New Worlds in Political Science
| Author | Patrick Dunleavy |
| Date | 01 March 2010 |
| DOI | 10.1111/j.1467-9248.2009.00834.x |
| Published date | 01 March 2010 |
| Subject Matter | Article |
New Worlds in Political Sciencepost_834239..265
Patrick Dunleavy
London School of Economics and Political Science
‘Political science’ is a ‘vanguard’ field concerned with advancing generic knowledge of political processes, while a
wider ‘political scholarship’ utilising eclectic approaches has more modest or varied ambitions. Political science
nonetheless necessarily depends upon and is epistemologically comparable with political scholarship. I deployBoyer’s
distinctions between discovery, integration,application and renewing the profession to show that these connections are
close woven. Two sets of key challenges need to be tackled if contemporary political science is to develop positively.
The first is to ditch the current unworkable and restrictive comparative politics approach,in favour of a genuinely
global analysis framework. Instead of obsessively looking at data on nation states,we need to seek data completeness
on the whole (multi-level) world we have. A second cluster of challenges involves looking far more deeply into
political phenomena;reaping the benefits of ‘digital-era’developments; moving from sample methods to online census
methods in organisational analysis; analysing massive transactional databases and real-time political processes (again,
instead of depending on surveys); and devising new forms of ‘instr umentation’, informed by post-rational choice
theoretical perspectives.
One does not set out in search of new lands without being willing to be alone on an
empty sea.
André Gide1
The undeveloped state of political science has been a theme for depressed navel gazing in
the profession over the last two decades. Our collective mood nowadays seems far removed
from the confidence that attended the first lectures in Cambridge by Sir John Seeley (1896),
the assertive pro-state position of early American political science in the hands of the
younger Woodrow Wilson (see Dryzek et al., 1995), the enthusiasm of the post-war
‘behavioural revolution’ or the early hopes of the rational choice pioneers. In the UK’s
‘Political Studies’or ‘Government’departments (the continued British naming equivocation
says it all) it is not hard to find people who agree heartily with Hal Abelson’s populist
judgement:‘Anything which uses science as part of its name isn’t: political science,creation
science, computer science’.2
Yet the mainspring of cur rent pessimism is actually localised chiefly in three areas that
should be manageable – the inescapable eclecticism of over-inclusive definitions of ‘politi-
cal science’; the apparent ‘reset to zero’ character of theory disputes; and the problems of
doing cumulative research in a fast-chang ing area of human behaviour. In the first part of
the article I seek to insulate forward thinking against these difficulties by defining an
explicitly ‘vanguardist’ (that is, non-inclusive) idea of ‘political science’, but one that
nonetheless rests on a wider substructure of ‘political scholarship’. Political scholarship still
predominates in our discipline and its role is both essential and fully legitimate.But political
science means something more.
doi: 10.1111/j.1467-9248.2009.00834.x
POLITICAL STUDIES: 2010 VOL 58, 239–265
© 2010The Author.Jour nal compilation © 2010 Political StudiesAssociation
In the second and third sections I briefly explore two ‘new worlds’for a refocused political
science, whose tackling or neglect will define our discipline for the next half-century. One
is to begin to build a genuinely global political analysis which can for the first time capture
political experiences across the whole world we have and treat them intelligently and
equitably. To do so will require that we ‘break the mould’ of comparative politics, whose
existing practice fetishises the nation state as a unit of analysis and focuses single-mindedly
on the same stale set of macro-institutional features, when we know that many different
micro institutions matter.
The second set of challenges is to look more deeply and precisely at political behaviours than
wehave so far attempted.We need to modernise and upg rade radically our ambitions to collect
evidence, our analysis of ‘fields’ of possibilities and our standards of proof and analysis.Key
pointers to a better future include using non-reactive web census methods, analysing
transactional and events data and embracing‘post-rational choice’theor ies and new methods.
Political Science and Political Scholarship
Science is built up with facts, as a house is with stones.But a collection of f acts is no
more a science than a heap of stones is a house.
Henri Poincaré3
A lot of learning can be a little thing.
Spike Milligan4
If our discipline means everything that gets done within departments labelled ‘Political
Science’ or ‘Political Studies’, its common purposes will seem elusive and its content
eclectic. An APSR editorial pessimistically observed:
Political science is a strange discipline. Indeed, it is hardly a discipline at all ... [R]ather than
being a distinct branch of learning, political science is a crazy quilt of borrowings from history,
philosophy, law, sociology, psychology, economics, public administration, policy studies, area
studies, international studies, civics, and a variety of other sources. Any real coherence in
political science exists only at the broadest conceptual level (Siegelman, 2002, p.viii).
A second area of concern for pessimists is that political science (like the rest of the social
sciences and humanities) will remain a ‘non-paradigm’ field in Kuhn’s terms, where
intellectual disputes constantly threaten to ‘reset to zero’ established results by challenging
their provenance, methods, empirical validity or theoretical premises. Because of this ‘pull
it up to examine the roots’approach, political science will not become what Randall Collins
(1994) famously called ‘High-Consensus,Rapid-Discovery Science’ as found in the physical
sciences. Beginning from around 1600 onwards and moving at an accelerating pace over
time all the STEM disciplines5were distinguished by ‘high consensus on what counts as
secure knowledge and rapid-discovery of a train of new results’ (Collins, 1994, p. 155). A
‘law of small numbers’ in intellectual disputes still operates in these disciplines (see Collins,
1998), but only at the research frontier itself:
It is the existence of the rapid discovery research front that makes consensus possible on old
results. When scientists have confidence they have a reliable method of discovery, they are
240 PATRICK DUNLEAVY
© 2010The Author.Jour nal compilation © 2010 Political StudiesAssociation
POLITICAL STUDIES: 2010, 58(2)
To continue reading
Request your trial