S cie n ce

Soc ial s
t M atte r
Tha
Social science is headed down
T
his paper asks what kind of social science we – scholars,
policy makers, administrators – should and should not
a dead end toward mere scientism, promote in democratic societies, and how we may hold social
scientists accountable to deliver what we ask them for.
becoming a second-rate version
My argument follows three main steps:
of the hard sciences. We need to 1. We should avoid social sciences that pretend to emulate natural
science by producing cumulative and predictive theory. The natural
recognise and support a different science approach simply does not work in the social sciences. No
predictive theories have been arrived at in social science, despite
kind of social science research – centuries of trying. This approach is a wasteful dead-end.
2. We should promote social sciences that are strong where natural
and so should those who demand science is weak – that is, in reflexive analysis and deliberation
about values and interests aimed at praxis, which are essential to
accountability from researchers. social and economic development in society. We should promote
value rationality over epistemic rationality, in order to arrive at
social science that matters.
by Bent Flyvbjerg 3. Policy makers and administrators should reward such praxis-oriented
social science, and they should penalise social science that has no
Professor of Planning, Department of Development and Planning, social and practical import, including social science which vainly tries
Aalborg University to emulate natural science. This would be accountability that matters.

Bent Flyvbjerg 38 foresight Europe / october 2005 > march 2006

 Two Types Why the Natural Science
of Social Science Model Does Not Work
Table 1 outlines two models for doing social science, the epistemic and
in Social Science
the phronetic models.
• The epistemic model finds its ideal in the natural science model for Inspired by the relative success of the natural sciences in using
doing science. Here the objective of the social scientist is to discover mathematical and statistical modelling to explain and predict natural
the theories and laws which govern social action, just as the objective phenomena, many social scientists have fallen victim to the following
of the natural scientist is to discover the theories and laws which pars pro toto fallacy: If the social sciences would use mathematical and
govern natural phenomena. Praxis, according to the natural science statistical modelling like the natural sciences, then social sciences, too,
model of social science, is social engineering which applies social theories would become truly scientific.
and laws to solve social problems. Often quantitative social scientists see economics as an ideal to follow,
A classic simile for this type of social science is the so-called “moon- because it is the “hardest” and thus seemingly most scientific of the
ghetto metaphor,” named for social scientists who argued during the social sciences. Economics has gone furthest with mathematical and
1960’s and 1970’s that if natural science and engineering could put a statistical modelling, but recently parts of political science and
man on the moon, surely social science could solve the social problems sociology have followed suit under the influence of rational choice and
of the urban ghetto (Nelson 1977). History proved them wrong. game theory. Commentators talk, for instance, about “economics envy”
• The phronetic model of social science takes as its point of departure the among political scientists (Stewart 2003).
fact that despite centuries of trying the natural science model still does not Such envy is misguided, for not even economics has succeeded in
work in social science: No predictive social theories have been arrived avoiding context (an issue we will consider in detail shortly) and
at as yet. The phronetic model is named after the Aristotelian concept becoming relatively cumulative and stable, like a natural science.
phronesis, which is the intellectual virtue used to deliberate about which Economists have been defined, jokingly but perceptively, as “experts
social actions are good or bad for humans. The basis of deliberation is who will know tomorrow why the things they predicted yesterday did
value rationality instead of epistemic rationality. not happen today.” Furthermore, it seems that the more “scientific”
At the core of phronetic social science stands the Aristotelian maxim academic economics attempts to become, the less impact academic
that social issues are best decided by means of the public sphere, not economists have on practical affairs. As pointed out by Sent (2002) in
by science. Though imperfect, no better device than public deliberation the Southern Economic Journal, Wall Street firms prefer to hire
following the rules of constitutional democracy has been arrived at for physicists, because they have a real as opposed to fake natural science
settling social issues, so far as human history can show. Social science background. Academic economists had little or no role to play in the
must therefore play into this device if it is to be useful. This is best final decisions concerning the North American Free Trade Agreement.
done by social scientists: (1) producing reflexive analyses of values and And though the US spectrum auction of frequency bands for additional
interests and of how values and interests affect different groups in cell phone use has been claimed as a victory for game theory, a closer
society, and (2) making sure that such analyses are fed into the process look at the developments reveals that the story is a bit more complex,
of public deliberation and decision making, in order to guarantee that according to Sent. In short, quantitative social scientists should hestitate
legitimate parties to this process, i.e., citizens and stakeholders, receive before insisting on emulating academic economics.
due diligence in the process. The underlying issue is that being scientistic does not amount to being
To sum up the differences: The epistemic or natural science model sees scientific. Regardless of how much we let mathematical and statistical
social scientists and social science professionals as technocrats who – modelling dominate the social sciences, they are unlikely to become
through their insight into social theories and laws – may provide scientific in the natural sciences sense. This is so because the phenomena
society with solutions to its social ills. The phronetic model sees social modelled are social, and thus “answer back” in ways natural phenomena
scientists and social science professionals as analysts who produce food do not. Weinberg (2001: 97), winner of the Nobel Prize in Physics and
for thought for the ongoing process of public deliberation, participation, an astute observer of what makes for success in science, is right when
and decision making. he observes that “it has been an essential element in the success of
science to distinguish those problems that are and are not illuminated
Table 1: Two models for doing social science by taking human beings into account.”
A crucial aspect of this distinction resides in the fact that the relevant context
Epistemic Social Science Phronetic Social Science of social action is human beings’ everyday background skills (Flyvbjerg
(the natural science model) (the reflexive model) 2001, chaps. 3-4). These skills are central in deciding what counts as the
relevant objects and events whose regularities social theory tries to explain
Epistemic rationality Theories and laws ----
of society and social action and predict. Context is not simply the singularity of each setting (as in a
laboratory), nor the distinctive historical and social paths taken to produce
such a setting, even if both may be important to understanding specific
Praxis Application of theories and laws Provision of input for public social phenomena. Ultimately, the human skills that determine the social
to solve social problems, i.e., deliberation and decision making, context are based on judgments that cannot be understood in terms of
social engineering i.e., democratic due diligence concrete features and rules. Therefore a “hard” theory of context in the
(instrumental rationality) (democratic rationality) social sciences is seemingly impossible. But if context decides what counts
as relevant objects and events, and if the social context cannot be
Value rationality ---- Reflexive analysis of values and formalised in terms of features and rules, then social theory cannot be
interests and how they affect different complete and predictive in the manner of much natural science theory,
groups in society which does not have the problem of self-interpretive objects of study.

Bent Flyvbjerg 39 foresight Europe / october 2005 > march 2006

One could reasonably ask: If no one can specify judgment in such a way
as to produce uniformly accurate predictions, does that mean that more Phronetic Social Science
modest and less successful efforts at dealing with judgment are useless,
as in, for instance, accounts of bounded rationality? And does it mean
that we cannot distinguish better from worse instances of judgment? The principal objective for phronetic social science is to understand
The answer is no on both counts. Such efforts may be useful. But they values and interests and how they relate to praxis. The point of departure
will not be science in the natural science sense. for this type of social science can be summarised in the following four
The above argument leads to the conclusion that social science is nei- value-rational questions, which must all be answered for specific,
ther “normal” nor “revolutionary” in the Kuhnian sense. Nor is it pre- substantive problematics, for instance in management:
or post-paradigmatic, as respectively Dreyfus (1991) and Schram (1) Where are we going?
(forthcoming) argue, because no paradigmatic phase has preceded the (2) Who gains and who loses, and by which mechanisms of power?
current situation or is likely to follow it. Kuhn’s concepts regarding (3) Is this development desirable?
paradigm change, that is, a new paradigm substituting for an older one (4) What, if anything, should we do about it?
after a scientific revolution were developed to fit natural science, and
they confuse rather than clarify when imported into social science. In Social scientists following this approach realise there is no global and
my analysis, social science is non-paradigmatic and is neither relatively unified “we” in relation to which the four questions can be given a final
cumulative nor relatively stable. In comparison, although natural answer. What is a “gain” or a “loss” often depends, crucially, on perspec-
science may be neither as rational nor as cumulative as believed tive: My gain may be your loss.
earlier, it still shows a type of stability and progress not found in social Phronetic social scientists are highly aware of the importance of
science. Social scientists who see the natural science model as an ideal perspective, and see no neutral ground, no “view from nowhere,” for
to follow sometimes claim that we have simply not yet discovered the their work. The “we” may be a group of social scientists or, more
various factors or rules that produced outcomes of significance. typically, a group including other actors as well. Phronetic social scientists
Appealing to context when arguing that social science can probably are well aware that different groups typically have different world views
never be explanatory and predictive in the manner of natural science and different interests, and that there exists no general principle by which
is therefore a “cop out,” according to this argument, promoted for all differences can be resolved. Thus phronesis gives us both a way to
instance by Laitin (2003: 168). analyse relations of power, and to evaluate their results in relation to
The argument is easy to counter. So far all attempts to analyse context in specific groups and interests.
social science as merely very complex sets of rules or factors have failed. The four value-rational questions may be addressed, and research
And if Laitin or other social scientists have found a way around this developed, using different methodologies. In other words, phronetic
problem they should rush to publish the evidence, because it would be a social science is problem-driven, not methodology-driven. The
real discovery and a sensation. It would open up the most important issue is not the individual methodology involved,
possibility, for the first time, that the social scien- even if methodological questions may have some significance.
ces could offer the type of theoretical explanation w e s t ill It is more important to get the result right – that is, to arrive
and prediction that today we find only in parts of “The reas
on at social sciences that effectively deal with deliberation,
the natural sciences. th e or y of judgment, and praxis in relation to the four value-rational
Ferrara (1989: 316, 319) has rightly pointed out lac
ka h a t it
nt is t questions, rather than being stranded with social sciences
that we need a theory of judgment in order to avoid judgme g h t in to that vainly attempt to emulate natural science at the cost of
contextualism, that is, the conclusion that validity n o t be brou .” taking judgment out of the picture.
ca n fo r m u la
and truth are context dependent, and that such a
retic a l Asking value-rational questions does not imply a belief in
theory does not exist as yet. The reason we still lack a theo linearity and continuous progress. We know enough
a theory of judgment, and therefore cannot explain about power to understand that progress is often complex, ephemeral,
and predict context, is that judgment cannot be brought and hard-won, and that setbacks are inevitable. Moreover, no one has
into a theoretical formula, as shown in Flyvbjerg (2001, chaps. 3-4). enough wisdom and experience to give complete answers to the four
When Laitin claims it is a cop out for social scientists to appeal to context questions, including social scientists. What should be expected, however,
in order to explain social phenomena, he accepts the burden of either pro- is that phronetic social scientists will indeed attempt to develop their
viding a theory of judgment, or of arguing that Ferrara is wrong in saying answers, however incomplete, to the questions. Such answers would be
we need such a theory in order to avoid appeals to context. input to ongoing dialogue about the problems, possibilities, and risks
We cannot, in principle, rule out that context, skills, and judgment may we face, and about how things may be done differently.
be studied in terms of elements which would make social science Focusing on values, phronetic social scientists are forced to face what is
explanatory and predictive in the manner of natural science (Flyvbjerg perhaps the most basic value-question of all, that of foundationalism
2001: 46-47). But for this to happen we would need a vocabulary in versus relativism – that is, the view that there are central values that
social science which picked out elements of human action that would be can be rationally and universally grounded, versus the view that one set
completely different from those abstracted from our everyday activities. of values is as good as another. Phronetic social scientists reject both
The elements would have to remain invariant through changes in back- of these positions and replace them with contextualism or situational
ground practices, in order to qualify as elements in context independent ethics. Distancing themselves from foundationalism does not leave
theory. No one has yet found such elements, and the logical possibility phronetic social scientists normless, however. They find their point of
that some day they may be discovered has little practical use. This departure in their attitude to the situation being studied. They seek to
possibility is merely in-principle, and cannot be used to conclude – as ensure that such an attitude is not based on idiosyncratic morality or
the proponents of epistemic social science would have it -- that the personal preferences, but on a common view among a specific reference
social sciences are pre-paradigmatic owing merely to historical group to which they refer. For phronetic social scientists, the socially
coincidence, to social science being young, or to a high degree of and historically conditioned context – and not the elusive universal
complexity in the social world (Dreyfus 1991). grounding that is desired by certain scholars – constitutes the most

Bent Flyvbjerg 40 foresight Europe / october 2005 > march 2006

effective bulwark against relativism and nihilism. Phronetic social
scientists realise that as researchers, their sociality and history is the only
solid ground under their feet; and that this socio-historical foundation
is fully adequate for their work.
As regards validity, phronetic social science, like any other social
science, is based on interpretation and is open for testing in relation to
other interpretations and other research. Thus the results of phronetic
social science may be confirmed, revised, or rejected according to
the most rigorous standards of social science, in relation to other
interpretations. This does not mean that one interpretation can be just
as good as the next, as relativism would have it, for each interpretation
must be based on validity claims. It does mean, however, that phrone-
tic social science will be as prepared to defend its validity claims as any
other research.
Phronetic social scientists also oppose the view that any given
interpretation lacks value because it is “merely” an interpretation. As
emphasised by Nehamas (1985: 63), the key point is the establishment
of a better option, where “better” is defined according to sets of validity
claims. If a new interpretation appears to better explain a given
phenomenon, that new interpretation will replace the old one, until it, We are talking about a difference in degree, not in kind, because
too, is replaced by a new and even better interpretation. This is other actors could still advance their own claims. To the phronetic
typically a continuing process, not one that terminates with “the right researcher, this is the reality of social science, in contrast to researchers
answer.” Social science and philosophy have not yet identified criteria by who act as if validity claims can and should be given final grounding
which an ultimate interpretation and a final grounding of values and (and with it, total acceptance). By substituting phronesis for
facts can be made. episteme, phronetic social scientists avoid trying to lift this
This work is dialogical in the sense that it incorporates, and, if impossible burden.
successful, is incorporated into, a polyphony of A first step in moving towards phronetic social sciences is
voices. No one voice, including that of the resear-
f socia l for social scientists to explicate the different roles of their
cher, may claim final authority. The goal is to pro- m p ts o research. The oft-seen image of impotent social sciences
duce input to dialogue and praxis in social affairs, “The a
tte e
to be c o m versus potent natural sciences is misleading and derives
rather than to generate ultimate, unequivocally verified
scie n c e scien c e from their being compared in terms of their epistemic
“knowledge”. Dialogue is not limited to the relation- is t e m ic qualities. If we instead compare the two types of
ship between researchers and the people they study, ‘real
’, ep a n d science in terms of their phronetic qualities we get the
nti o n
but may include anyone interested in and affected by draw atte fr o m opposite result: strong social science and weak natural
e s a w a y
resourc s where
the subject under study, and may be started by parties science. From that perspective, the attempts of social
besides the researchers. science to become “real”, epistemic science draw
Thus, phronetic social science explicitly sees itself as those ar
ea u ld
c o attention and resources away from those areas where
not having a privileged position from which the final cia l s c iences nd social sciences could make an impact, and toward
truth can be told and further discussion arrested. We can- s
o p a c t, a
not think of an “eye turned in no particular direction,” as make
an im er e th ey areas where they do not, never have, and probably
h never
reas w , and normal and predictive sciences.
will, obtain any significance as Kuhnian
Nietzsche (1969: 119) says. “There is only a perspective
to w a rd a e
seeing, only a perspective ‘knowing;’ and the more affects
o n o t, n ever hav l, One useful task of organisation research
we allow to speak about one thing, the more eyes, different d r w il
eyes, we can use to observe one thing, the more complete probably
neve n c e
practised on the basis of the guidelines

s a presented here is to provide concrete examples

ignific d and detailed narratives of the ways in which
will our ‘concept’ of this thing, our ‘objectivity,’ be” (emphasis
tai n a n y n
in original). Hence, “objectivity” in phronetic social science is ob
u h n ia n normal a power and values work in organisations and
not “contemplation without interest” but employment of “a as K e s .” with what consequences, and to suggest how
variety of perspectives and affective interpretations in the ic ti v e s cienc power and values could be changed to work
service of knowledge” (emphasis in original).
pred with other consequences. Insofar as organisa-
It might be feared that the dialogue sought by phronetic social tional situations become clear, they are clarified by detailed
science will easily degenerate into a cacophony in which the stories of who is doing what to whom. Such clarification is a principal
loudest voice carries the day. But the significance of any given concern for phronetic organisation research, which explores current
interpretation will depend on the extent to which its validity claims are practices and historic circumstances to find avenues to praxis.
accepted in this dialogue, and phronetic social scientists recognise that The task is to identify, and deliberate about, the problems, possibilities,
acceptance typically occurs in competition with other claims. If, on the and risks that organisations face, and to outline how things could
contrary, the arguments of researchers carry special weight in the be done differently – all in full knowledge that we cannot find ultimate
dialogue, it would likely derive from their having spent more time on, answers to these questions, or even a single version of what the
and being better trained at, establishing validity than other actors. questions are.

Bent Flyvbjerg 41 foresight Europe / october 2005 > march 2006

Conclusions
Two scenarios may be outlined for the future of social science. References
In the first – and today, dominant – scenario, it is scientism, Bartky, Sandra Lee (1990). Femininity Latour, Bruno (1999). Pandora’s Hope:
and Domination: Studies in the Essays on the Reality of Science Studies.
the belief that science holds a reliable method of reaching the Phenomenology of Oppression. New York, Cambridge, MA: Harvard University Press.
truth about the nature of things, which continues to dominate NY: Routledge. Leijon, Svante (1993 Ed.) God Kommunal
the social sciences. But scientism in social science will continue Brytting, Tomas (2001). Att Vara Som Gud? Organisation: Lyssning, Tolkning och Handling,
Moralisk Kompetens i Arbetslivet. KFi-rapport 22, Göteborg.
to fail, because the reality of social science does not and cannot Malmö: Liber. Leijon, Svante (1996). Arbete i Brytningstid:
live up to the ideals of natural science. Consequently, social Brytting, Tomas, Hans De Geer, Gunilla Tankar om Tillvarons Dualiteter, BAS,
science will increasingly degenerate as a scholarly activity, and Silfverberg (1997). Moral i Verksamhet: Göteborg.
Ett Etiskt Perspektiv på Företag och Arbete.
will find it more and more difficult to gain public support and Stockholm: Natur och Kultur.
Miller, Peter (1994). “Accounting and
Objevtivity: The Invention of Calculating
funding for its activities. Clegg, Stewart (1989). Frameworks of Power. Selves and Calculable Spaces,” in
The second scenario replaces scientism with phronesis. Here the Newbury Park, CA: Sage. A. Megill (Ed.) Rethinking Objectivity.
Durham: Duke University Press, pp: 239-264.
purpose of social science is not to develop epistemic theory, but Clegg, Stewart (1997). “Foucault, Power and
Organisations,” in A. McKinlay and K. Starkey Miller, Peter (2003). “Management and
to contribute to society’s practical rationality by elucidating (Eds.) Foucault, Management and Accounting,” in Theodore M. Porter and
where we are, where we want to go, and what is desirable according Organisation Theory: From Panopticon to Dorothy Ross (Eds.) The Cambridge History
Technologies of Self. London: Sage, pp: 29-48. of Science, vol. 7, The Modern Social
to different sets of values and interests. The goal of the phrone- Sciences. Cambridge: Cambridge University
Crush, Jonathan (1994). “Scripting the
tic approach becomes contributing to society’s capacity for Compound: Power and Space in the South Press, pp: 565-576.
value-rational deliberation and action. The contribution may be African Mining Industry,” Environment and Miller, Peter and Nikolas Rose (1997).
Planning D: Society and Space, 12: 301-324. “Mobilising the Consumer: Assembling
a combination of concrete empirical analyses and practical the Subject of Consumption”, Theory,
Dean, Mitchell (1991). The Constitution of
philosophical-ethical considerations – “fieldwork in philosophy,” Poverty: Toward a Genealogy of Liberal Culture and Society, 14: 1-36.
as Pierre Bourdieu called his own version of phronetic social Governance. London: Routledge. Minson, J. (1993). Questions of Conduct:
Sexual Harassment, Citizenship and
science. In this scenario social scientists actively ensure that Dean, Mitchell (1999). Governmentality:
Government. London: Macmillan.
Power and Rule in Modern Society.
their work is relevant to praxis. The aim is to make the line Thousand Oaks, CA: Sage. Nehamas, Alexander (1985). Nietzsche:
between research and the world direct and consequential. Donzelot, Jacques (1979). The Policing of Life as Literature. Cambridge, MA:
Harvard University Press.
If we want more phronesis in social science, we need to do Families. New York: Pantheon.
Dreyfus, Hubert (1991). “Defending the Nelson, Richard R. (1977). The Moon and the
three things: Difference: The Geistes/Naturwissenschaften Ghetto: An Essay on Public Policy Analysis.
• First, we must drop all pretence, however indirect, of emulating Distinction Revisited,” in Akademie der New York: W.W. Norton.
the relative success of the natural sciences in producing Wissenschaften zu Berlin, Einheit der Nietzsche, Friedrich (1969). On the Genealogy
Wissenschaften, Internationale Kolloquium of Morals. New York: Vintage Books, 1969.
cumulative and predictive theory, for their approach simply der Akademie der Wissenschaften zu Berlin, Procacci, Giovanna (1993). Gouverner la
does not work in social science. Bonn, 25-27 June 1990. misère: La question sociale en France
Berlin: Walter de Gruyter. (1789-1848). Paris: Éditions du Seuil.
• Second, we must address problems that matter to groups in the
Ewald, François (1986). L’Etat providence. Rabinow, Paul (1989). French Modern:
local, national, and global communities in which we live, and we Paris: B. Grasset. Norms and Forms of the Social Environment.
must do it in ways that matter; we must focus on issues of Ewald, François (1996). Histoire de L’Etat Cambridge, MA: MIT Press.
context, values, and power, as advocated by great social scientists providence: Les Origines de la solidarité. Rabinow, Paul (1996). Essays on the
Paris: Grasset. Anthropology of Reason. Princeton: Princeton
from Aristotle to Machiavelli to Max Weber. Ferrara, Alessandro (1989). “Critical Theory University Press.
• Finally, we must effectively and dialogically communicate the and its Discontents: On Wellmer’s Critique of Rabinow, Paul (1999). French DNA: Trouble in
results of our research to our fellow citizens and carefully listen Habermas,” Praxis International 8: 305-320. Purgatory. Chicago: The University of Chicago
Flyvbjerg, Bent (1998). Rationality and Power: Press.
to their feedback. Democracy in Practice. Chicago: The Rose, Nikolas (1985). The Psychological
University of Chicago Press, 1998. Complex: Psychology, Politics and Society
If we do this – focus on specific values and interests in the context Flyvbjerg, Bent (2001). Making Social Science in England, 1869-1939. London: Routledge
Matter: Why Social Inquiry Fails and How It and Kegan Paul.
of particular power relations – we may successfully transform social Can Succeed Again. Cambridge: Cambridge Rose, Nikolas (1996). Inventing Our Selves:
science into an activity performed in public and for different University Press. Psychology, Power and Personhood.
publics, sometimes to clarify, sometimes to intervene, sometimes to Flyvbjerg, Bent, Nils Bruzelius, and Werner Cambridge: Cambridge University Press.
Rothengatter (2003). Megaprojects and Risk:
generate new perspectives, and always to serve as eyes and ears in An Anatomy of Ambition. Cambridge
Schram, Sanford F. (forthcoming). “Return
to Politics: Perestroika, Phronesis, and
efforts to understand the present and deliberate about the future. University Press. Post-Paradigmatic Political Science.”
We may, in short, arrive at social science that matters. Harcourt, Bernard E. (2001). Illusion of Order: In Sanford F. Schram and Brian Caterino (Eds.)
The False Promise of Broken Windows Making Political Science Matter: The
Policy makers and university administrators should reward this Policing. Cambridge, MA: Flyvbjerg Debate and Beyond. New York: New
type of praxis-oriented social science and they should penalise Harvard University Press. York University Press.
social science that has no social and practical impact, including Hardy, Cynthia and Stewart Clegg (1996). Sent, Esther-Mirjam (2002). “Review of
“Some Dare Call It Power,” in Clegg, ‘Making Social Science Matter: Why Social
social science which vainly tries to emulate natural science. Hardy and Nord (Eds.) Handbook Inquiry Fails and How It Can Succeed Again’
This would be accountability that matters. of Organisation Studies. London: Sage, by Bent Flyvbjerg,” Southern Economic
pp: 622-641. Journal, 68: 732-734.
Johansson, Ulla (1998). Om Ansvar: Stewart, Sharla A. (2003). “Revolution From
Ansvarföreställningar och Deras Betydelse Within.” University of Chicago Magazine 95,
för den Organisatoriska Verkligheten. no. 5, http://magazine.uchicago.edu/0306/
Lund: Lund University. features/index.shtml.
Laitin, David (2003). “The Perestroikan Weinberg, Steven (2001). “What Science
Challenge to Social Science,” Politics Can Explain.” The New York Review of Books,
and Society, 31: 163-184. September 20: 97.
Latour, Bruno (1996). Aramis, Or the Love of
Technology. Cambridge, MA: Harvard
University Press.

Bent Flyvbjerg 42 foresight Europe / october 2005 > march 2006