made' (Giddens , 1979, p. 205). Although Giddens does not state
these objections, I take them to be essentially theoret ical: the relevance
of these writings to his own project has much to do with Goffrnan's
(1959) accounts of 'the presentation of selfin everyday life', which draw
attention to the knowledgeability of social actors in ways which, as I
show below, Hagerstrand's formulations do not. The comparison
which van Paassen makes between Hagerstrand's theses and Elias's
'figurations' is more instructive in this regard than he intends: van
Paassen (1976) pp . 338-9; Elias (1978a). (Since completing this essay, I
have read Giddens 's contribution to the present volume, which seems to
confirm the interpretation I offer here.)
91. Thrift (1983) p.39fn. I have a number of reservations about this
proposal. First, even if these 'links' can be given substantive meaning -
which I doubt - they are clearly not confined to the labour process, and
their identification requires an inclusive theorisation of the 'anatomy'
of society. Second, I regard the connection between the compositional
and the contextual as a methodological one, which is bound in to the
relation between what I describe below as 'causal powers' and their
'fields of effectivity'.
Derek Gregory 335
..
"0c
...
"0c
c c
.. .
0 0 BIOGRAPHY FORMATION AND
SOCIALIZA TlON
E E
.
0
o gr--
.
Language acquisition
Personality development
III III
ESTABLISHMENT . Development of
REPRODUCTION AND consciousness
TRANSFORMATION OF
POWER RELA TlONS
'-- (STRUCTURE) TRANSFORMATION OF NATURE
I I
Occur simultaneously
/
r
-- -
-
_/ -
Jr ........ _ . . . . _
(j' 6
SKANE
Srrr ms ha mn
the right either to cultivate any of the numerous narrow strips in the
village's three fields or to sit on the village council that co-ordinated
and controlled major local agricultural projects.'?
As a group these villages had much in common in terms of the
agricultural, household, religious and festive projects through which
socialisation occurred and social relations between groups and the
sexes were reproduced. Certain common forms of agricultural co-
operation among adjacent villages and of kinship links with nearby
villages were also observable. In addition, there were case-to -case
similarities with regard to the nature of economic contacts occurring
between village residents and either urban businessmen or the rural
inhabitants of the forested and more inland areas of Skane.!"
However, each of the villages was a distinctly becoming place where
the details of biography-formation and everyday life varied some-
346 Enclosures and Social Change in Skane
formally enacted for Skane until 1803,a single large landlord, Rutger
Maclean, innovated this type of enclosure in 1783 when he began
replanning the four villages which comprised his estate on the south-
western plains.P Maclean rearranged the holdings of individual
peasants so that they no longer included from sixty to one hundred
land parcels, but instead were confined to a single rectangular farm
where fallowing was prohibited and a rotation of fodder roots,
barley, and clover or vetches was mandatory. This replanning not
only facilitated an increase in the number of peasant farmsteads, but
it also required the construction of new buildings and roads and was
synonymous with the breaking-up ofthe settlement cluster at the core of
each village. While the drastic re-ordering which resulted from
Maclean 's exerciseof power met with protest and resistance - half his
tenants revoked their contracts - it eventually brought considerable
economic success." By the 1790s other estate owners in Skane were
inspired by this success to enclose the villages under their control.
With the scattered incidents of peasant unrest that accompanied the
bad harvests at the turn of the century, and with the French Revolu-
tion still echoing, some later imitators of Maclean were also
motivated to break up the population concentrations of their estate
villages in order to minimise the potential for future disturbances."
According to the provisions of the 1803 enskifte decree, once a
request for enclosure was forwarded to the county governor, a com-
missioned surveyor supervised by specially appointed civil witnesses
was to work out as equitable as possible a reapportionment of
property and of any previously uncultivated common lands. Alth-
ough the configuration of new single-parcel farms resulting from
enskifte varied from case to case, the general result was the same in the
typical large village of the south-western and south-eastern plains.
The village core was quickly thinned out , for only a few peasant
homes persisted; many landless inhabitants remained but others had
their small houses torn down and were completely displaced; and the
great majority oflanded peasant households now conducted most of
their activities in relative isolation from one another (see Figure
14.3).25
At first there was widespread reluctance to undertake enskifte,
particularly because of the village disintegration which it required .
Although by 1812individual resistance to this form of enclosures had
not totally broken down, the number of peasants willing to initiate
the often unpopular procedure was clearly increasing in the face of
pronounced grain-price increases. (Between 1803 and 1812 the
• t v.>
t ..
~
\ / . [•• • 18 00
r··~:h~:~l._
1790 r.: ·~;:l: . I .!:t19 ' \ 1809
'00
o . '00
I • • I
fi~rfffI
r --114
f:;·V~
··~/~!.C..
<I" ..~lt
,
\: « .;
· ': <"':--1. , ,
METERS
:"113ti}}:::'@ ::i 22
.J......-ll j;.. t:l :.. 1.
i1'2l!.:'i:
:""""1i
I:.. ~I./~::;f~J1;· 4 :
~.-. :· "~" .·m. 223~ 'jE t)
,, _ ~ d~.
\ ' --- ,.•
•
.:.IIl!-'I'/I.
~ ~ .•. 'el
z.::..:.\."__~_ ~~ 25\!~Ii
'-I :~.:: : i l '- ~'f":-
r" l 91J:-:'" ::' i ' :-= " I '~ _26e.,
~ .~t.,j: tf+ \ VPJ=!
m
-. ···· -...L"'-"-",-:.:. JLlJi>::
• - ~ ";::':;' . __
-';m
/.
: . .i
r" .~ (~ 8"
i• ~ , "I." :'~~ " . "~:;:-~~": ...... :;.
\ " --:::l --~ ·:o. {311.
: 6 _or . ... . . . . . .,.... 10 ·.
! .\.
C:.~ :::::~~:: <--:,.;
.~'" :;:..
.......- \.Ll-t\
·~.:-= ~~ -:J . r" ·T · ";
: .;--<" .. : "32
••'"•. -··... .,......... 1•••
_J . ~ ~ ' . ... t,;
'"'
(" - ~~
' . ' ... r.....l•. >'~
' I' ;' . _
~;~ '~ .~\ . ~ '~" -:!
\/
~· 0 l
.)...~
~'t!"··'~·tm
'J••' "
L_\.. _f
FIGURE 14.3 The core of the village of Hyllie, south of Malmo, before 1790 and after 1809
enskifte. Solid bars indicate buildings occupied and used by households, each of which is represen-
ted by a number. Note church and vicarage at centre. Modifiedfrom Nordholm , Skanes aldre
ekonomiska geografi, I 121.
Allan Pred 349
average market price for rye, barley and oats in Skane more than
doubled.P" With the help of state subsidies to cover the various costs
involved, there were few villages remaining on the plains of Skane
that had not undergone enskifte by 1826.27 Thus, the spatial organisa-
tion of that area was little affected by lagaskifte, or the final form of
enclosures enacted the following year. 2 8
Informed questions
What were some of the basic components of the daily paths followed
by landed-peasant household members during the decades im-
mediately preceding enskifte enclosures on the plains of Skane? What,
in other words, were some of the general characteristics of the path-
project intersections that were part of biography formation and the
village-level unfolding of the structuration process? What were the
power relations underlying daily paths? What characteristics of prac-
tice-based consciousness predisposed some landed peasants to in-
itiate enclosure proceedings? How were daily paths and power rela-
tions altered as a consequence of enskifte enclosures? Or, to combine
all these questions: how did the social become the spatial and the
spatial become the social in the becoming villages of Osterlen and
south-western Skane?
Most of these questions cannot be answered in a definitive manner,
but there is a tremendous wealth of material available and what
follows reflects but a small fraction of the available evidence and the
reinterpretive arguments which can be developed from it.
hold was in motion from 3 or 4 a.m. until about 9 p.m., the day
usually began with a sandwich and a shot of aqua vitae before all the
males and some of the females left the house . Subsequent snacks and
meals, each accompanied by a swig or two, occurred at 6 a.m., 7.30
a.m., 10a.m., 12noon to 1p.m. (followed by a rest), 3 p.m., 5 p.m. and
sometime during early evening.'! When ploughing, sowing, harrow-
ing, fertilising, harvesting or other field-centred dominant agricul-
tural projects were undertaken, all but the last of these additional
repasts were eaten at or near the work-site . Since most field-strips
were highly elongated and at some distance from the homestead, it
follows that a large amount of time was taken up by women and
young boys bringing out food to the fields and, perhaps, by periodic
down-field movement toward the village centre.
At one level, major agricultural projects required considerable
daily path synchronisation and 'synchorisation' (or spatial co-ordin-
ation) among peasant household members as well as livestock and
implements . Spring and autumn ploughing, for example, required the
use ofanything from three to seven pairs of horses and oxen, depend-
ing on soil characteristics and the crop to be planted. The large
number of creatures needed to pull a plough in turn called for the
presence ofthree or more adult males and any younger boys or girls in
the household who might be available to help out or shout en-
couragement (Figure 14.4).32 Fertilising projects, with their heavy
labour inputs, constitute another good example. Manure was loaded
and transported to various dispersed field-parcels and immediately
spread out in late Mayor June, to be dug in some months later. In the
south-east, at least, participation in the initial phases was not con-
fined to household males and the draught animals and wagons neces-
sary for transportation. The daily paths of female servants and
daughters also had to be co-ordinated insofar as they were engaged in
loading and spreading. (Because of the slowness of wagon movement
over rudimentary roads and the volume of household time-resources
diverted from other projects, fertilising activities decreased in inten-
sity with increased distance from the village centre.P" .Harvesting
projects represented the extreme case of daily-path synchronisation
and synchorisation, normally involving all a household's men and
women (who were responsible for most of the raking, binding, and
stacking) as well as draught animals, wagons and various hand-held
implements.t"
Seen in isolation, the household execution and co-ordination of
ploughing, fertilising and harvesting projects, and attendant eating
Allan Pred 351
24 ~------+----rl-----;~~----(
20
18
12
Until relatively recently many Swedish scholars held that the enskifte
enclosures brought about an abrupt and fundamental change in the
consciousness and personality of landed peasants. Prior to enskifte
the landed peasants were supposedly 'completely free of individual-
ism', their freedom to undertake land-use or other changes fettered
either negatively by bytvang - the socially-based compulsion to co-
operate - or positively by a sense of community. After enskifte and
village disintegration, individualism supposedly blossomed among
landholding peasants, their personal initiative to innovate no longer
hindered by neighbours. One simple fact, however, indicates that the
historical border between the presence and absence of individualism
was not that clearly CUt.4 2
While enskifte was reportedly unpopular because of the costs
354 Enclosures and Social Change in Skane
As the land redistribution and new house, fence, and road construc-
tion required by enskifte enclosures was completed in any village, the
daily paths of landed peasant households and the population as a
whole were quickly altered - now stamped by new and modified
agricultural projects .
Many of the path changes were connected with the more intensive
cultivation of land that became necessary and feasible with the im-
plementation of enskifte. The moving and construction costs incurred
as holdings were consolidated forced some landholders to sell off a
portion of their new property units. In more exceptional instances
new landed peasant households were also formed because family
356 Enclosures and Social Change in Sktme
a)
Time
,---1---.
24
20 ~ J
r----
16 I
I
I
12 I
I
I
I
8 L _
b)
Time
24
20 ----~~~~===========~
I-
I
16 I
I
I
12
I
I
8 I
I
L
4
O'----~~---------c======:::r-
DWELLING
AND THINHED OUT
FARMYARD VILLAGE CORE
L---J
CONSOLIDATED
HOLDING
- - - Actual path
Potential time-space reach in village core and
fields
o 1000 METERS
'--------',
COD(E)A
Notes
1. For critical comments on the treatment ofplace and sense of place by the
'new humanistic' geographers, see Cosgrove (1978); Hudson (1979);
Sayer, A. (1979a); Ley (1981); Pred (1983).
2. For other comments on places, regions and spatial organisation as a
social product, see Soja (1980); Urry (1981c); Gregory (1981; 1982b).
3. Giddens (1979) pp. 106-7; Giddens (1981b) pp. 39,45.
4. Pred (1984); see also Pred (1981c; 1981a; 1983).
5. There are a considerable number of authors who may be linked with the
theory of structuration. The particular set of works influencing my own
theoretical formulations are as follows: Berger and Luckmann (1967);
Giddens (1976; 1979; 198Ib); Kosik (1976); Bhaskar (1978a, 1979);
Bourdieu (1977);Touraine (1977);Williams (1977); Shotter (1983). Note
also the lucid introduction in Gregory (1982b).
6. A more detailed discussion of the language of time-geography will be
found in Hiigerstrand (1970; 1974); Pred (1973; 1978);Parkes and Thrift
(1980); Carlstein (1982); Gregory, this volume.
7. For all its advantages, the Vidalian tradition is not without its limitations
as a foundation for conceptualising places and regions in processual
terms: see Gregory (1981); Pred (1984).
8. Although social theorists and philosophers define and interpret the
elusive concept of power as a social relation in often disparate ways, the
attributes specified here are normally explicitly or implicitly present.
9. Note related arguments about 'time- space distanciation' on the one
hand and social integration and system integration on the other in
Giddens (1981b).
10. Once a person makes a commitment to partake in any project at a given
time and site it becomes impossible for them: (i) simultaneously to do
something else somewhere else; (ii) to join any other spatially separate
project which starts at another time but temporally overlaps with some
portion of the project in question ; or (iii) to join another project that
presents no simultaneity conflicts but which is out of 'reach' because of
travel-time requirements. Those unfamiliar with time-geography will
find that these constraints are more readily understandable when depic-
ted diagrammatically: see Lenntorp (1976); Martensson (1979).
11. Note comments on unknowing in Thrift (1983) and this volume.
12. Cf. Hagerstrand (1977); Carlstein (1982).
13. Cf. Hiigerstrand (1974a).
14. Gregory (l982a) pp . 188-9.
15. von Linne (1959 [1751]) pp.174, 177, 197,238.
16. Dahl (1942) pp. 13-16 and map facing p.16; Hanssen (1952) pp. 51-7.
Although Dahl's data on village density are based on mid-seventeenth
century sources, there was very little change in the number of settlements
before the mid-eighteenth century.
17. For a much fuller and more nuanced account of the social composition
of pre-enclosure villages on Skane's fertile plains, see Hanssen (1952)
pp.48-61; Hanssen (1976) pp.33-61. For related observations on
Allan Pred 363
55. von Linne (1959 [1751]) p.269; Hallenborg (1910-13) p.345 ; Barck
(1904-8) p.373.
56. Nicolovius (1957 [1847]) p. 20; Dahl (1942) p. 207; Bringeus (1971) p. 46.
57. Frykman and LOfgren (1979) p. 179; Frykman (1977) p. 197; Hellspong
and Lofgren (1974) pp. 81, 138; Lofgren (1974) p. 31; (1975) pp. 7,241,
28; (1978) p. 103.
58. Thunell (1952) p. 47; Frykman and LOfgren (1979) p. 30.
59. Fridlizius (1957) p.59; Utterstrom (1957) p. 540.
60. Bringeus (1964) pp. 70-1 , 74; Frykman (1977) pp. 112-16.
61. Wigstrom (1891) pp . 14, 23; Fredin (1946) p. 198. See also note 10.
62. Asplund (1983); Frostin (1932) pp. 9, 11; Bruzelius (1978 (1876]) pp . 59,
73; Frykman (1977) p. 131 ; Cf. Pred, 1983.
63. Engstrom (1928)p. 16;(1928a) p. 34; Frostin (1932)p. 11; Thunell (1952)
p.53 .
64. Bruzelius (1978 [1876]) pp . 37, 40; Hanssen (1952) pp. 73-84.
65. Hanssen (1952) pp. 352,362-82; von Linne (1959 [1751]) pp .183 , 203,
249,267,318-19; Hallenborg (1910-13) pp. 300, 318, 340, 359; Anckar-
svard (1910-13) p. 82; Campbell (1928) pp. 19-23, 44ff.
66. For example, Hanssen (1952) p.85; Berg and Svensson (1934) p. 165;
Frykman (1977) p. 123.
67. Areskoug (1949) p.83; Hanssen (1952) p.88; von Linne (1959 (1751])
pp.82, 96, 252, 271, 290, 300, 308, 385; Ejder (1969) pp.318, 405;
Bringeus (1971) pp . 51, 54.
15
Flies and Germs: A Geography of
Knowledge
NIGEL THRIFT
Introduction
This little vignette in which the child uses what knowledge is available
to her to interpret her mother's interdiction, illustrates the central
theme of this chapter, namely that what we know is an important
constraint on what we can think and do. It also illustrates two other
important themes which grow out of and into this central theme . The
first is that knowledge is historically specific. Thus , before 1871,
knowledge of 'germs' did not exist." The mother could not have
known of their presence and the child would not have had to answer
her interrogator. Second, knowledge is geographically specific. Even
now, the knowledge of germs does not stretch everywhere in the
world and in these nescient places mothers do not have to tell their
children something they cannot know .
This chapter thus forms part of a continuing attempt to develop a
theory of situated social action . Such a theory must continually
intersect with both the presences and absences of social structure.'
and with the continual turmoil of social groups in conflict in a
complex process of'structuring'4 but I would argue that it still forms
a relatively separate focus of attention. 5 Elsewhere, I have outlined
the three most important elements of this project, which are the study
of the process of personality formation, the study of the organisation
of sociability and the study of the availability of knowledge (where
Nigel Thrift 367
'Social knowing' is the term I will use to denote the continual process
of the creation of the stocks of knowledge" within a society upon
which actors, who are always members of various social groups, can
draw in the production of their life and the reproduction or transfor-
mation of the forms oflife ofsocial groups (their own and others) and
ultimately of society.9 I see this process as being an essentially recur-
368 A Geography of Knowledge
sive one, the result of practices that both provide the medium for and
the outcome of social structure.to
ACTOR 1. ACTOR 2 .
UNKNOWN
/
/ /
L----r--~------~----
/ / /
/
/
/
/
/
/
WHAT IS 'KNOWN '
Types of knowledge
Unconscious knowledge
Unconscious knowledge is based upon forgotten practices still re-
membered in the limning of action . As Bourdieu puts it, 'the uncons-
cious is never anything other than the forgetting of history which
history itself produces by incorporating the objective structures it
produces in the second natures of'habitus'." Interpreting the uncons-
cious as a historically changing and geographically variable phe-
nomenon, let alone the variations in the type of knowledge it
provides, is the domain of a historical and geographical psychology
that as yet hardly exists.-"
Practicalknowledge
Practical knowledge can be defined as that informal (but not
therefore unstructured) type of knowledge that is learnt from the
experience of watching and doing in highly particular contexts in
direct mutual interaction. It still forms 'the massive central core of
human thinking' that goes unrecorded in so many histories of social
thought ."
Practical knowledge has four major components. First, it is unar-
ticulated; that is, it is based upon practices that have, over the course
of history, become naturalised so that much of its content now
inhabits the realm of the undiscussed and , certainly, the uncodified.
To have to ask about practical knowledge is to miss the point of
practical knowledge." Second, practical knowledge is part of a con-
tinuous and repetitive flow of conduct which takes place in finite time
and which is oriented towards doing. This is a world in which 'I think'
(cogito) is inseparable from 'I can' (practico) and in which long-term
goals are submerged by immediate objectives. Third, practical
knowledge is local, that is, it is knowledge produced and reproduced
in mutual interaction that relies on the presence ofother human beings
on a direct, face-to-face basis." Such knowledge is deeply imbued
with both historical and geographical specificity,taking its cues from
local contexts each with their own particular ensemble of practices
and associated linguistic usages. Nowhere is this made clearer than in
historical accounts of local communities which include glossaries of
terms that describe practices (and the experience of those practices)
now long past. These are words which not only capture the attributes
of a particular practice but also begin to conjure up the ghost of the
374 A Geography of Knowledge
The Kuranko of Sierra Leone use the word kile ('path' or 'road') as
a metaphor for social relationship. For instance, the adage nyendan
bin to kile a wa ta an segi describes the way a particular species of
grass (used for thatching) bends back one way as you go along a
path through it, and then bends back the other way as you return
along the path. This movement to and fro of grass along a pathway
is used as a metaphor for the movement of people and of goods and
of services within a community; it is a metaphor for reciprocity.
Thus, in Kuranko one often explains the reason for giving a gift,
especially to an in-law, with the phrase kile ka nafaga, 'so that the
path does not die'. However , if relat ions between affines or neigh-
bours are strained, it is often said that 'the path is not good between
them' (kile nyuma san tema), and if a person disappoints a friend
then people may comment a ma kile nyuma tama a boma ('he did
not walk on the good path with his friend')."
Dutt lists a few of the beliefs or the omens of Bihar with respect to
insect and pest attacks on their crops . In particular, he notes that
there are many farmers who believe that certain caterpillars fall
from the sky with the rain, and their cure for the attack is accordin-
gly usually religious. The reason for the beliefis that the caterpillars
emerged from their eggs in cloudy, rainy conditions, and their life-
cycle is not understood. That man is born of woman, and a calfof a
Nigel Thrift 375
Empirical knowledge
The stock of empirical knowledge is built up as a result of the general
process of rationalisation of knowledge," which has a double inter-
pretation as the proffering ofa rational explanation and the organisa-
tion of knowledge in a systematic fashion. Empirical knowledge, like
practical knowledge (with which it continues to share many
similarities), is bent towards the mastery of the conditions of exis-
tence, but it is exercised within a learning process which is not only
cumulative but systematised and co-ordinated over large tracts of
space and over longer time-horizons , particularly by modern state
and economic institutions. The modern state is based upon the
various practices associated with surveillance (that is, the accumula-
tion of information on the population, the supervision of the popula-
tion and the characterisation of the population in such a way that it
can be supervised) and proceeds from the institution of regular armies
and the systematic registration ofbirths and deaths via the census and
fingerprinting to modern computer data-banks. Similarly, the
capitalist economy is based upon the co-ordination of the labour
process and the complex exchange of commodities, both of which
require the extensive collection of systematic knowledge, and
proceeds from double-entry bookkeeping via modern accounting
practices to - again - the computer. In both bases, the institutions
engendered by these practices produce a network of space-time
locations at which knowledge is stored, received and transmitted. The
schooling system forms a particularly crucial set of nodes within this
network."
What, then, are the major characteristics of empirical knowledge?
Above all, it is heterodynamic." that is, it is acquired by virtue of an
actor's membership of class and other social groups , and it is distan-
ciatedr" that is, it is removed in both time and space from the
experiences and events it describes. Empirical knowledge does not
depend for its acquisition upon the direct presence of people, but is
transmitted through institutions and technologies which allow per-
376 A Geography of Knowledge
Natural philosophy
The third type of knowledge (empirical knowledge) overlaps with
knowledge gleaned from 'natural philosophy'. I use 'natural philoso-
phy' in the most catholic ofsenses, to denote knowledge that attempts
to unify a number of bodies of knowledge into one whole, as
knowledge about knowledge." It has at least three interrelated
characteristics. First, it requires much time both to absorb a number
of bodies of knowledge and to synthesise them." As Kierkegaard put
it, 'Life can only be understood backwards but it has to be lived
378 A Geography of Knowledge
Almost every English village had its 'cunning man', its white
magician who told those who had been robbed how to recover their
property, advised on propitious times and seasons for journeys and
foretold the future . Nor were such beliefs limited to ignorant
villagers."
But, at the same time as this, there existed another magic in England,
one enjoying a 'renaissance' in the universities and at court thanks to
the discovery in the fifteenth century of texts attributed to Hermes
Trismegistus.v The Hermetic tradition, meant to be based on the
magical religion of the Egyptians , had spread rapidly (at least in part
Nigel Thrift 379
The example of the rise and fall of magic illustrates two major
themes. First, the connections between practical knowledge, natural
philosophy and, later, empirical knowledge are by no means one-
way : from natural philosophy to empirical knowledge and so on
down to the person in the street. For example, practical knowledge
can intervene in the building of empirical knowledge or natural
philosophy, and this can happen in the most unexpected ways . Thus:
Further:
In everyday life these types of knowledge are, more often than not,
unproblematically fitted together by each actor in an ongoing
bricolage, closeted away in the stream ofeach person's daily conduct.
This is not to say there are no problems, however; and these are
compounded by variations in the availability of printed knowledge
over space and by social group.
:
,)0'1 54.
<? .. :• ..••• •
.•4.5 • .. •
•
•
.... ....
• • ••
.• ; · ·':60 - • • • .. .
~ :
) • • ••• • • ·IO~.OON 1
'"Z-' C:-~ . . . 92 .
• 134
•••
NUMBER Of CHAPMEN
•
309 20 9 10 19
•
~9
3' 2 I
• • s.utl',, "pf IOolS L....--...J 100•• ,
The first four editions were luxurious folio publications that only a
few could have afforded and , when taken together, account for only
40 per cent of the Encyclopedies in existence prior to 1789. The great
mass of the Encyclopedies sold came from the cut-rate quarto and
octavo editions printed between 1777and 1782. Of those produced in
France some 50 to 65 per cent were quartos and all these subscriptions
can be traced, at least to the booksellers , so allowing Figures 15.4and
15.5 to be drawn."?
The class-specific nature of these patterns can be elicited by the
example of the town of'Besancon.?" In some respects Besancon does
not seem the most fertile ground for the Encyclopedic - it was a
provincial capital based upon religion and administration, a veritable
bastion of Bourbon bureaucracy; but it also had a public library, a
cabinet litteraire and four booksellers . Literacy rates were high (95
per cent for men, 60 per cent for women). In the event, the Ency-
clopedie sold well- to noblemen, to the military, to parliamentarians,
to doctors and to lawyers. In contrast, the artisans, shopkeepers, day-
labourers and servants who made up the other three quarters of
Besancon's population do not appear at all amongst the subscribers,
nor do the peasants and shopkeepers of the surrounding province of
Franche-Comte. Certainly some of these may have consulted the
Encyclopediein the reading clubs of the area, but for the vast majority
access to its knowledge was severely limited - by a whole host of
closely interwoven factors such as distance, class, time and ability to
read . Indeed, in France as a whole:
• e •
!>1l0
100
e • •
100
10 10 10
ll,IbctCrlpll0n l l - J 100\,.,
lCJ~
:.J'>?\ ~~~
Moscow.
arnbu,'
~ Lunden . The Netherlands • Warsaw
~
• ~r~eIS and Ypres
.. The Rhineland p
Liege • • rague
SWitzerland • MUnich
- Pe"
_ . Milan Venice
Turin ' M
:': \ :
8~
e:::J
••
100
;0
30 10 ) .431,1 ~oo ~m,
• • • 0 subm ophon1
before the nineteenth century but which now become different facets
of the same process of mobilisation of social groups around various
specialised stores of knowledge . The 'constituency' of the various
mass media is strongly differentiated by class and other social group-
ings to such a degree that the actual fund of knowledge common to all
social groups (and likely to be spatially homogeneous) may well be
quite small; the same point applies to mass schooling systems." The
first tendency is the explosion in the overall stock of knowledge in the
nineteenth century as a result of the growth of the state, the capitalist
economy and the various sciences which brings with it specialists in
particular stocks of knowledge . The second tendency is the seizure by
various social groups upon this growing body ofknowledge (over and
above that knowledge which becomes common to all or most of the
population) as a fertile source of 'cultural capital'82 with which to
differentiate themselves from other social groups and thereby gain
economic and social advantages. (Thus the education system
provides the grounds of heterogeneity as well as homogeneity.)
The result is that spatial variation in the distribution of empirical
knowledge still exists and may even be relatively stronger than in the
past, but is increasingly tied to the social distribution of empirical
knowledge in a pattern of sequestered life-spaces." Thus, the social
distribution ofempirical (and practical) knowledge is associated with
institutional nodes like home, school, university or officewhich form
a set of points that selectively channel the life-paths of actors accord-
ing to their membership of a particular social group." This channell-
ing results in the acquisition of particular common kinds of
knowledge (and the limits on that knowledge) that ultimately ensure
the reproduction of that group as a socio-spatial entity . Furthermore
the organisation of these nodes into distinctively sequestered life-
spaces occurs at different scales. Working-class life-spaces are even
now predominantly local: a local school is followed by a local job.
Middle class life-spaces are more spatially extensive: a local school
may be followed by a university in a different location and then a job
somewhere else again.
These claims can be made more concrete by another example from
the eighteenth century, a study of two particular doctors, James Clegg
(1679-1756) and Richard Kay (1716-51).85 They were members of
the same social group, solidly Presbyterian and middle class, and
although they lived out their lives some kilometres apart they shared
common interests, had friends in common'" and had heard sermons
Nigel Thrift 389
preached by the same people. Indeed, they may even have met. Their
respective life-paths and the daily-paths they rode and trod for a week
in July 1745, are depicted in time-geographic fashion in Figure 15.6.
Immediately the common nodes of this particular social group's form
of life stand out - home, school, chapel.
What can we say about the stocks of knowledge upon which Clegg
and Kay drew? First, they would have had a common pool of natural
philosophy. Their upbringing was, in both cases, based upon a num-
ber of religious schools and dissenting academies. Second, they would
both have had a common store of practical knowledge, a closed
homo dynamic code founded in and naturalised by the localised
routines offamily, farm and chapel. Third, they would both have had
access to much the same supply of empirical knowledge - from
school, from meetings of the Book Society and philosophical lectures
and from various books and journals." In particular, of course, they
would have shared much of the same medical knowledge which was
primitive but becoming more systematic as physiology and anatomy
slowly developed." Their diaries show they were well-versed in draw-
ing blood, applying leeches and glysters, setting broken limbs and
removing tumours. But there is one important change between
Clegg's life-path and that of Kay . Clegg was essentially an ordained
Presbyterian minister with a small farm who practised medicine to
make ends meet, a practice common in remote rural areas of England
but one that was already dying out in his lifetime. His medical
knowledge was picked up by some reading, of course, but also
through apprenticing himself to another doctor in Macclesfield. His
right to practise without a licencewas guaranteed by a medical degree
obtained in absentia from Aberdeen on the recommendation of other
(Presbyterian) doctors. In some respects Clegg was an 'amateur'
doctor and the localised life-space of his life-path reflects this fact. In
contrast, Kay had a year's formal training at Guy's Hospital in
London, and his diary makes it clear that medicine was his first
priority and that his living came from the practice of it. Here we see
the beginnings of the formation of the medical profession as we now
understand it, of a segment ofa new professional middle class and of a
new social distribution of empirical knowledge with the attendant
qualifications serving as its cultural capital. We can also observe the
corresponding change in Kay's life-space, when compared with that
of Clegg - only a year in London perhaps, but soon to burgeon into a
much longer period of formal training.
390
'"'
""
li lt
LONDON .
,60
oral to a print culture, with print still only 'one of rather than 'the
major' means of disseminating and receiving political knowledge.
Interpretation of print is concrete and literal. This is not so surprising.
In England:
the ability to read was only the elementary technique. The ability to
handle abstract and consecutive argument was by no means
inborn; it had to be discovered against almost overwhelming dif-
ficulties - the lack of leisure, the cost of candles (or spectacles), as
well as educational deprivation. Ideas and terms were sometimes
employed in the early Radical movement which, it is evident, had
for some ardent followers a fetishistic rather than rational value.
Some of the Pentridge rebels thought that a 'Provisional Govern-
ment' would ensure a more plentiful supply of 'provisions'; while,
in one account of the pitmen of the north-east in 1819, 'Universal
Suffrage is understood by many of them to mean universal suffer-
ing' . . . 'if one member suffers, all must suffer'.I02
twentieth century when , for example, other media than print come
into play (each with their own distinctive ways of disseminating,
rationing and biasing knowledgej.l'" when mass schooling systems
guarantee a print culture's existence and when organised political
parties hold sway. These changes are not within the scope of this
chapter. But at the most general level, the difference between the
nineteenth and twentieth centuries are probably not as great as might
at first be assumed. In most social groups political commitment
remains diffuse, those who read political books remain in a minority,
those who know about politics and political institutions are relatively
few.106 It can be safely assumed that political knowing and unknow-
ing remains differentially distributed amongst social groups, over
space and in time.'?"
Conclusions
Four main threads can be teased out from this chapter. First, the
availability of knowledge must be seen as an important component in
the construction ofa theory of situated social action. It is not the only
component; it is probably not the most important component, but
neither is it a component that can any longer be passed over .
Ignorance may be no excuse in the eyes of the law, but it is a fact of
social life that can no longer be overlooked. Second, the sociology of
knowledge can be given a more coherent base, since it is now general-
ly agreed that:
I hope that this chapter goes some way towards beginning to over-
come this difficulty. Third, epistemology must be seen for what it is.
Thus, it:
Notes
CORRESPONDING
SPATIO - TEMPORAL CORRESPONDING
FORM OF MANIFESTATION TIME' SCALE
LIFE · PATH
I PRACTICES
I DUREE OF
DAY · TO · DAY LIFE
100. Other libraries and other sources of political knowledge also grew up at
the time, of course, although usually of a non-radical nature; for
example, those of the Mechanics Institutes, the Society for the Dif-
fusion of Useful Knowledge, chapel libraries, even a few school
libraries.
101. Vincent (1981) p. 118.
102. Thompson (1968) p. 783.
103. Calhoun (1982) ch. 8.
104. Foster (1974) p. 147.
105. See reviewsin Barnett, Corrigan, Kuhn and Wolff (1979); Swingewood
(1977); Corrigan and Willis (1980).
106. Dennis, Lindberg and McCrone (1971).
107. For books, see Mann (1970); Mann and Burgoyne (1969).
108. Abercrombie (1980) p. 176.
109. Barnes (1982) p. 93.
110. One seeming gap in this essay is the work of Foucault, but this is a result
of his lack of emphasis on practices.
111. Thrift (1979). See Bhaskar's (1979, 1980) comments on the activity-,
concept- and space-time dependence of theories.
Bibliography
(Place of publication of books is London unless otherwise stated .)
trial Societies.
Carlstein, T., Parkes, D. and Thrift, N. (eds) (1978) Timing Space and
Spacing Time (3 vols).
Carney, J., Hudson, R. and Lewis, J. (eds) (1980) Regions in Crisis.
Carter, H. (1981) The Study of Urban Geography.
Castells, M. (1971) La Question Urbaine (Paris).
Castells, M. (1976a) 'Is There an Urban Sociology?' in Pickvance (ed.)
(1976).
Castells, M. (1976b) 'Theory and Ideology in Urban Sociology' in Pickvance
(ed.) (1976).
Castells, M. (1977) The Urban Question.
Castells, M. (1978) City , Class and Power.
Castells, M. (1983) 'Crisis, Planning and the Quality of Life: Managing the
New Historical Relationships between Space and Society' Environment
and Planning D: Society and Space, vol. 1, pp.3-21.
Castoriadis, C. (1971) History and Revolution.
Cawson, A. and Saunders, P. (1983)' Corporatism, Competitive Politics and
Class Struggle' in King (ed.) (1983) pp.I-27.
Chapman, G. P. (1983) 'The Folklore of the Perceived Environment in
Bihar', mimeo.
Charlesworth, A. (1979) Social Protest in a Rural Society (Norwich).
Chaytor, H. J. (1945) From Script to Print (Cambridge) .
Chisholm, M. and Oeppen , J. (1973) The Changing Pattern ofEmployment:
Regional Specialisation and Industrial Localisation in Britain.
Chorley, R. J. (ed.) (1973) Directions in Geography.
Christaller, W. (1966) Central Places in Southern Germany (Englewood
Cliffs, New Jersey).
Clanchy, M. T. (1979) From Memory to Written Record: England /066-1307.
Clark, G. and Dear , M. (1984) State Apparatus: Structures and Language of
Legitimacy.
Clark , G . and Gertler , M. (1983) 'Migration and Capital', Annals of the
Association of American Geographers, vol. 73, pp. 18-34.
Clark, P. A. (1982) 'A Review of Theories of Time and Structure for Organ-
isational Sociology', University of Aston Management Centre, Working
Paper 24.
Clarke, J., Critcher, C. and Johnson, R. (eds) (1979) Working Class Culture:
Studies in History and Theory.
Clarke, S. (1977)'Marxism, Sociology and Poulantzas' Theory of the State',
Capital and Class, vol. 2, pp . 1-31.
Claval, P. (1977) 'Le Marxisme et l'espace', L 'Espace Geographique, vol. 3,
pp.145-64.
Cliff, A. D. (1968) 'The Neighbourhood Effect in the Diffusion of Innova-
tions', Transactions of the Institute of British Geographers, vol. 44.
pp.75-84.
Cliff, A. D. (1979) 'Quantitative Methods: Spatial Diffusion', Progress in
Human Geography, vol. 3, pp. 143-52.
Cliff, A. D., Haggett, P. and Versey, G. R. (1981) Spatial Diffusion: an
Historical Geography of Epidemics in an Island Community (Cambridge) .
Cohen, G. A. (1978) Karl Marx's Theory ofHistory: a Defence (Oxford and
Princeton) .
Bibliography 409
Daunton, M. (1981) 'Down the Pit: Work in the Great Northern and South
Wales Coalfields, 1870-1914', Economic History Review, vol. 34,
pp.578-97.
Davis, M. (1980)'Why the US Working Class is Different', New Left Review,
vol. 123, pp.3-46.
Dawley, A. (1976) Class and Community: the Industrial Revolution in Lynn
(Cambridge, Massachusetts).
Day, G. et al. (eds) (1982) Diversity and Decomposition in the Labour Market
(Farnborough).
Dear, M. and Scott, A. J. (eds) (1981) Urbanisation and Urban Planning in
Capitalist Society.
Dennis, J., Lindberg, L. and McCrone, D. (1971) 'Support for Nation and
Government among English Children', British Journal ofPolitical Science,
vol. 1, pp. 25-48.
de Schlippe, P. (1956) Shifting Cultivation in Africa : the Zande System of
Agriculture.
Doe, V. S. (ed.) (1978) The diary of James Clegg of Chapel-en-le-Frith
1708-1755 (2 vols) (Matlock).
Doeringer, P. B. and Piore, M. F. (1971) Internal Labour Markets and
Manpower Analysis (Lexington, Massachusetts) .
Duncan, J. (1980) 'The Superorganic in American Cultural Geography',
Annals of the Association of American Geographers, vol. 70, pp. 181-98.
Duncan, S. (1974) 'The Isolation of Scientific Discovery: Indifference and
Resistance to a New Idea', Science Studies, vol. 4, pp. 109-34.
Duncan, S. and Goodwin , M. (1982) 'The Local State and Restructuring
Social Relations', International Journal of Urban and Regional Research,
vol. 6, pp. 157- 86.
Durkheim, E. (1964) The Division of Labour in So ciety (London and New
York) (first published 1893).
Engels, F (1969a) The Condition ofthe Working Class in England (St Albans).
Engels, F. (1969b) 'The Housing Question' in Marx and Engels (1969) vol. II.
Engstrom, B. (1928) 'Aldermansskifte och Bystiimma i Skiinska och Danska
Byar' , Fataburen, pp. 18-39.
Engstrom, B. (1928) Byalagen i Bara hiirad (Lund).
Ericksen, E. G. (1980) The Territorial Experience (Austin).
Erixon, S. (1934) 'Svenskt Byliv', Sjiilvstyrelsen i svenskt samhiillsliv,
pp.35 -45.
Erixon, S. (1956) 'Bebyggelsestruktur och Bysamfiilligheti Sverige', Nordisk
Kultur, vol. 13, pp.275 -294.
Erixon, S. (1978) Byalag och Byaliv (Stockholm).
Eyles, J. (1981) 'Why Geography Cannot be Marxist: Towards an Under-
standing of Lived Experience', Environment and Planning A, vol. 13,
pp.1371 -88.
Nagel, E. (1955)'On the Statement "The Whole is Greater than the Sum ofits
Parts" in Lazarsfeld and Rosenberg (eds) (1955) pp. 519-27.
Needleman, L. (ed.) (1968) Regional Analysis (Harmondsworth).
Neisser, H. (1965) On the Sociology of Knowledge (New York) .
Nerlich, G. (1976) The Shape of Space (Cambridge).
422 Bibliography
Paci, M. (1981 a) 'Class Structure in Italian Society' in Pinto (ed.) pp. 206- 22.
Paci, M. (1981b) 'Internal Migrations and the Capitalist Labour Market' in
Pinto (ed.) pp. 33-46. :
Pahl, R. (1968) 'The Rural-urban Continuum' in Pahl (ed.) (1968)
pp.263-97.
Pahl, R. (ed.) (1968) Readings in Urban Sociology .
Pahl, R. (1975) Whose City? (2nd edn).
Pahl, R. (1980)'Employment, Work and the Domestic Division of Labour',
International Journal of Urban and Regional Research, vol. 4, pp. 1-20.
Pahl, R. (1981) 'Employment, Work and the Domestic Division of Labour'
in Harloe and Lebas (eds) pp. 143-63.
Pahl, R. and Dennett, J. (1981) 'Industry and Employment on the Isle of
Sheppey', HWS Project Report, University of Kent at Canterbury.
Pahl, R. and Gershuny, J. I. (1979) 'Work Outside Employment: Some
Preliminary Speculations', New Universities Quarterly, vol. 34,
pp.120-35.
Pahl, R. and Wallace, C. (1983)'Household Work Strategies and the Polar-
isations and Divisions of Work Practices' in Redclift and Mingione (eds)
(1983).
Palm, R. and PrOO, A. (1978) 'The status of American Women: a Time-
geographic view' in Lanegran and Palm (eds) (1978) pp. 99-109.
Park, R. (1952) Human Communities (New York).
Parkes, D. and Thrift , N. (1980) Times. Spaces and Places: A Chronogeogra-
phic Perspective (New York).
Bibliography 423
Parsons, T., Bales, R. and Shils, E. (1953) Working Papers in the Theory of
Action .
Pelling, Ha. (1967) The Social Geography of British Elections.
Penn , R. (1982) 'The Contested Terrain: A Critique of R. C. Edwards'
Theory of Working Class Fractions and Politics' in Day (ed.) (1982)
pp .93-106.
Perry, D. and Watkins, A. (eds) (1978) The Rise ofthe Sunbelt Cities (Bever-
ley Hills).
Pickvance, C. (ed.) (1976) Urban Sociology: Critical Essays.
Pickvance, C., (1981)'Policies as Chameleons: An Interpretation ofRegional
Policy and Office Policy in Britain' in Dear and Scott (eds) (1981)
pp.231-65.
Pinto, D. (ed.) (1981) Contemporary Italian Sociology (Cambridge) .
Piore, M. (1979) Birds of Passage (Cambridge).
Pollard, A. (1980) 'Teacher Interests and Changing Situations of Survival
Threat in Primary School Classrooms ' in Woods (ed.) (1980) pp. 34-60.
Popper, K. (1972) Conjectures and Refutations: The Growth of Scientific
Knowledge.
Poster, M. (1975) Existential Marxism in Postwar France (princeton).
Poster , M. (1979) Sartre 's Marxism .
Poulantzas, N. (1975) Classes in Contemporary Capitalism.
Poulantzas, N. (1978) State, Power, Socialism.
Pred, A. (1973) 'Urbanisation, Domestic Planning Problems and Swedish
Geographic Research', Progress in Geography, vol. 5, pp . 1-76.
Pred, A. (1977) 'The Choreography of Existence: Comments on Hager-
strand's Time-geography and its Usefulness', Economic Geography,
vol. 53, pp . 207-21.
Pred, A. (1978) 'The Impact of Technological and Institutional Innovations
on Life Content: Some Time-geographic Observations', Geographical
Analysis, vol. 10, pp. 345- 72.
Pred, A. (1981a) 'Social Reproduction and the Time-geography of Everyday
Life', Geograjiska Annaler B, vol. 63, pp. 5-22.
Pred, A. (198Ib) 'Power , Everyday Practice and the Discipline of Human
Geography' in Pred (ed.) (1981) pp. 30-55.
Pred, A. (ed.) (1981c) Space and Time in Geography: Essays Dedicated to
Torsten Hiigerstrand, Lund Studies in Geography, Series B, No . 48.
Pred, A. (1983)'Structuration and Place: On the Becoming of Sense of Place
and Structure of Feeling' Journalfor the Theory ofSocial Behavior, vol. 13,
pp .45-68.
Pred, A. (1984) 'Place as Historically Contingent Process: Structuration and
the Time-geography of Becoming Places', Annals of the Association of
American Geographers, vol. 74, pp .279-97.
Pred, A. (forthcoming) Becoming Places, Practice and Structure : the Emer-
gence and Aftermath of Enclosures in the Plains Villages of Southwestern
Sktine 1750-1850 (Cambridge).
Price, R. (1983) 'The Labour Process and Labour History', Social History,
vol. 8, pp.57-75.
Przeworksi, A. (1977) 'Proletariat into a Class: The Process of Class Forma-
tion from Karl Kautsky's The Class Struggle to Recent Controversies',
Politics and Society, vol. 7, pp .343-401.
424 Bibliography
Saville, J. (1960) 'Trade Unions and Free Labour: The Background to the
TaffVale Decision' in Briggs and Saville (eds) (1960) pp.317-50.
Sawers, L. and Tabb, W. K. (eds) (1984) Sunbelt/Frostbelt: Urban Develop-
ment and Regional Restructur ing (New York).
Sayer, A. (1979a) 'Epistemology and Conceptions of People and Nature in
Geography', Geoforum, vol. 10, pp. 19-44.
Sayer, A. (1979b) 'Theory and Empirical Research in Urban and Regional
Political Economy: A Sympathetic Critique', University of Sussex, Urban
and Regional Studies Working Papers, no. 14.
Sayer, A. (1981)'Abstraction: A Realist Interpretation', Radical Philosophy,
vol. 28, pp. 6-15.
Sayer, A. (1982) 'Explanation in Economic Geography: Abstraction versus
Generalisation', Progress in Human Geography, vol. 6, pp.68-88.
Sayer, A. (1984) Method in Social Science: A Realist Approach.
Sayer, D. (1979) Marx 's Method. Ideology, Science and Critique in 'Capital'
(Brighton).
Schofield R. S. (1968) 'The Measurement of Literacy in pre-industrial
England' in Goody (ed.) (1968a) pp. 318-25.
Schofield, R. S. (1973)'Dimensions of Illiteracy 1750-1850', Explorations in
Economic History, vol. 10, pp.437-54.
Schutz, A. (1962) Collected Papers: vol. 1: The Problems of Social Reality
(The Hague).
Schutz, A. and Luckmann, T. (1973) The Structures ofthe Life-world (Evan-
ston, Illinois).
Scott, A. (1980) The Urban Land Nexus and the State.
Seabrook , J. (1978) What Went Wrong?
Sekscenski, E. S. (1980) 'Women's Share of Moonlighting Nearly Doubles
1969-1979', Monthly Labour Review, vol. 103(5), pp. 57-8.
Shaw, M. (1979) 'Wolverhampton 1871 and 1971 ', Population Trends,
vol. 18, pp.17-23.
Shotter, J. (1983) '''Duality of Structure" and "Intentionality" in an
Ecological Psychology: Towards a Science of Individuality', Journal for
the Theory of Social Behaviour, vol. 13, pp. 19-43.
Simmel, G. (1950) 'The Metropolis and Mental Life' in Wolff (ed.) (1950)
pp.409-24.
Smart, J. J. C. (1963) Philosophy and Scientific Realism.
Smith, D. (1980) 'Tonypandy 1910: Definitions of Community ', Past and
Present, vol. 87, pp. 158-84.
Smith, D. (ed.) (1980) A People and a Proletariat: Essays in the History of
Wales 1780-1980.
Smith, D. (1982) Conflict and Compromise: Class Formation in English
Society 1830-1914. .
Smith, N. (1979)'Geography, Scienceand Post-positivist Modes of Explana-
tion' Progress in Human Geography, vol. 3, pp. 356-83.
Smith, N. (1982) Uneven Development: The Production of Nature under
Capitalism (unpublished Ph.D thesis, The Johns Hopkins University,
Baltimore: Department of Geography).
Smith, S. L. (1983) 'From Capitalist Domination to Urban Planning: Three
Stages in the Urban Politics of Dartford 1814-1980', mimeo.
426 Bibliography
Tabb, W. K. and Sawers, L. (eds) (1978) Marxism and the Metropolis: New
Perspectives in Urban Political Economy (New York) .
Tambiah, S. J. (1968) 'Literacy in a Buddhist Village in N.E. Tha iland' in
Goody (ed.) (1968a) pp. 85-131.
Taylor, M. and Thrift, M. (forthcoming) Capital, Location and Organisation.
Taylor, P. (1971) 'Distance Transformation and Distance Decay Functions'
Geographical Analysis, vol. 3, pp. 221-38.
Taylor, P. and Johnston, R. J. (1979) The Geography ofElections (Harmond-
sworth).
Thomas, B. (1969) 'The Migration of Labour into the Glamorganshire
Coalfield, 1861-1911' in Minchinton (ed.) (1969) pp.37-56.
Thomas, G. and Shannon, C. Z. (1982)'Technology and Household Labour:
Are the Times A'Changing?' mimeo.
Thomas , K. (1971) Religion and the Decline of Magic: Studies in Popular
Beliefs in Sixteenth- and Seventeenth-century England.
Thompson, E. P. (1961) 'The Long Revolution', New Left Review, vol. 9,
pp. 24-33 and vol. 10, pp. 34-9.
Thompson, E. P. (1968) The Making of the English Working Class (Har-
mondsworth).
Thompson, E. P. (1978) The Poverty of Theory and Other Essays.
Thompson, J. B. and Held, D. (eds) (1982) Habermas: Critical Debates.
Thrift , N. (1979) 'Limits to Knowledge in Social Theory: Towards a Theory
of Practice', Department of Human Geography, Australian National
University, seminar paper .
Thrift, N. (1983)'On the Determination of Social Action in Space and Time'
Environment and Planning D: Society and Space, vol. 1, pp . 23-57.
Thrift, N. J. and Forbes, D. K. (1983) 'A Landscape with Figures: Political
Geography with Human Conflict', Political Geography Quarterly, vol. 2,
pp.247-63.
Thrift, N. and Pred, A. (1981) 'Time-geography: A New Beginning',
Progress in Human Geography , vol. 5, pp. 277-86.
Thunell, E. (1952) Livet pa Kabbarp 1 & 6: 1814-1864 (Lund).
Touraine, A. (1977) The Self-production of Society (Chicago).
Townsend, P. (1979) Poverty in the United Kingdom (Harmondsworth).
Ungerson, C. (1981) 'Women, Work and the "Caring Capacity of the Com-
munity" ', mimeo.
Union of Radical Political Economics (1978) Review of Radical Political
Economics, vol. 10.
Urry, J. (1981a) The Anatomy of Capitalist Societies: The Economy, Civil
Society and the State.
Urry, J. (1981b) 'De-industrialization, Households and Forms of Social
Conflict', Lancaster Regionalism Group, Working Paper, no. 3.
428 Bibliography