hudson fuzzy concepts.pdf

Regional Studies, Vol. 37.6&7, pp. 741–746, August/October 2003

Fuzzy Concepts and Sloppy Thinking:

Reﬂections on Recent Developments in Critical

Regional Studies

RAY HUDSON

Department of Geography and International Centre for Regional Regeneration and Development Studies, Wolfson Research

Institute, University of Durham, Durham DH1 3LE, UK. Email: ray.hudson@durham.ac.uk

(Received November 2002)

H UDSON R. (2003) Fuzzy concepts and sloppy thinking: reﬂections on recent developments in critical regional studies, Reg.

Studies 37 , 741–746. There have been considerable changes in approaches to the study of regions and regional uneven

development in recent years. Much of this has been associated with cultural and institutional ‘turns’ in the social sciences,

involving a greater reliance upon case studies, qualitative forms of evidence and analysis, and di

erent concepts of theory. For

Ann Markusen these developments are problematic, both in themselves and in their resultant increasing detachment from public

policy. I have some sympathy with her critique but in some respects see it as misconceived. Not least this is because of its

conﬂation of quantitative methods with a traditional conception of theory that sits oddly with her call for greater policy and

political involvement. Such concerns are better served by the theoretical traditions and practices of critical theory and political

economy.

ﬀ

Fuzzy concepts

Regional studies

Empirical evidence

Concepts of theory

Public policy

H UDSON R. (2003) Des concepts confus et des pens´es

H UDSON R. (2003) Unklare Konzeptionen und schludriges

laxistes: des r´ﬂexions sur des d´veloppements r´cents dans

Denken:

Betrachtungen

¨ ber

seit

kurzem

auftretende

le domaine des ´tudes r´gionales critiques, Reg. Studies 37 ,

Entwicklungen bei kritischen Regionalstudien, Reg. Studies

741–746. Dans les derni`res ann´es, il y a eu d’importants

37 , 741–746. Ans¨tze zum Studium von Regionen und

changements des fa¸ons d’´tudier les r´gions et les d´s-

regional ungleichmßiger Entwicklung sind in den letzten

´quilibres r´gionaux. Pour la plupart, cela s’explique par

Jahren betr¨chtlichem Wandel unterworfen gewesen. Viel

des ‘tournures’ culturelles et institutionnelles prises dans les

davon ist kulturellen und institutionellen ‘Wenden’ in den

sciences sociales, n´cessitant une d´pendance accrue vis-`-

Sozialwissenschaften zu zuschreiben, die sich vermehrt auf

vis des ´tudes de cas, des formes qualitatives de preuves et

Fallstudien, qualitative Beweisformen und Analysen, sowie

d’analyse, et des concepts de th´orie di ﬀ ´rents. Pour Ann

verschiedene

theoretische

Konzepte

st ¨ tzen.

F ¨ r

Ann

Markusen, ces d´veloppements-l` s’av`rent probl´matiques

Markusen

sind

diese

Entwicklungen

als

problematisch,

en soi et en termes de leur s´paration accrue de la politique

sowohl an sich, als auch in ihrer sich daraus ergebenden

publique. Dans une certaine mesure, on est d’accord avec sa

zunehmenden

Losl ¨ sung

von

der

¨ ﬀ entlich

verfolgten

critique, mais ` certains ´ gards on la consid ` re mal con ¸ ue. A

Politik. Der Autor hat in gewisser Weise Verst ¨ ndnis f ¨ r ihre

commencer parce qu’elle regroupe les m ´ thodes quantitatives

Kritik, sieht sie jedoch in mancher Beziehung als von falschen

autour d’une notion traditionnelle de la th ´ orie qui va

Voraussetzungen ausgehend an. Dies ist haupts ¨ chlich so

lourdement de pair avec sa demande d’un engagement poli-

wegen ihrer Zusammenfassung von quantitativen Methoden

tique et de politique accru. Les traditions th´oriques et les

mit einer traditionellen Konzeption von Theorie, die schlecht

pratiques de la th´orie critique et de l’´conomie politique

zu ihrem Aufruf zu st ¨ rkerem Engagement in Bestrebungen

sont plus utiles ` de tels soucis.

und Politik paßt. Solchen Anliegen ist besser mit den theore-

tischen Traditionen und dem Praktizieren von kritischer

Concepts confuses

Etudes r´gionales

Theorie und Volkswirtschaft gedient.

Preuves empiriques

Concepts de th´orie

Politique publique

Unklare Konzeptionen

Regionalstudien

Empirische Beweise

Konzeptionen der Theorie

ﬀ

entliche Politik

DOI: 10.1080/0034340032000108822

http://www.regional-studies-assoc.ac.uk

742

Ray Hudson

INTRODUCTION

The ‘turns’ of the last decade form the context for

Ann Markusen’s provocative and thoughtful article in

Regional Studies (M ARKUSEN , 1999). Markusen raised

In the last decade or so there has been an increasing

‘turn’

cultural,

institutional

and

evolutionary

important issues, which she believes have been glossed

over in recent years – and there is certainly some

perspectives in the social sciences and in the analysis of

regions and regional development. As a result, regional

validity to this claim. She has sought to open a debate,

and regional analysts ought to be grateful to her for

analysts have increasingly explored the signiﬁcance of

concepts such as knowledge (codiﬁed and tacit), learn-

raising the issues, to which this (at the invitation of

Gernot Grabher and Robert Hassink) is a contribution.

ing (in its varied forms) and trust in seeking to under-

stand

why

some

regions

are

economically

I am in agreement with some – but not all, as will

become clear below – of her critique, not least as she

successful and socially progressive than others. At least

in part, these ‘turns’ were in response to what were

seeks to defend advances made by political-economic

analysts, and raises important issues that deserve to be

seen by some as the limitations of the political-economy

approaches that had previously been in the ascendant.

taken seriously. In other ways, however, as I hope to

show, she in turn glosses over some critical issues of

For some, exploring these other approaches o ﬀ ered

an alternative, rather than a complement, to political

epistemology and theory.

Markusen begins (p. 869) by recognizing the impor-

economy approaches, which were seen as overly struc-

tural and deterministic, allowing too little (if any) space

tant advances made by analysts of regional development

(such as Bluestone and Harrison, Harvey, and Massey 1 ),

for human action and political intervention. At the risk

of stating the obvious, this clearly represents but one

informed and inspired by Marxist and institutionalist

approaches. In particular, she identiﬁes three shared

reading of much of political economy. Moreover, it is

one which many of those who seek both to defend the

features that characterized such work: strong contesta-

tion of prevailing interpretations of urban and regional

merits of a political economy approach and explore the

scope for fruitful engagement with other intellectual

phenomena; adroit use of empirical evidence to force

home their contentions; and the advocacy of collective

traditions in the social sciences, would reject. For others

(myself included: for example, see H UDSON , 2001)

action to overcome the problems addressed by their

analyses. While each of these three points is crucial,

then, this represents a way of adding to the advances

previously made in understanding regional uneven

the second is perhaps the most salient in the context

of her critique of ‘fuzzy concepts’. As she puts it, since

development via explorations of a variety of political-

economy perspectives, with Marxian political economy

the mid-1980s the approach to critical analysis of

regional

perhaps the most prominent of these. This is an argu-

ment against totalizing meta-narrative (singular) but in

development:

‘exempliﬁed

Bluestone,

Harrison, Harvey, Massey et al. , has been supplanted by

a largely uncontested and increasingly obtuse dialogue,

support of meta-narratives (plural). It recognizes that

not all meta-narratives are equally valuable and empha-

relying on what I will call ‘‘fuzzy concepts’’. In some

cases, no attempt is made to o ﬀ er evidence at all.

sizes the need to attend carefully to the absences and

limitations of any given meta-narrative. As such, it

Elsewhere, evidence marshalled is highly selective.

Methodology is little discussed’. This is a pretty damn-

draws attention to the ways in which the knowledge

created within one meta-narrative might be related

ing critique and indictment, lambasting academics on

the grounds of sloppy theorizing, at best selective use

constructively to knowledges created within the para-

meters of other meta-narratives.

of evidence, and a failure to address questions of

methodology. To which one could add, implicit in her

Whatever the precise reading and interpretation of

recent developments, however, what is beyond any

identiﬁcation of the strengths of preceding approaches,

a failure to take seriously their social responsibilities as

reasonable doubt is that the exploration of diverse and

heterogeneous approaches to the study of regions and

intellectuals. The key question, therefore, becomes:

how justiﬁed is this critique?

regional development has led to a lively and more

interesting debate, both within and beyond the acad-

I will structure what follows around major themes

that Markusen herself identiﬁes: ‘fuzzy conceptualiza-

emy. Not least, the growing interest in the social

sciences in the spatiality of social life has ensured that

tion’; falling standards of evidence; and policy distance. 2

I will seek to summarize what I see as the core of her

this debate within the academy has become truly cross-

disciplinary, and has beneﬁted enormously as a result.

critique and in turn reﬂect critically upon it with the

intention of generating debate and discussion around

The exploration of diverse theoretical perspectives,

however, led to a parallel concern with equally diverse

some important issues

forms of evidence – or more precisely, a diversity of

views as to what constituted valid evidence which

FUZZY CONCEPTUALIZATION

challenged not only some of the tenets of political

economy but also the positivistic approaches to regional

For Markusen a ‘fuzzy concept’ is a one that lacks

conceptual clarity and is di cult to operationalize –

analysis, to which political-economy perspectives were

themselves a response.

and in this it seems to me that she conﬂates two separate

Fuzzy Concepts and Sloppy Thinking

743

issues, conceptualization and operationalization, with

monetary price categories. Some have attempted to

solve the problem via a one-to-one quantitative map-

important implications for the way she develops her

argument. A ‘fuzzy concept’ is one that ‘posits an

ping of prices to values (or vice versa). However, this

misses the decisive point that these are concepts of

entity, phenomena or process which possesses two or

more alternative meanings and thus cannot be reliably

qualitatively di ﬀ ering theoretical status. As such, they

cannot be quantitatively equated. Value categories may

identiﬁed or applied by di ﬀ erent readers or scholars’.

She elaborates on how one might detect conceptual

in this sense be ‘invisible’ to the available empirical

evidence of data on prices but this does not invalidate

fuzziness: ‘A simple question which evokes fuzziness of

concept where it inheres is the following: ‘‘How do I

their theoretical value. 3 As M ASSEY , 1995, p. 307,

forcibly argues, the law of value forms the ‘absolutely

know it when I see it?’’ ’ (p. 870). This brings the issue

of the relationship – the visible relationship – of concept

essential basis for some central concepts – exploitation

for instance’ but ‘the Byzantine entanglement into

and evidence centre-stage.

This is ﬁne as far as it goes but Markusen wants to

which the ‘‘law of value’’ has fallen make it . . . unusable

in any empirical economic calculus’. But Massey is

go further. Revealingly, she places great emphasis upon

replicability as the reason for seeking this one-to-one

adamant that this ‘unusability in empirical economic

calculus’ is deﬁnitely not a reason to abandon value

correspondence between concept and evidence: she

deﬁnes the criteria for ‘sound academic work’ in terms

concepts, which for her retain a key theoretical role –

and on this, I am ﬁrmly with Massey.

of ‘someone else [being able] to replicate ﬁeld work

and verify the validity of conclusions’ (p. 870). Now

Moreover, there is another point of critical impor-

tance, a key assumption underpinning the basis of

such ‘replicability’ is a central concern of positivistic

spatial/social science (modelled on a particular and

Markusen’s argument: that is, that ‘replicability’

requires a key assumption that the objects and subjects

none too sophisticated view of the form of theory and

procedures of physical and natural sciences). There

of analysis will remain unchanging to allow such rep-

licability to be possible in the ﬁrst place. However, this

are, however, other conceptions of theory that would

challenge this emphasis on replicability (and by implica-

is a big – indeed heroic – assumption. Ultimately, it

is untenable, for it involves the objectiﬁcation (or

tion, predictability) as a key criterion for assessing the

value of theory. For example, those defending a critical

‘thingyﬁcation’:O LSSON , 1975) of subjects. It fails to

acknowledge that people, unlike natural objects and

realist conception of theory would stress that whether

particular causal powers and liabilities are realized

things, learn and consciously alter their behaviour in

light of this, and may deliberately seek to subvert

depends both on the presence or of absence of other

structures of causal powers and upon time-space speciﬁc

replicability (and so predictability and potential control:

see below) as part of their strategies as knowledgeable

contingent circumstances. The presence or absence of

a given empirically observable behaviour or outcome

and rational social actors.

There is, however, also some ambiguity in Marku-

cannot therefore be used to assess the validity of a

theory and, as such, replicability cannot be a decisive

sen’s own stance towards ‘fuzzy concepts’. She admits

in some cases ‘fuzziness’ may be a necessary stage in

criterion. Rather than replication, corroboration via

the convergence of evidence, perhaps of various types

the maturing of a concept. This is perfectly acceptable

since concepts are social constructs which require social

drawn from a variety of sources, is a much more

appropriate criterion for assessing the validity and value

production, and not ready-formed objects, ‘out there’

awaiting discovery. However, in other instances, ‘fuzzi-

of such theory. The emphasis is more appropriately

placed upon triangulation rather than replication.

ness’ may simply reﬂect theoretical imprecision and

sloppiness of thought (which is not defensible). How-

In short, while stressing the primacy of visible rela-

tionships between concepts and evidence, Markusen’s

ever, she also acknowledges that conceptual ‘fuzziness’

may reﬂect a deliberate tactical move (for example, to

views on these relationships are overly simplistic. This

emphasis

renders

her

line

critique

profoundly

create a broad umbrella under which a political move-

ment might be mobilized), which raises some interest-

dangerous and potentially disabling, theoretically and

politically, particularly for those sensitive to the tradi-

ing questions given her wish for theory to inform

politics and practice, but also presumably for politics

tions of Marxian political economy, critical realism and

critical theory more generally. In addition, she fails to

and practice, in the form of evidence, to inform theory.

So is this a call for an explanation of when ‘fuzziness’

acknowledge the issue of relationships between con-

cepts of di

ﬀ

ering theoretical status, which in turn can

is permissible and desirable – and if it isn’t, shouldn’t

it be?

impinge upon the issue of how empirical evidence

relates to theoretical concepts. The point may be

For Markusen, the general ‘proliferation of fuzzy

concepts’ is accompanied by ‘an increasing emphasis

illustrated by reference to the relationships between

price and value analyses in Marxian political economy.

on process, rather than structure, agency and perfor-

mance’ so that ‘processes themselves become causal

There is a long (unresolved, indeed unresolvable)

debate as to how best to deal with the ‘transformation

actors’ (p. 870). Markusen’s concern here is not simply

one of theoretical looseness and ambiguity, however; it

problem’ and theorize relationships between value and

744

Ray Hudson

also becomes a practical and political issue: ‘The

conception of ‘traditional’ (as opposed to critical:

H ORKHEIMER , 1972) theory. ‘Traditional’ theory rests

fudging or suppression of agency and causality in fuzzy

concepts and process-preoccupied writing is problem-

on a belief that there are underlying regularities to be

revealed by some combination of inductive empirical

atic, particularly for planners, policy makers and poli-

tical activists’. As one who has for a long time held the

analysis deploying extensive methodologies (S AY E R ,

1984) and deductive theorizing. As a result, as is well-

view that the point of theory is to change the world

for the better (recognizing all that is buried in that

known, inherent to such a conception of theory and

predictive power as the test of its validity is the potential

term), I have considerable sympathy with her view that

theory needs to serve practical and socially progressive

for social control (L EWIS and M ELVILLE , 1977). For

critical theorists and critical realists, however, predictive

purposes (and see below). I agree wholeheartedly with

her view that ‘good theory must encompass both

power is not a central issue in assessing the validity

of a theory (see above). Given Markusen’s laudable

process and institutions, both structure and agency’

(p. 871). Indeed, the history of Marxian political econ-

aspirations for regional analysts to help change the

world for the better via engagement with policy and

omy can be read as a debate between those who place

relatively more weight upon agency than structure,

political communities and so on, her preference for

‘traditional’ theory is odd.

and vice versa (A NDERSON , 1984). However, I have

diculty at this point in following Markusen’s line of

The point can be made in another way, with refer-

ence to Markusen’s views as to the roles of qualitative

argument. Having always thought of process as requiring

a speciﬁcation of the links between actors, agents and

research and case studies in regional analysis. A corollary

of apparently privileging quantitative evidence is that

structures, I have diculty in conceptualizing process

as an alternative to the speciﬁcation of these links.

qualitative methods and case studies are seemingly

relegated to a supporting role – they can ‘enrich’

an understanding derived from the construction of

FALLING STANDARDS OF EVIDENCE

‘traditional’ theory based on quantitative evidence and

a particular conception of ‘statistical representativeness’.

Markusen argues reasonably that ‘[e]vidence, broadly,

speaking, is fundamental to the credibility of con-

This is clear from her complaint that ‘authors of

qualitative accounts often fail to make the case that the

tending theories of regional development’ but goes on

to claim – which some will ﬁnd less reasonable –

particular case study is representative or that the ﬁndings

from it are generalizable’ (p. 872). It is certainly the

that ‘our standards of evidence seem to have slipped

dramatically in the last decade’. Markusen recognizes

case that they may not be representative in a statistical

sense, i.e. be a representative randomly drawn sample

that evidence can – and I’d argue, must – take a variety

of forms, qualitative and quantitative, as appropriate. It

of some putative underlying population, to which the

results of analysis may be generalized, with a given

seems to me that any other position is simply indefens-

ible. She points out that ‘new quantitative regional

probability of error. They may, nonetheless, satisfy

the criterion of representativeness required from the

comparative techniques such as Isserman, Beaumont

and Rephan’s quasi-experimental methods are breaking

perspective of critical realism, and be representative of

key causal processes and mechanisms. As such, they

new ground’, going on to add that ‘[q]ualitative tech-

niques have also advanced signiﬁcantly in regional

may well be much more valuable than yet another

statistical or quantitative analysis of spatial pattern in

studies, as judiciously chosen and interpreted case stud-

ies enrich our understanding of regional development’

revealing the social processes that underlie regional

development, in revealing the ways in which causal

(p. 872). Is it mere coincidence that while she gives a

speciﬁc

example

quantitative

methodological

mechanisms may or may not be activated in speciﬁc

contingent circumstances. Markusen states at one point

advances, Markusen simply talks in general terms about

advances in qualitative methodologies?

(p. 872) that ‘it is not my intention to engage in a

debate on epistemology’. However, it seems to me that

In short, it seems to me that Markusen is primarily

concerned with a decline in the use of quantitative

addressing some of the issues her paper raises requires

precisely such an engagement and recognition that

evidence, especially that which would lead to a one-

to-one modelling of the relationships between a con-

erent

epistemological assumptions. The constitution of ‘valid

ﬀ

erent

conceptions

theory

make

ﬀ

cept and empirical evidence. 4 For example, she argues

that: ‘It is common to hear scholars refer to a divide

evidence’ and the meaning of key concepts such as

‘representative’ are contested and dependent upon the

between the quantitative people and the theorists’,a

remark that for me evokes memories of some of the

theoretical context in which they are grounded.

The need to engage with some of these broader

mindless empiricist data dredging of past spatial science

approaches to regional analysis but for Markusen seems

issues seems to me to be critical and relates back to the

need to recognize the signiﬁcance of meta-narratives

to imply that theory construction necessarily requires

quantitative data. This leaning towards a particular sort

and grand theories (both plural). It is, without doubt,

important to avoid a totalizing and deterministic inter-

of evidence is, however, symptomatic of something

much more serious – a preference for a particular

pretation of capitalist economies and the patterns of

Fuzzy Concepts and Sloppy Thinking

745

regional uneven development that are integral to them

prevalence of ‘fuzzy concepts’ has been accompanied

by a retreat from engagement with regional policy,

but it is equally important to recognize that there

are broad social structural relations (of class, gender,

compounded by a ‘relative neglect of the state’s role in

regional development [which] exacerbates the distance

ethnicity, for example) that have determinate e

ects.

However, the precise e ﬀ ects that they do or do not

ﬀ

between theory, research and policy’. In her terms, this

growing policy distance is no bad thing in so far as she

have are contingent matters, and must be revealed

by careful, theoretically-informed empirical research,

is correct in her analysis of the growing dominance of

fuzzy concepts since ‘under-researched fuzzy concepts

typically via an intensive case study (of factories, ﬁrms

or regions, for example) that will necessarily involve

are more tolerable the less we expect them to guide

action’ (p. 873). Now if it is the case that concepts are

collecting and interpreting a variety of forms of evi-

dence, both quantitative and qualitative. What it will

‘fuzzy’ and imprecise, then there can be little argument

that we should apply the precautionary principal before

not involve – because this would simply be inappro-

priate – is some sort of statistical analysis (whether

rushing to apply them in practice, as a guide to public

policy. However, the key question is whether more

simplistic or sophisticated) of quantitative data in search

of empirical regularities, with their prediction as the

‘qualitative’ concepts such as learning, tacit knowledge,

trust, are necessarily ‘fuzzy’ or are only (as with harder

test of a ‘good’ theory. Being able to make some

comments as to how future geographies of regional

‘quantitative’ concepts) ‘fuzzy’ when the product of

bad

uneven development may evolve may indeed be help-

ful, practically and politically. However, the point is

rather

than

good

social

science

research

and

theorizing.

Where

that prediction cannot be the test of a good critical

theory, although it certainly is a key criterion against

much

agreement

with

Markusen, however, is in her insistence that ‘regional

research should be policy-relevant and concerned not

which the validity a ‘traditional’ theory is to be judged.

The precise point of a critical theory, and the criterion

only with eciency but also with normative goals

loosely

against which it is to be judged, is to change the

underlying

considered

progressive:

equity,

democracy;

structures

that

generate

socially

unac-

human rights; environmentally benign development’

(p. 873). If this smacks of an adherence to aspects of a

ceptable and politically regressive outcomes (for

example, in the form of ‘problem regions’), the persist-

modernist agenda and a belief in socially progressive

social science and practice that, to a degree at least, is

ence

which

would

allow

prediction

based

extrapolation of past trends to be possible.

grounded in a recognition of the validity of meta-

narratives, then that seems to me to be exactly what is

Clearly, then, there is a continuing role for structural

analyses, drawing on the legacies of political-economy

required. Equally, I would accept the implication of

her analysis that the national state has a continuing

approaches to understanding regional uneven develop-

ment. However, it is important to insist that this

crucial role to play in this regard (as does, in the

European context, the EU). This is the case in at least

is conceptualized as structural determination, which

deﬁnes limits within which agency can be exercised,

two regards: (1) via progressive public expenditure

policies, both in the direct provision of required goods,

and avoid sliding back into a structural determinism

that closes o ﬀ space for human action and political

services and ‘hard’ and ‘soft’ infrastructures in eco-

nomically marginalized and peripheral regions; and

practice, and the potential to alter structural parameters.

As M ASSEY , 1995, pp. 303-04, eloquently put it,

(2) in providing material support and resources to local

and regional agencies and institutions seeking to address

recognition of such structures ‘is not the same as the

commitment to, or adoption of, a meta-narrative view

developmental problems, both via the mainstream and

social economies (A MIN et al. , 2002) as part of evolving

of history. None of the structures . . . need to be

assumed to have any inexorability in their unfolding

multi-scalar systems of governance. In short, if local

and regional ‘pro-activity’ is to become the accepted

. . . outcomes are always uncertain, history and geo-

graphy have to be made’. The sort of research that

route to economic and social regeneration in depleted

communities

reveals how history and geography – indeed historical

geographies – are made is surely the sort of research

and

marginalized

regions,

then

this

requires underpinning via (supra)national state support.

This emphasis on ‘local’ solutions in turn highlights

needed to reveal why regional uneven development

takes the forms that is does, research drawing from and

the fact that the state may be only one of a range of

social actors engaged in policies for places. Trades

in turn informing critical theory, with the emancipat-

ory aim of seeking progressive change in the map of

unions, for example, may also seek to shape regional

policy agendas, along with local environmental groups.

regional inequality.

However, social forces at the local/regional levels alone

cannot be expected to deal with the consequences of

POLICY DISTANCE

capital ﬂight and restructuring and the retreat of

national state responsibilities for addressing the con-

I have considerable sympathy with the third strand of

Markusen’s critique, although reservations about some

sequent problems. Furthermore, and again I strongly

agree with Markusen on this point, any simplistic

aspects of her route to it. She argues that the growing

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: