Besides these individuals, a variety of key
issues within the geographical tradition drew
heavily on evolutionary motifs. Statements
ofenvironmental determinismby figures
such as Ellen Semple, Elsworth Huntington
and Griffith Taylor were invariably couched
in evolutionary categories, with climate,
migration and natural selection routinely play-
ing the leading roles. From a more radical per-
spective, Peter Kropotkin found in a modified
Darwinism the grounds for championing col-
lectivism, opposing Spencerian individualism
and connecting up the philosophy of natural
science with anarchism. This essentially
Russian reading of evolution drew inspiration
from the St Petersburg naturalists, all of whom
had conductedfieldworkin Siberia, in condi-
tions markedly different from the tightly
packed-in niches of the tropical world – in par-
ticular, from the work of Karl Kessler, who set
out the law of mutual aid. The transference of
ideas aboutcommunitybetween sociology and
ecology, couched within an evolutionarypol-
itical economy, also found expression in
geography in the tradition ofhuman ecology
(Mitman, 1992). Debates about acclimatiza-
tion were likewise connected up to questions
about heredity andadaptation(Livingstone,
1987a; Anderson, 1992). And early theoretical
statements aboutregional geography, such
as those of Herbertson and Geddes, were sup-
ported by appeals to the need for elucidating
evolutionary mechanisms in specific contexts
(Livingstone, 1992).
Within contemporary human geography,
issues raised by Darwinism continue to sur-
face. The legitimacy of transferring biological
categories to the social order, for example,
continues to be the subject of debate (cf.bio-
power), as are matters rotating around the
understanding of hownatureandculture–
to employ two abstractions – are to be concep-
tualized. Recently too, enquiries have been
undertaken into the geographyofDarwinism,
with the aim of ascertaining the ways in which
evolutionary theory circulated around the
world and was differently embraced, mobil-
ized and resisted in the light of localcultural
politics(Livingstone, 2006). dl
Suggested reading
Bowler (1989); Livingstone (2006); Stoddart
(1966).
data archive A central repository of access-
ible data sets. Researchers are encouraged to
deposit original data sets (e.g. survey data)
there to enable others to conductsecondary
data analyses. Many countries now have
such archives (usually sponsored by research
funding agencies that encourage – in some
cases require – researchers to deposit data
sets to encourage social scientific advances
through the re-analysis of existing data sets
alongside the creation of new sources). These
archives are increasingly used to store and
disseminate not only original data sets col-
lected by academic researchers but also
those collected by public-sector bodies (such
ascensuses). Although most of those archives
focus on quantitative data, other types are now
stored in formats that allow them to be made
readily available to other researchers – as in
Qualidata, part of the UK Data Archive
housed at the University of Essex (seesoft-
ware for qualitative research) – and
archives that store maps in digital form, and
theinternethas facilitated linking archives
to enable international data-sharing. As well
as storing data, most archives offer training
and other forms of user support to facilitate
analyses of the data sets held. rj
Suggested reading
See theGuide to primary social science data and
related resources available on the Internet(http://
http://www.chass.utoronto.ca/datalib/other/) and the
Social science data archives (http://www2.fmg.
uva.nl/sociosite/databases.html)
data mining Computer-based automated
procedures for searching large and complex
data sets in order to identify spatial patterns
and relationships, either to confirm existing
or to generate newhypotheses.Inspatial
analysis, Openshaw’sgeographical analy-
sis machineandgeographical explanation
machineexemplify data mining algorithms, as
do some aspects ofgeodemographics. Some
term these practices ‘data dredging’, since
they are based on little prior knowledge and
exemplify inductive thinking (seeinduction)
whereby explanations are sought after patterns
are identified (cf.exploratory data analy-
sis): however, they can also be abductive (cf.
abduction;geocomputation). Many of the
structured algorithms for data mining use
artificial intelligenceapproaches. rj
Suggested reading
Berry and Linoff (1997).
decentralization A process of spatial change
generated by centrifugal forces (cf.centrifu-
gal and centripetal forces). Within urban
areas (seeurbanization), demands for space
Gregory / The Dictionary of Human Geography 9781405132879_4_D Final Proof page 143 1.4.2009 3:15pm
DECENTRALIZATION