Comp. by: VPugazhenthi Stage : Revises1 ChapterID: 9781405132879_4_Q Date:31/
3/09 Time:19:14:09 Filepath:H:/00_Blackwell/00_3B2/Gregory-9781405132879/appln/
3B2/revises/9781405132879_4_Q.3d
the creation/re-creation of geographies ofun-
even development and the (class-based)
power relations that underpinned them.
Other critiques emanated from what became
known ashumanistic geographyand the var-
iety of approaches tosocialandcultural
geographythat emerged in subsequent dec-
ades, which focused on spatial science’s impli-
cit characterization of decision-making as
either deterministic (and thus denying the
free will inherent inhuman agency) or influ-
enced by economic factors only, thereby ig-
noring the social and cultural contexts of
decision-making– not least those associated
withgenderandethnicity.
To some extent, these critiques were antici-
pated by geographers working with quantita-
tive methods in the spatial science genre.
Many soon became aware that the complexity
of contemporarysocietymeant that the rela-
tively simple locational models that they were
applying were insufficiently sophisticated to
incorporate the wide range of contextual and
other variables that can influence spatial be-
haviour. They therefore shifted their focus
away from those models that portrayed a
rigid, geometrical spatial order towards more
general models of locational decision-making
that emphasized context – what became
known as behavioural geography. They
sought order – common patterns of decision-
making – in non-deterministic situations,
much as suggested in the attempt through
structuration theoryto eliminate the bin-
ary dualism between structure and agency.
Behavioural geography’s core argument is not
that people are irrational decision-makers but,
rather, that they are boundedly (as against
perfectly) rational: the boundedness reflects
the limits to their information, their ability to
process it and the utility functions that they
apply when evaluating options. This does not
mean that everybody has to be treated as both
unique (having characteristics that nobody
else shares in total) and singular (sharing no
characteristics with others), so that no general
patterns can be found and theories developed.
Rushton (1969) makes this point in contrast-
ing studies ofbehaviour in space– for example,
mapping individual journey-to-work routes –
with those identifyingrules of spatial behaviour,
the general patterns within all those choices
that can be uncovered from the individual
pieces of data using quantitative procedures,
to suggest common behavioural decisions.
The reasons for such commonalities can then
be explored by, for example, identifying
shared information underpinning such de-cisions, as in Peter Gould’s pioneering work
onmental maps, the spatial depictions that
we use when evaluating the places we want to
visit, move to and so on (Gould and White,
1993 [1974]).
This general approach characterizes the
contemporary use of quantitative methods in
human geography. Although some have
claimed that such work is now virtually absent
from human geography’s portfolio and pro-
spectus, much work using quantitative
methods continues to be reported in the dis-
cipline’s general and specialist journals. As
with its predecessors, it concentrates on aggre-
gate patterns in space and their generating
processes: the description, explanation and
(in some cases) prediction of general patterns
of spatial behaviour and their reflection in the
landscape. (Note, however, the recent emer-
gence of a ‘new economic geography’ within
economics, and the creation of a journal –
Journal of Economic Geography– to explore
commonalities between that approach and
the work of geographers who continue to
undertake theoretically inspired quantitative
work.)
Much contemporary work follows the be-
havioural geography rather than the original
spatial scienceparadigm, therefore. Most of
it is theory-driven, based on generalized hy-
potheses of how people behave, but much of
the analysis involves exploring data assembled
to identify patterns rather than explicit
hypothesis-testing. This very largely reflects
the ‘immaturity’ of such studies: our know-
ledge of how people make spatially relevant
decisions and behave is limited, and so the
best we can come up with at this stage is
broad expectations. Inelectoral geography,
for example, there are theoretical arguments
thatself-interestinfluenceshow people vote,
with economic concerns underpinning that
self-interest. But it may be expressed in differ-
ent ways: some may vote sociotropically,
according to their views of the national, re-
gional and/or local economic situation (voting
for the governing party[ies] if they are optimis-
tic about the economic situation but against if
they are pessimistic); some may vote egocen-
trically, according to their perceived personal
financial situation; and some vote altruistic-
ally, in the interests of their (geographical)
neighbours even when these don’t coincide
with their own (Johnston and Pattie, 2006).
Those theoretically based expectations gener-
ate testable hypotheses but – as with the grav-
ity model – they cannot predict in advance the
size of the relevant coefficients for each type ofGregory / The Dictionary of Human Geography 9781405132879_4_Q Final Proof page 609 31.3.2009 7:14pm Compositor Name: ARajuQUANTITATIVE METHODS