The ethics of engaged scholarship

in a complex world

DEBORAH AVANT, NAAZNEEN H. BARMA, GEORGE F.

DEMARTINO AND ILENE GRABEL *

Is it not in the interest of social science to embrace complexity, be it at some sacrifice of its
claim to predictive power?1

Imagine two worlds in which scholars engage to promote social betterment. The
two worlds are alike in all respects but for their epistemic conditions. In the first,
scholars have the possibility of sufficient, even if incomplete, knowledge and can
know well enough the causal models of the domains in which they operate. The
models are largely appropriate to the nature of the problems that arise and give
adequate guidance for policy and other interventions. The world they confront
is ergodic, with the past providing sufficient guidance to the future. They can
know at least approximately how decision-makers and target populations will
respond to their advice. They can also know, with relative certainty, about the
relevant scope conditions and salient contingencies that will have a bearing on
the impact of their interventions. Is this a caricature of how engaged scholars
approach their work? We think not, since a significant portion of mainstream
scholarship routinely speaks of the world as if it can be sufficiently known.
Now consider a second world characterized by the impossibility of sufficient
knowledge. Given inescapable uncertainty, scholars cannot ever know enough
to be sure of the causal mechanisms operating in a non-ergodic, complex and
unpredictable world. Many problems are ‘wicked’ rather than simple.2 A wicked
problem ‘is not well bounded, is framed differently by various groups and individ-
uals, involves large scientific to existential uncertainties, and tends not to be well
understood until after the formulation of a solution’.3 In this world, causal models
are always deficient in one way or another, but scholars cannot know precisely

* We would like to thank the Carnegie Corporation of New York, which funded the research upon which this
article is based. We are grateful to the scholars who generously shared with us their reflections on their public
engagement experiences and to Sevde Acabay-Oguc, Kelci Burckhardt and Audrey Elliott for their research
assistance with the collection of the reflections.
1
Albert O. Hirschman, Rival views of market society and other recent essays (first publ. 1986), quoted in Jeremy Adel-
man, ed., The essential Hirschman (Princeton: Princeton University Press, 2013), pp. 214–47 at p. 243.
2
Horst W. J. Rittel and Melvin M. Webber, ‘Dilemmas in a general theory of planning’, Policy Sciences 4: 2,
1973, pp. 155–69, https://doi.org/10.1007/BF01405730; Kate Crowley and Brian Head, ‘The enduring chal-
lenge of “wicked problems”: revisiting Rittel and Webber’, Policy Sciences 50: 5, 2017, pp. 539–47, https://doi.
org/10.1007/s11077-017-9302-4.
3
Robert Lempert, ‘Embedding (some) benefit-cost concepts into decision support processes with deep uncer-
tainty’, Journal of Benefit-Cost Analysis 5: 3, 2014, pp. 487–514 at p. 488, https://doi.org/10.1515/jbca-2014-9006.

International Affairs 100: 1 (2024) 159–180; doi: 10.1093/ia/iiad292

© The Author(s) 2024. Published by Oxford University Press on behalf of The Royal Institute of International Affairs.All rights
reserved. For permissions, please email: journals.permissions@oup.com

INTA100_1_FullIssue.indb 159 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
how. They cannot know enough about the people they hope to serve—about how
they will behave as the world changes, or about their values and goals, which, too,
will evolve over time. They cannot know what imaginable goals are achievable,
or how to achieve them.
For well over a century, since the emergence of the modern social sciences,
many academics have been trained to conduct research as if they inhabited the first
world. This is particularly true of economics training (a discipline shared by two
of the authors of this paper) where an appealing positivist epistemological story
has dominated in the profession.4 But similar epistemic presumptions are present
in political science (disciplinary home to the other two authors) and beyond.
Social science training focuses on knowing what is knowable today or in some
proximate future. This is what methods training aspires to do—to guide scholars
in ascertaining the causal relations between some variables (xi) and other variables
(yi), which are presented as accessible to social scientists. Armed with the newest
and most sophisticated research techniques, initiates to our professions are led to
believe that their sciences are trending towards more adequate knowledge over
time. We know more today, the reasoning goes, so there is less we don’t know.
The progressive view of science assures scholars that their practice will promote
rather than undermine social betterment and that the quality of their interven-
tions will improve over time.
Over the past few decades, however, there has been increasing recognition of
the deep uncertainties that pervade the natural and social world. In part, this is
due to an important shift in ontological presumptions that guide social science
research. The quantum revolution in physics, with fundamental uncertainty at its
core, has been widely accepted. Research on the importance of intersubjectivity
and the social construction of reality has become a central part of many social
science disciplines. Scholars in the social sciences have also begun to explore a
quantum turn. Complexity theory and network theory have grown in influence
and focus on understanding dynamics and relationships, rather than only causal
claims. In this situation, as Friedrich Kratochwil and John Ruggie wrote about
regime theory in International Relations decades ago, ‘we have the most debili-
tating problem of all: epistemology fundamentally contradicts ontology!’5
We might now consider a third world—one characterized by the ontological
complexity of the second, but populated by decision-makers and scholars who
think they inhabit something like the first. Hic sunt dragones! Here, the contra-
diction between ontology and epistemology is not only a threat to productive
research, as Kratochwil and Ruggie argued: it is also a threat to productive policy
engagement. We submit that the world is made substantially more perilous by an
egregious misrecognition of a confounding world as adequately knowable. When
engaged academics mischaracterize their uncertain world as simple and accessible,
4
David Colander, ‘From muddling through to the economics of control: view of applied policy from J. N. Keynes
to Abba Lerner’, History of Political Economy 37: suppl. 1, 2005, pp. 277–91, https://doi.org/10.1215/00182702-
37-Suppl_1-277.
5
Friedrich Kratochwil and John Gerard Ruggie, ‘International organization: a state of the art on an art of the
state’, International Organization 40: 4, 1986, pp. 753–75 at p. 764, https://doi.org/10.1017/S0020818300027363.
160
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 160 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
and when they act on this mischaracterization, they risk causing gratuitous harm
as they work to promote social betterment.
Economics exemplifies the problem, as recent history illustrates. By the early
2000s a sense of a ‘great moderation’, or a reduction in macroeconomic fluctua-
tions starting from the mid-1980s, had taken hold among leading macroecono-
mists. James Stock and Mark Watson, even as they characterized it, were careful
to attribute most of this reduction to ‘good luck in the form of smaller economic
disturbances’, and they concluded with this warning to the profession: ‘the quies-
cence of the past fifteen years could well be a hiatus before a return to more turbu-
lent economic times’.6 Unfortunately, that warning was dismissed, as leading
macroeconomists sought to take credit for improved economic performance.
For instance, in 2007 the macroeconomist Christina Romer, who would soon
be appointed by President Barack Obama as chair of his Council of Economic
Advisers, congratulated the profession on the adequacy of macroeconomic
theory, just before a global crisis—which she did not anticipate—was about to
destroy innumerable livelihoods: ‘We have seen the triumph of sensible ideas and
have reaped the rewards in terms of macroeconomic performance … The costly
wrong turn in ideas and macropolicy of the 1960s and 1970s has been righted,
and the future of stabilization looks bright.’7 Romer was by no means an outlier
in what amounted to a Titanic moment for the profession. Her naive view of
the capacities of modern-day macroeconomists to pilot the economy was widely
shared just as the world economy was to be wrecked by an overlooked iceberg.
Ben Bernanke, the chair of the Federal Reserve, expressed unqualified confi-
dence when speaking about financial trends right up to the onset of the crisis. As
late as July 2008, Bernanke told all who would listen that financial markets were
adequately regulated. Just weeks before the crisis emerged, Bernanke assured a
meeting of the House of Representatives Financial Services Committee that ‘the
GSEs8 are adequately capitalized. They are in no danger of failing.’9 Two months
later the US Treasury was forced to purchase up to US$100 billion in GSE securi-
ties to prevent a global financial implosion. We are reminded in this context of the
famous quip attributed to the humourist Josh Billings, as quoted by Lucy Keeler
in 1913: ‘It ain’t so much men’s ignorance that does the harm as their knowing so
many things that ain’t so.’10

6
James H. Stock and Mark W. Watson, Has the business cycle changed and why?, NBER Macroeconomics Annual 17
(Cambridge, MA: National Bureau of Economic Research, 2002), pp. 159–218 at p. 200, https://www.nber.
org/system/files/chapters/c11075/c11075.pdf. (Unless otherwise noted at point of citation, all URLs cited in
this article were accessible on 5 Oct. 2023.)
7
Christina Romer, cited in Virginia Postrel, ‘Macroegonomics’, The Atlantic, April 2009, https://www.theat-
lantic.com/magazine/archive/2009/04/macroegonomics/307319.
8
Government-sponsored enterprises, namely the Federal Home Loan Mortgage Corporation (known colloqui-
ally as Freddie Mac) and the Federal National Mortgage Association (Fannie Mae).
9
David M. Herszenhorn and Steven R. Weisman, ‘Republican leader in US house confident in rescue plan’, New
York Times, 17 July 2008, https://www.nytimes.com/2008/07/17/world/americas/17iht-17fannie.14558589.
html.
10
Lucy Elliot Keeler, ‘My garden beasts’, The Atlantic 112, 1913, pp. 134–41, https://www.theatlantic.com/maga-
zine/archive/1913/07/my-garden-beasts/645223.
161
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 161 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel

A new epistemic awareness?

During our respective careers—now spanning three to four decades—we have
begun to see a promising shift among some members of our respective disciplines.
Today, there is greater talk, not only of uncertainty but also of its epistemic impli-
cations, than was the case when we were trained. This understanding was likely
encouraged by the rapid series of history-altering ‘black swans’ in the social world
from the 1980s forward that even the most astute social scientists failed to antici-
pate.11 The list ranges from domestic US developments (from the rise of the Tea
Party movement to the election of Donald Trump and the 6 January 2021 insur-
rection at the US Capitol Building) to international events (including the fall of
the Berlin Wall and the subsequent collapse of the Soviet Union, the extraor-
dinary rise of China as an economic powerhouse, a succession of deep financial
crises, the rise of illiberalism in democratic societies, the dramatic cultural and
political effects of social media, and the Russian invasion of Ukraine). Aware-
ness of the implications of uncertainty for what we know and how we know it
is uneven, to be sure. But we find growing appreciation of the limits to certainty
in academic writing, and in the applied work that scholars undertake beyond the
university campus.
Over the course of the past century insightful scholars have emphasized
epistemic limits even as their disciplines have struggled to repress the problem.
John Dewey and Arthur Bentley most famously emphasized these limits in their
book Knowing and the known.12 Among twentieth-century scholars engaged with
policy-making, Albert Hirschman’s work perhaps best conveys an awareness of the
dangers associated with the epistemic error of ‘knowing’ too much. Hirschman
warned against would-be social engineers, armed with ‘paradigms’ that grossly
oversimplified the economies of the global South, who readily introduced sure-
fire policy solutions to pressing problems—solutions that backfired more often
than they succeeded.13 The raw material for a different kind of engagement by
scholars promoting social betterment has been available for a long time.
Today, moving away from presuming a knowable, controllable world has
become more widely acceptable. The theologian Sharon Welch draws a distinc-
tion between the ‘ethic of control’ held by privileged communities, which posits
them as enjoying the right and the ability to bring about worlds that they value,
and the ‘ethic of risk’ operating in marginalized communities, which harbour
no illusion that they will ever know enough or that they can dictate outcomes.14
Even some traditional economists echo Welch’s sensibilities about knowledge and
control. ‘[The] world we live in is not an ergodic world’, wrote the economist

11
Nassim Nicholas Taleb, The black swan: the impact of the highly improbable (New York: Random House, 2007).
12
John Dewey and Arthur F. Bentley, Knowing and the known (Boston, MA: Beacon Press, 1949).
13
Albert O. Hirschman, ‘The search for paradigms as a hindrance to understanding’, World Politics 22: 3, 1970,
pp. 329–43, https://doi.org/10.2307/2009600. See also discussion of Hirschman’s work in Ilene Grabel,
When things don’t fall apart: global financial governance and developmental finance in an age of productive incoherence
(Cambridge, MA: MIT Press, 2017), ch. 2.
14
Sharon Welch, A feminist ethic of risk, 2nd edn (Minneapolis, MN: Fortress, 2000).
162
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 162 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
Douglass North. ‘For an enormous number of issues that are important to us, the
world is one of novelty and change; it does not repeat itself.’15
The clear implication for those seeing the world in this way is that epistemology
without space for uncertainty is a fool’s errand.16 We probe promising steps to
accept and centre uncertainty below. First, we report on interviews with a range
of publicly engaged scholars reflecting on the types of engagement and ethical
dilemmas they have faced in their applied work. Their reflections lend support
to our belief that the epistemic premise of the first world we describe above is
illusory and that self-aware scholars confront uncertainty in most, if not all, facets
of their work. We offer insights gleaned from the practice of thoughtful scholars
about how to intervene in worlds we cannot adequately know or control, and
we discuss pedagogical practices to accelerate the ongoing shift towards epistemic
self-awareness.

Learning from self-aware engaged scholars

The Sié Chéou-Kang Center for International Security and Diplomacy at the
University of Denver’s Josef Korbel School of International Studies has under-
taken a programme of research, training and outreach on public engagement,
funded since 2014 by the Carnegie Corporation of New York. As part of this
research on responsible engagement, Deborah Avant and Naazneen Barma inter-
viewed—or collected written reflections from—more than 25 engaged scholars
between 2021 and 2023 to learn about the ethical challenges they experienced and
navigated as part of their engaged scholarship related to peace and security, with
a particular focus on how uncertainty and complexity affected their engagement.
These scholars hail from a variety of disciplines in the social sciences that touch
on issues of peace and security at local, national or international levels. They are
mostly based at North American universities, many in interdisciplinary schools
of international affairs or public policy. They all indicated their interest in ethical
issues via their participation on the Sié Center’s ethical engagement panel.17 We
draw on these publicly available reflections and the interviews on which they
are based, as well as the case vignettes in our ethical engagement curriculum, as
an empirical basis for developing a framework on the types of ethical challenges
associated with publicly engaged scholarship and how scholars might navigate
challenges associated with uncertainty and complexity.
Academics who seek to inform policy-making and other forms of decision-
making do so through a variety of practices of engaged scholarship.18 Below,

15
Douglass C. North, ‘Dealing with a non-ergodic world: institutional economics, property rights, and the
global environment’, Duke Environmental Law & Policy Forum 10: 1, 1999, pp. 1–12 at p. 3, https://scholarship.
law.duke.edu/delpf/vol10/iss1/2/.
16
George F. DeMartino and Ilene Grabel, ‘Irreparable ignorance, protean power, and economics’, International
Theory 12: 3, 2020, pp. 435–48, https://doi.org/10.1017/S1752971920000263; George F. DeMartino, The tragic
science: how economists cause harm (even as they aspire to do good) (Chicago: University of Chicago Press, 2022).
17
These reflections and details on the ethical engagement curriculum and research are available online at https://
www.sieethicalengagement.com.
18
Bruce W. Jentleson and Ely Ratner, ‘Bridging the Beltway–ivory tower gap’, International Studies Review 13: 1,
163
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 163 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
we present a stylized continuum of scholarly engagement at various degrees of
remove from the ‘ivory tower’ (figure 1).

Figure 1: A stylized continuum of engaged scholarship practices

Arm’s Policy- Testimony/ One-off Ongoing Embedded

length writing/ think tank consulting consulting/ in policy-
research media work co-creation making
engagement

Even scholars who mostly inhabit the ivory tower engage indirectly with the
public sphere by addressing contemporary events at arm’s length in refereed publi-
cations. A further step is taken into the public sphere when scholars comment on
contemporary events through various media channels and writing outlets aimed at
public or policy audiences. An additional degree of engagement takes place when
scholars participate in policy-adjacent conversations—for example, by giving
public testimony, participating in think tank events, or contributing to public
commissions and think tank reports. More direct forms of engagement include
consulting with policy-makers and practitioners, either on a one-off basis or via
ongoing relationships, including those that involve the co-creation of applied
research. Finally, scholars are most directly engaged when they are embedded in
policy-making bodies, whether through termed fellowships or longer appoint-
ments to positions in government and other institutions.
We initially surmised that the nature of the challenges posed by uncertainty
could intensify over the different types of engaged scholarship. Scholars who
choose to remain in the ivory tower might think that they can sidestep the problem
of uncertainty. But they cannot know how their published research will be picked
up, interpreted, and used by advocates and decision-makers beyond the academy.
It may be simplified for use in public debate by those with agendas contrary to that
of the scholars generating the work. Network analysis of human rights advocacy
might, for example, be used by governments looking to track activists and repress
rights campaigns. Such risks can be amplified by media engagement and policy-
writing, in which scholars seek to bring public attention to their research in hopes
of affecting public discourse and policy decisions. Those who engage with the
media cannot know how people with their own agendas will interpret, repackage
and even exploit the scholarship on which their engagement is based.
Researchers giving testimony in policy-making forums, participating in think
tank events and reports, and providing arms-length consulting encounter parallel
challenges that might be intensified by having the ear of decision-makers who
sometimes wish to infer from research the findings that support their preconcep-

2011, pp. 6–11, https://doi.org/10.1111/j.1468-2486.2010.00992.x; Naazneen H. Barma and James Goldgeier,

‘How not to bridge the gap in international relations’, International Affairs 98: 5, 2022, pp. 1763–81, https://
doi.org/10.1093/ia/iiac102.
164
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 164 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
tions and preferences.19 In consulting relationships, both one-off and ongoing,
scholars also face uncertainty regarding what restrictions might be placed on them
by the client once the work is underway—especially in terms of who ‘owns’ and
how to use shared data and research outputs. In addition, scholars face uncertainty
about how findings might be framed by different parties to the research, particu-
larly if the client seeks to use the research for its own interests. This is especially
salient in contested environments. Ongoing consulting relationships may resolve
some of these uncertainties while amplifying others. As the relationship deepens,
researchers may come to understand how the client behaves and this information
could provide insight into whether the policy interlocutor is trustworthy. But
knowing that an interlocutor has problematic tendencies raises new uncertainties.
Is it better to remain in an imperfect working relationship with a client, with
opportunities to nudge them towards better decision-making, or is it better to
terminate the work? If the latter, should one blow the whistle on the client? The
researcher cannot know the effects of any of these strategies.
Scholars embedded in policy organizations face many of these uncertainties,
complicated by the additional uncertainty as to how they will manage the role
conflict that arises when holding multiple identities with competing responsi-
bilities. Those committed to their scholarly role may worry about becoming
habituated to the norms and practices of the organization in which they (newly
or temporarily) serve, and may fear making decisions that serve the organiza-
tion, rather than the broader concerns that drew them to the work initially. This
tension is aggravated by the fact that embedded scholars necessarily serve the needs
of those with whom they have contracted. It is not clear how a researcher should
think about the balance between their scholarly principles and service to a policy
organization in which they are embedded. Moreover, taking part in engaged
scholarship may give scholars new insights that change their initial perspective.
There are no simple decision rules to dictate which compromises are legitimate.
While parsing the challenges associated with different types of engagement is
worthwhile, our interviews revealed that similar dilemmas rear their heads across
the array of engagement types, and they can be eased as well as intensified by
greater proximity. Based on how the scholars with whom we spoke described
the challenges they experienced and navigated, we identified four broad concerns
appearing across the engagement continuum. Importantly, the way individual
scholars understood and responded to uncertainty was critical to the way they
managed these engagement dilemmas.

19
Stephen M. Walt, ‘International affairs and the “public sphere”’, Foreign Policy, 22 July 2011, https://foreign-
policy.com/2011/07/22/international-affairs-and-the-public-sphere; Jeremy Shapiro, Who influences whom?
Reflections on US government outreach to think tanks, commentary, Brookings, 4 June 2014, https://www.brook-
ings.edu/blog/up-front/2014/06/04/who-influences-whom-reflections-on-u-s-government-outreach-to-
think-tanks.
165
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 165 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel

The insider dilemma

The first dilemma that researchers described revolved around uncertainties
concerning whether they could be effective in pushing for change from the inside.
Some worried that moving from arms-length engagement to closer connections
to policy-makers might lead them to become complicit in perpetuating behaviour
they deemed problematic. Others, though, saw close engagement as potentially
leading to reshaping policy-makers’ perspectives in a positive way. In our inter-
views, this dilemma was sharpest among people who were involved in longer-
term relationships with their policy interlocutors and learned more about them
over time.
One climate change scholar engaged with the United States intelligence
community professed uncertainty about many aspects of climate change research.
There is little doubt that climate change is happening, and they were certain about
the importance of the issue and the need for the US government to give it more
attention in its planning. There is much less certainty, though, over what climate
change means for social systems, including violence and instability. The scholar’s
goal in engaging was to garner attention from the US government. Participating
in a scenario-building process, they felt, did succeed in raising attention to climate
change. Yet after their engagement, the scholar came to feel that the exercise
was insufficient. At the outset, they had concerns about the ethics of their inter-
locutors, and these concerns intensified during their interactions. For instance,
even though climate change has received more attention within the intelligence
community, it is addressed almost exclusively through a geopolitical lens focusing
on US security. The US intelligence community appears reluctant to consider
the impacts of climate change for different populations around the world or to
suggest that the US should be working harder to reduce its emissions. Given this
and similar issues, the scholar has not continued their engagement and would
likely not re-engage if asked. In their view, the US government is not invested in
addressing the existential threat of climate change in a productive way.
Another scholar, focused on technology and ethics, engaged with computer
scientists and engineers writing artificial intelligence (AI) algorithms for the United
States military to address the humanitarian implications of those algorithms. The
scholar was uncertain about potentially being implicated in building programmes
that would one day take lives, and they had been critical of the way that even
humanitarian discourse could justify violence. They chose to engage, however,
because they thought they could inform the writers of these programmes—who
were otherwise unfamiliar with both international humanitarian law and the
ways in which military culture can interact with machine learning in unexpected
and unpredictable ways. They hoped to guide better design that could avoid key
pitfalls and felt that those they engaged with listened and understood. They were
generally fairly satisfied with the interactions, and have maintained the relation-
ships they built.

166
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 166 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
Still another scholar spoke of concerns about complicity in the context of
giving presentations and teaching in authoritarian countries. They felt that this
kind of engagement was important for supporting curious and otherwise ‘belea-
guered’ academics in those countries and for providing information useful to
students and scholars. Because they were often engaging around subjects unrelated
to the repression in these countries, they were free to speak openly. But they
also worried that they might be legitimizing ‘window-dressing’ and helping to
‘burnish regimes’ unintentionally by just being present. Even worse, they worried
that in trying to open dialogue they could inadvertently say something that
endangered or harmed those whom they were trying to support. In discussing
how they dealt with these issues on a case-by-case basis, they reported feeling
generally comfortable with how they had managed these situations in the past.
But they were concerned that the intensification of authoritarian actions would
make this harder to manage in the future, introducing an important temporal
dimension into the calculus.
Scholars engaging with the private sector in multi-stakeholder forums also
experienced this dilemma. One engaged with businesses through the United
Nations Global Compact in an effort to encourage responsible behaviour. Their
engagement with corporate entities, they thought, led other scholars to be suspi-
cious of them and they worried about their reputation as an independent scholar.
They were also concerned that developing the relations of trust with business
leaders that were required for successful engagement could compromise their
independence. The scholar did not want to be a tool for enabling businesses to
varnish their reputations without making serious commitments to UN global
governance objectives. In the end, they felt they were able to achieve a balance
between cultivating trust and ensuring independence.
Similar concerns about complicity were expressed by scholars who had
embedded roles in policy-making and policy processes—but they were experi-
enced differently. One worked on Russia policy in the US government, beginning
in January 2022. Another worked closely with the US military collecting data
on Afghanistan. Both reported worries about complicity and their determination
to maintain an appropriate commitment to paramount concerns—the former to
avoid a third world war, and the latter to ensure their work was not used for human
targeting. The first reported their interactions with policy-makers as leading to
their own learning. As they became personally vested in Russia’s strategic defeat,
their scholarly views changed in what they saw as a positive way as a consequence
of their engagement. The second, on the other hand, very consciously resisted
being drawn into the value judgements and operational incentives of those with
whom they worked.
For the scholars experiencing the insider dilemma, their differing levels
of confidence in social science-derived knowledge on the one hand and the
quality of policy judgements on the other led to different outcomes involving
continuing—even deepening—their engagement, as opposed to ending it. The
way that different scholars described how they experienced and navigated this

167
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 167 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
dilemma reveals the particular importance of scholars’ sense of the certainty of
their own knowledge, compared to their impression of the knowledge and ethics
of their policy interlocutors.

The integrity dilemma

The second dilemma our engaged scholars reflected on involved navigating situa-
tions where policy actors pressed for a spin on research findings that reflected their
political imperatives as opposed to what scholars saw to be an honest reporting of
those findings. The problem arose both in one-off consultancies and in embedded
settings.
For one scholar, this arose as they arrived in Iraq to assess how humanitarian
workers navigate the ethics of working in conflict zones. They had established
terms with the regional head of the organization with which they were working.
Yet interference from the project manager on the ground, combined with poor
project design, convinced them that they had no option other than to disengage
entirely. The researcher specifically expressed the concern that the organization
they were working with was ‘deliberately asking questions in a way that does not
allow any uncertainty or ambiguity or tension to come out of it … they knew
well enough how to ask questions so that … they were just going to create the
right data that gave the response [they wanted]’. The organization’s unwillingness
to embrace uncertainty in the research was a red flag for the scholar. Another
experienced pressure to produce stronger findings than could be warranted given
the time and resources available.20 The pressure on this researcher came not from
the policy client, but from a third-party consultant the client had hired to manage
research. Though they were able to work through to a satisfying result, they noted
that it required a willingness to understand and work within a complex set of
relationships.
Two others experienced pressures to shift their conclusions as they presented
their research findings. One talked of this ranging from a fairly innocuous request
to delete one paragraph to pressure to shift emphasis from less to more favour-
able findings, to asking that they assert results that were not supported by their
research. The second described a more complicated set of issues when they were
contracted by an international initiative on business and human rights. The scholar
viewed their role as providing information relevant to this initiative and the policy
it favoured, but the initiative saw the engagement as validating the policy path
they were already set on. When the scholar publicly presented research relevant
to, but questioning parts of, the existing policy, the policy audience was disap-
pointed and ended the engagement. While confident in their findings, the scholar

20
The issue of unreasonable time constraints forcing scholars to reach judgements prematurely, based on
inadequate data and analysis, was raised almost universally by engaged economists whom George DeMar-
tino surveyed. George F. DeMartino, The economist’s oath: on the need for and content of professional economic
ethics (Oxford: Oxford University Press, 2011). See also William R. Allen, ‘Economics, economists, and
economic policy: modern American experiences’, History of Political Economy 9: 1, 1977, pp. 48–88, https://
doi.org/10.1215/00182702-9-1-48.
168
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 168 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
also wondered whether better communication about their role would have led to
a more productive engagement.
Yet another scholar reflected on embedded work with an NGO in Afghan-
istan to set up an experiment on the impact of cash pay-outs on government
legitimacy. Although the results showed no improvements to legitimacy, people
working with the NGO pointed to many changes in the country that could still
have contributed to positive change. The scholar’s certainty about the experiment
design led them to believe that its results should have been more disruptive and
demonstrated that the policy was not working.
For the scholars who discussed the integrity dilemma, a key strategy for
managing it was acknowledging the possibility of competing research and organi-
zational imperatives. Establishing incentives and clear lines of communication, as
well as a willingness to share accountability and publicly communicate the impli-
cations of findings with a view to the organization’s interests, can help to soften
this dilemma. One potentially useful strategy for scholars that extends beyond
good communication is pre-commitment, in which they establish and make clear
to prospective clients just what they are and are not prepared to do in their engaged
work.21 Some scholars have adopted their own personal codes of conduct, which
they make available to prospective clients.

The complexity dilemma

The third dilemma involves policy-makers turning to academics for simple, declar-
ative answers that repress or ignore uncertainty even as engaged scholars recognize
that reality is much more complicated. Many respondents reported seeing more
complexity than could easily be addressed. This dilemma arose in arms-length
engagement and in policy-writing, as well as in longer-term engagement with
think tanks and policy communities.
One scholar describes encountering this throughout their career and writing
specifically to counter simple narratives as a way of establishing room for more
complexity in policy debates. Another similarly referenced the difficulty of even
talking about the ethics of conflict, given new practices and technologies that
enable the lethal use of force outside the traditional war zones to which just war
debates apply. They described their struggle to be explicit about moral contradic-
tions in their work. To offer constructive guidance and not simply criticism, they
developed ‘two guiding assumptions … about what limited force should look
like’ to build in restraints and ‘to stave off some of the potential abuses’. As they
worked in this area, they felt one of their greatest contributions was in devel-
oping a language that was more productive in speaking about the ethical issues
surrounding the use of drones. A third scholar discussed how language is crucial in
engaging on responses to mass atrocities. They advocate, based on their research on
the Rwandan genocide and recovery from it, for the use of first-person language
instead of labels like ‘genocidaire’ or ‘perpetrator’, thus foregrounding the evil
21
DeMartino, The economist’s oath.
169
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 169 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
of an act instead of the evil of a person. They reflected how people working in
spaces that have experienced atrocity ‘are often deeply uncomfortable’ about such
language changes and that, in their experience, such recommendations must be
made with caution and delicate considerations around context.
Another reflection brought up how the training we give to graduate students
at professional policy schools affects the language and the concepts they use in
their careers. When the scholar in question first began working on cyber security,
they found many policy-makers speaking about norms as substantial entities that
could be used instrumentally. The idea of norms as an intersubjective practice
was something unfamiliar to many trained in the rationalist language surrounding
neo-liberalism and neo-realism. In much of their engagement, the scholar
described a ‘shared uncertainty’ about the nature of a problem ‘still up for grabs’.
They described the policy-makers they worked with as concerned to ‘get up to
speed’ on the issue at hand and to ‘get people to take the issue seriously’. They
navigated these interactions by meeting those in the policy community on their
own terms and trying to discover the problems they were grappling with.
A similar strategy was used by a researcher to help address ‘blind spots’ as they
engaged with practitioners on identity politics in the wake of the filmed killing
of George Floyd in May 2020 by a police officer following the former’s arrest
in Minneapolis. In helping people ‘unlearn’ common misunderstandings, the
researcher found that using academic language and social science tools like careful
argument and evidence helped tamp down polarized reactions. They argued that
‘trying to add different layers by giving new forms of information or new ways
of thinking about data and as an intersectional black feminist … other ways of
coming at the problem’ were productive strategies for encouraging people to see
more complexity.
Two researchers reflected on encouraging more complex and nuanced views
among policy communities with which they were already deeply involved. One
understood this as posing reputational difficulties for them if they moved too
far away from entrenched perceptions too quickly. They carefully thought about
how to present their work on a ‘global NATO’ to nudge the policy community
towards a more complex, nuanced view of NATO but without seeming too far
out of the mainstream consensus to be taken seriously. One of their strategies
was based on the realization that a particular recommendation might never be
accepted but ‘the point in writing was to say we have to think about this issue
in this institution in a different way’. The second scholar worked hard to create a
complex, detailed report on the future of Afghanistan and the different options
for US policy involving many stakeholders. They described using simple versions
of formal models of strategic interaction to explore with some degree of certainty
what would likely happen when the US withdrew its troops from the country.
They also explicitly used scenarios to identify the more uncertain elements of
potential pathways and the various implications. These efforts were ultimately
unsuccessful; the report had little effect given incoming President Joe Biden’s
more simple, pre-formed narrative on Afghanistan.

170
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 170 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
For scholars who reflected on the complexity dilemma, strategies for managing
it included paying deliberate attention to language and context, as well as focusing
on understanding the goals of their policy interlocutors. These scholars also
manifested a willingness to engage over the long term with a policy community
to help shape the language and goals over time. There is good reason to expect
that most research will not generate the kind of clarity and certainty often desired
by policy-makers and advocates. Conveying informed uncertainty effectively—
by emphasizing context, contingency and complexity—is an important skill for
engaged scholars.22

The conflict-of-interest dilemma

The fourth dilemma that our respondents raised was around how engagement can
present actual or apparent conflicts of interest. Some of the expressed concerns
revolve around funding and other dimensions around ownership of research
findings. Other concerns surround ethical interaction with local research partners
and interlocutors during field research. This dilemma arose mostly in arms-length
engagement, but also in other types of engaged research.
Accepting funding for research was the first manifestation of this dilemma.
A scholar pointed out the obvious problem of doing research when funders
expect or demand a particular result, but acknowledged that most situations
are more nuanced. The same scholar also noted that the source of funding can
raise concerns. Government funding—for example, from the US Department of
Defense—can be questionable to some scholars, whereas for others it serves as a
vote of confidence in one’s research from a powerful actor. In the case of private
foundations, the source of the funding that initially capitalized the institution can
also introduce dilemmas for scholars. Some of the oldest foundations in the US
were founded on the wealth of so-called robber-barons who used unsavoury and
exploitative practices to amass their wealth. Should their actions raise concerns for
scholars accepting their funding? Or does the time elapsed since the foundation was
capitalized render these concerns less important, especially if the funding is now
being used to support research that attempts to serve the public good? For founda-
tions that have been capitalized more recently, though, it may be more difficult to
look away from the practices that generated the resources to the research that is
being supported. This has been a particular concern given inflows of money from
ideologically driven groups or organizations that are also funding contemporary
political projects. Ultimately, the scholar who raised concerns about the source
of funding encouraged other researchers to think through their own boundaries
and try to reserve judgement about others’ choices. Another researcher worried
that ongoing consulting with commercial enterprises could be seen as a conflict of
interest. One of their key strategies was to ensure distance from corporate entities
by refusing to accept any corporate funding.

22
Barma and Goldgeier, ‘How not to bridge the gap’, p. 1773.
171
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 171 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
Other concerns revolve around how the research itself is conducted, and
whether a scholar has ownership over any part of the research in the event of
conflict with a client. One of our interviewees pulled out of their consultancy, but
was nonetheless able to use some of the research generated with the approval they
had secured from their university’s institutional review board. The complications
of withdrawing from a research consultancy could be especially consequential for
junior scholars who cannot afford to forgo the publication of research in which
they have invested time. Another scholar emphasized how crucial it is to invest
time and care in pre-establishing, via written contracts specifically, ownership
over research design, findings, and products.
Conflict-of-interest dilemmas in democracy research conducted with local
interlocutors also came up in our interviews. One scholar emphasized that the
imperative to get the right research setting, and to choose cases based on consis-
tent theoretical criteria, must be balanced against the necessity of gaining access
and the responsibility to do this in a way that is safe for local research partners.
A second scholar similarly expressed concerns about exposing local respondents,
even anonymized, to potential retaliation by repressive states. They also discussed
whether it was important to adjust language to gain research access to local inter-
locutors. For instance, the researcher wondered whether they should abandon
politically weighted concepts like ‘democracy’. Although they decided against this
strategy to speak clearly to American policy-making audiences, they acknowl-
edged the stresses this placed on local partners who necessarily worried about
being seen as complicit in an American foreign policy agenda.
Scholars for whom the conflict-of-interest dilemma emerged spoke in different
ways about the importance of creating distance from certain sources of funding,
and of having clear terms of reference vis-à-vis research counterparts and other
stakeholders. Several of the reflections point to how relationships between
scholars, interlocutors and other stakeholders (for example, funders) shift over
time, sometimes to the benefit of scholars.
As is clear from this brief discussion of four general dilemmas of engaged schol-
arship, our respondents—often explicitly—engage uncertainty and complexity in
many aspects of their applied research and policy engagement. We emphasize that
this pool is unlikely to be representative of engaged scholars more generally—in
fact, we chose these respondents in part owing to our expectation that they would
convey useful insights about the ethics of engagement in uncertain and complex
conditions. We now turn to the matter of how different approaches to uncertainty
might inform responses to these dilemmas, as well as to the general lessons we
draw from our investigation.

Reflections on engaged scholarship in a fourth world

We began this paper by imagining three worlds characterized by varying epistemic
conditions. We now imagine a fourth world, in which scholars and the decision-
makers they serve recognize and explicitly centre epistemic limits in their work

172
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 172 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
given the fundamental complexity of the world in which they live.23 Here,
scholars fit their epistemic expectations more closely with emerging ontological
presumptions. They recognize that strategies implemented even by well-trained
and virtuous actors will have unintended effects, some of them harmful—and that
pursuing social betterment means taking on responsibility for anticipating and
preparing for unwelcome consequences. Scholars, along with those they study
and support, thus explore how to navigate responsibly the risks associated with
acute limits to the known and knowable.
One could infer that the appropriate response to uncertainty is to avoid engage-
ment with the world for fear of causing harm. We do not hold that view. It
is a truism that not acting can also cause harm, especially when the status quo
involves deprivation and misery. Instead, we explore in this section how engaged
scholars can act responsibly in pursuit of social betterment, even in the face of
the unknown and unknowable. We present three lessons that cut across our inter-
views—accept deep uncertainty; engage with stakeholders; and value descrip-
tion, narrative and fantasy—and then discuss the Decision Making under Deep
Uncertainty (DMDU) approach to uncertainty and complexity in confrontation
with wicked problems. We offer these as starting-points for future research and
conversation about engaged scholarship in the fourth world that scholars who
acknowledge uncertainties are attempting to build.

Accept deep uncertainty

Economics is particularly illustrative of a shift towards appreciation of deep uncer-
tainties, if only because over the past century economists were trained to believe
that economic science was largely adequate to the task of knowing and, hence,
controlling the economy. Abba Lerner’s The economics of control, published in 1944,
sent generations of economists into practice believing they could achieve the
unachievable.24
Economists and political scientists today are expressing a new humility about
the capacities of their respective sciences to know and control. Since the 2008
crisis, it is not uncommon to find it in economists’ statements—like that of Peter
Orszag, Robert Rubin and Joseph Stiglitz, who write: ‘In our collective experi-
ence, fiscal policy should instead be informed by copious amounts of humility,
particularly given the role of impossible-to-predict events (including pandemics,
wars, and bubbles).’25 Naazneen Barma’s book on peacebuilding explores the
23
Our characterization of these four worlds is a heuristic device, distinct from the positing of alternative poten-
tial worlds that is common in scenario analysis. We note, however, that others have used the concepts of
scenario analysis to similarly explore the evolving landscape in which publicly engaged scholarship takes
place. For example, see Jordan Tama, Naazneen H. Barma, Brent Durbin, James Goldgeier and Bruce W.
Jentleson, ‘Bridging the gap in a changing world: new opportunities and challenges for engaging practitioners
and the public’, International Studies Perspectives 24: 3, 2023, pp. 285–307, https://doi.org/10.1093/isp/ekad003.
24
Abba Lerner, The economics of control (London: Macmillan, 1944); David Colander, ‘Muddling through and policy
analysis’, New Zealand Economic Papers 37: 2, 2003, pp. 197–215, https://doi.org/10.1080/00779950309544384.
25
Peter R. Orszag, Robert E. Rubin and Joseph E. Stiglitz, Fiscal resiliency in a deeply uncertain world: the role of
semiautonomous discretion, Peterson Institute for International Economics Policy Brief no. 2 (Washington DC:
Peterson Institute for International Economics, 2021), p. 2, https://www.piie.com/publications/policy-briefs/
173
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 173 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
fundamental unknowability and endogeneity of what might happen with any
attempt at socio-political engineering. Barma delivers a critique of the positivist,
probabilistic mainstream that purports to know certain things within some bounds
of confidence.26 Séverine Autesserre goes even further, questioning how social
scientists have conceptualized peace as well as how it unfolds and what scholars
and policy-makers can do to promote it. In a way reminiscent of sociologist Bruno
Latour, Autesserre follows the actors, taking note of what people on the ground
in different conflicts understand by peace and what they hope for from those who
would help them achieve it.27
The engaged scholars we interviewed exhibit an awareness of epistemic limits,
and the need for humility. The scholar we mention above whose experience refer-
ences the way that their work in the US government shifted their perspective
on the importance of defeating Russia in Ukraine provides one example. In the
ethical engagement curriculum, Deborah Avant describes how parroting what
political scientists thought they knew about the potential for regulating private
security blinded her to what turned out to be promising regulatory pathways.28
Her experience of participating in the multi-stakeholder efforts to rein in this
industry has made her much more attentive to different processes through which
governance can develop, which had been largely overlooked by the profession.
Innovative scholars are now pressing much further on the matter of epistemic
limits. Some of these scholars are at the forefront of the complexity approach to
social science. Brian Arthur, a leader in the field at the Santa Fe Institute, empha-
sizes dynamic features of the economy that present formidable limits to economic
knowledge. He writes: ‘[T]he economy is not something given and existing but
forms from a constantly developing set of institutions, arrangements, and techno-
logical innovations’.29 While standard ‘equilibrium’ economics ‘emphasizes order,
determinacy, deduction, and stasis, complexity economics emphasizes contingency,
indeterminacy, sense-making, and openness to change’.30
Complexity economics is generating tentative conclusions about causality that
deepen awareness of ineradicable uncertainties. Arthur describes the method this
way: ‘Complexity economics thus sees the economy as in motion, perpetually
“computing” itself—perpetually constructing itself anew.’31 What becomes of

fiscal-resiliency-deeply-uncertain-world-role-semiautonomous-discretion.
26
Naazneen H. Barma, The peacebuilding puzzle: political order in post-conflict states (Cambridge, UK: Cambridge
University Press, 2017). See also George F. DeMartino, Ilene Grabel and Ian Scoones, ‘Economics for an
uncertain world’, unpublished paper (2023).
27
Séverine Autesserre, The frontlines of peace: an insider’s guide to changing the world (Oxford: Oxford University Press,
2021); Deborah Avant, ‘Introduction’, International Security Studies Forum Roundtable 14-2 on Autesserre,
Frontlines of peace, H-Diplo, 21 Oct. 2022, https://networks.h-net.org/node/28443/discussions/10965072/h-
diplo-issf-roundtable-14-2-avant-autesserre-frontlines-peace; Bruno Latour, Reassembling the social: an introduc-
tion to actor network theory (Oxford: Oxford University Press, 2005).
28
The ethical engagement curriculum is available at https://www.sieethicalengagement.com.
29
W. Brian Arthur, Complexity economics: a different framework for economic thought, Santa Fe Institute Working
Paper #13 (Santa Fe, NM: Santa Fe Institute, 2013), p. 1, https://www.santafe.edu/research/results/working-
papers/complexity-economics-a-different-framework-for-eco.
30
W. Brian Arthur, Complexity and the economy (New York: Oxford University Press, 2013), p. 19, emphasis
added.
31
Arthur, Complexity and the economy, p. 1.
174
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 174 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
causal explanation under this approach? Explanation takes the form of a ‘solution’
that seeks to identify ‘a pattern, a set of emergent phenomena, a set of changes
that may induce further changes, a set of existing entities creating novel entities’,
where the goal is not the discovery of universally applicable theorems, ‘but the
deep understanding of mechanisms that create these patterns and propagations of
change’.32 Central to this approach is the emphasis on non-equilibrium, which
emerges not just from external shocks but also endogenously, from circumstances
and processes that define the economy.
The consequences of this world-view for theorists and engaged scholars are
dramatic. Here, decision-makers confront a world characterized by complex
systems and wicked problems. In this kind of world, the idea of finding optimal
strategies and outcomes is quixotic. Arthur notes:
To the degree that outcomes are unknowable, the decision problems they pose are not
well-defined. It follows that rationality—pure deductive rationality—is not well-defined
either, for the simple reason that there cannot be a logical solution to a problem that is not
logically defined.33

We are not suggesting that causal logic is of no use, but that the domain in
which it provides adequate guidance is not always well defined in advance, and
so the use of causal models can mislead. In the excitement of modelling causality
with an eye to confronting an important policy challenge, it is easy to lose sight
of the fact that our models necessarily engage assumptions and are therefore
always fallible. Scholars should thus foreground humility and alert stakeholders to
uncertainty about the assumptions they employ. Among twentieth-century social
scientists, perhaps no one better grasped the point than Albert O. Hirschman. As
the epigraph that begins this paper demonstrates, Hirschman warned of the poten-
tially problematic fallout from the social science hunt for, and implementation of,
comprehensive programmes of social engineering in what he understood to be an
ineluctably complex social world.

Engage with stakeholders

Foregrounding uncertainty underlines the scholarly mission to facilitate collab-
orative exploration with stakeholders that seeks to reveal possible causal connec-
tions and possible futures, rather than to impose a scientifically warranted singular
truth on them. We see this in several reflections on the process of trying to change
things from the inside, where the scholars sought to help decision-makers shake
off narrow preconceptions that blocked learning and consideration of new ideas.
The results of our respondents’ efforts were mixed in this regard. One scholar was
not sure their interlocutors were as open to learning as they had hoped. Another,
32
Arthur, Complexity and the economy, p. 19.
33
Arthur, ‘Complexity economics’, p. 4. Behavioural economists, too, have hammered home the idea that real
agents in the economy are not rational in the way presumed by twentieth-century neo-classical econom-
ics. But the approach does not open the door to uncertainty. Instead, the methods are designed to identify
‘predictable’ patterns of behaviour, as the title of Dan Ariely’s influential text indicates. Dan Ariely, Predictably
irrational: the hidden forces that shape our decisions (New York: HarperCollins, 2008).
175
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 175 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
though, found those they engaged with were more open to learning. Still another,
similarly, found the businesses they engaged with were open to learning about
their potential public roles. One scholar’s reflection describes their engagement
as akin to therapy sessions, helping policy-makers think though their respective
concerns. Several pointed out the way that academic language can help people find
some space from their initial political reactions, which can allow greater potential
for learning and exchange. But the exchange must be seen as two-way. One reflec-
tion centres a growing trend towards applied research ‘co-created’ by scholars
and practitioners, whereby the desire to engage responsibly leads to partnered
approaches to research design, conduct and dissemination.34
This kind of interaction may be more readily available to scholars who engage
with policy-makers more closely. Though we initially surmised that ethical
dilemmas would compound with closer engagement, in some instances greater
collaboration can ease tensions. More embedded forms of engagement might
lead to a more fluid learning dynamic between scholars and policy interlocutors,
thereby making it easier to navigate tensions than in more isolated instances of
engagement where the scholar–interlocutor relationship does not build over time.

Value description, narrative and fantasy

An important shift is under way in how we are urged to think about our social
scientific enterprises. We refer here to the shift in orientation towards social
science as the construction of ‘narratives’ rather than the discovery of unassail-
able truths. Centring uncertainty suggests a rebalancing of attention to appre-
ciate the importance of ‘description’ to both useful theorization and to productive
academic engagement.35
Narratives drive how people understand their experience and thus whether
descriptions are seen as accurate. Narratives drive agents’ engagement with the
world, shaping their beliefs and behaviours in directions that do not accord with
more abstract notions, such as the rational actor model which has grounded
economics and much of political science for over a century.36 Contemporary
scholars have rediscovered the insights of economists like G. L. S. Shackle, who
argued (rightly in our view) that economics should be understood as a study of
ideas, not of objective categories like prices and GDP. According to Shackle:
‘Economics is about thoughts. It is therefore a branch or application of epistemics,
the theory of thoughts.’37 Dani Rodrik picks up the thread: ‘Yet without ideas …
34
See Susanna P. Campbell, ’Dapo Oyewole and Haley J. Swedlund, ‘Bridging the gap between research and policy:
lessons from co-creation in the aid sector’, The Duck of Minerva, 3 June 2023, https://www.duckofminerva.
com/2023/06/bridging-the-gap-between-research-and-policy-lessons-from-co-creation-in-the-aid-sector.html.
35
Deborah Avant, ‘The role of description’, in Jennifer Cyr and Sara Wallace Goodman, eds, Doing good qualita-
tive work (Oxford: Oxford University Press, 2023 forthcoming).
36
Deirdre N. McCloskey, The rhetoric of economics (Madison, WI: University of Wisconsin, 1985); Robert J. Shiller,
Narrative economics: how stories go viral and drive major economic events (Princeton: Princeton University Press, 2019);
Jens Beckert, Imagined futures: fictional expectations and capitalist dynamics (Cambridge, MA: Harvard University Press,
2016); and Frederick W. Mayer, Narrative politics: stories and collective action (New York: Oxford University Press, 2014).
37
G. L. S. Shackle, Epistemics and economics: a critique of economic doctrines (reprint, New York: Routledge, 1992 [first
edn 1972]), p. xx.
176
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 176 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
the concept of self-interest is empty and useless … In truth, we don’t have “inter-
ests”. We have ideas about what our interests are.’38 Pragmatists concur, turning
the story of interests and ideas on its head by arguing that neither are set ahead of
time, but are settled in specific situations.39
These scholars suggest how academic narratives not only describe the world
but shape it—often in unpredictable and deeply performative ways.40 Notably,
both Rodrik and Shiller assign some blame for the financial crisis of 2008 to the
stories economists told about the efficiency of financial markets. Shiller writes:
This mania was the product not only of a story about people but also a story about how the
economy worked. It was part of a story that all investments in securitised mortgages were
safe because those smart people were buying them … To a remarkable extent we have got
into the current economic and financial crisis because of a wrong economic theory—an
economic theory that itself denied the role of the animal spirits in getting us into manias
and panics.41

Nor should we forget the relevance of imagination. Indeed, making causal

claims entails specifying counterfactual claims. As Philip Tetlock and Aaron
Belkin claim:
We can avoid counterfactuals only if we eschew all causal inference and limit ourselves
to strictly noncausal narratives of what actually happened (no smuggling in causal claims
under the guise of verbs such as ‘influenced’, ‘responded’, ‘triggered’, ‘precipitated’, and
the like).42

But counterfactual worlds exist only in our imaginations—and so our causal

claims about our actual world depend on the veracity of claims made about imaginary
worlds that exist only in our minds. For instance, the overworked question, ‘Would
the First World War have happened had the Archduke Ferdinand not been assas-
sinated in Sarajevo in June of 1914?’ can be adjudicated only by specifying what
would have happened in an alternative world in which the Archduke was not
assassinated. But of course, that alternative world was foreclosed the moment that
the assassination occurred—it can exist, then, only in our imaginations.43 How we
imagine our world can be important for the options we entertain for dealing with

38
Dani Rodrik, Straight talk on trade: ideas for a sane world economy (Princeton: Princeton University Press, 2017):
pp. 159, 163 (emphasis in original); cf. Frank H. Knight, Risk, uncertainty, and profit (Boston and New York:
Houghton Mifflin Company, 1921), ch. 7; Shackle, Epistemics and economics.
39
Louis Menand, The metaphysical club: a story of ideas in America (New York: Farrar, Straus and Giroux, 2001),
p. xi.
40
Other scholars who have contributed much to our understanding of the performative force of social science
narratives include Albert O. Hirschman and, more recently, J. K. Gibson-Graham, and Stephen Resnick and
Richard Wolff.
41
Robert Shiller, ‘A failure to control animal spirits’, Financial Times, 12 May 2009, https://www.ft.com/
content/453e55ca-0c0c-11de-b87d-0000779fd2ac. See also George F. DeMartino, ‘The economic crisis and
the crisis in economics’, in Martha Starr, ed., Consequences of economic downturn: beyond the usual economics (New
York: Palgrave Macmillan, 2011), pp. 25–44.
42
Philip E. Tetlock and Aaron Belkin, eds, Counterfactual thought experiments in world politics: logical, methodological,
and psychological perspectives (Princeton: Princeton University Press, 1996), pp. 3–4.
43
For a more detailed examination of the role of fictional counterfactuals in the hunt for causality in economics,
see George F. DeMartino, ‘The specter of irreparable ignorance: counterfactuals and causality in economics’,
Review of Evolutionary Political Economy, vol. 2, 2021, pp. 253–76, https://doi.org/10.1007/s43253-020-00029-w.
177
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 177 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
it, as one of our interviewees recounts in their reflection on trying to encourage
a broader view of what NATO could be.

One productive model: DMDU

Economist Edward Leamer brings together these lessons in the introduction to his
macroeconomic textbook, where one would normally expect a stout defence of
the scientific credentials of economics, which for a century held itself to be the
physics of the social world. He instead speaks of economics as storytelling.
You may want to substitute the more familiar scientific words ‘theory and evidence’ for
‘patterns and stories’. Do not do that. … The words ‘theory and evidence’ suggest an inces-
sant march towards a level of scientific certitude that cannot be attained in the study of the
complex self-organizing human system that we call the economy. The words ‘patterns and
stories’ much more accurately convey our level of knowledge, now, and in the future as
well. It is literature, not science.44

It is hard to overstate just how fundamental is the shift that Leamer’s claims
represent. Today’s epistemically self-aware economists are projecting an image
of economics as a far humbler intellectual enterprise that seeks to promote social
betterment in a world that defies adequate understanding. Leamer and others
warn us against expecting social science wizards to know the unknowable, or to
control the uncontrollable.
Engaged scholars and practitioners have begun to put this advice into practice.
One community of practitioners is leading the way—those engaged in what
is called Decision Making Under Deep Uncertainty, or DMDU.45 DMDU
foregrounds the epistemic problem we have explored here. It presumes deep
uncertainty regarding the nature of the systems that drive events in the world (and
how those diverse systems interact), the models that are constructed to map those
systems, the agents who will be affected by policy choices and how those agents’
values will evolve over time, and the innumerable contingent factors that condi-
tion the world in which any policy decision will take effect. It therefore steers
clear of the standard approaches by which engaged scholars often seek to inform
decision-making. DMDU rejects the ‘predict then act’ model that characterizes
standard policy advocacy, which seeks to find the optimal policy given a probabi-
listically knowable future. Instead, it looks to discover policies that are apt to be
‘robust’ across a very wide range of possible futures. The strategy involves testing
particular policy proposals via simulations against many thousands of possible
future worlds, each characterized by adjustments in the inputs, parameters and

44
Edward E. Leamer, Macroeconomic patterns and stories (Heidelberg: Springer Berlin, 2009), p. 3 (emphasis in
original).
45
The reader interested in learning about this approach should consult Vincent A. W. J. Marchau, Warren E.
Walker, Pieter J. T. M. Bloemen and Steven W. Popper, eds, Decision making under deep uncertainty: from theory
to practice (Cham, Switzerland: Springer, 2019) and the website for the Society for Decision Making Under
Deep Uncertainty, https://www.deepuncertainty.org.
178
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 178 12/18/23 9:53 AM

 The ethics of engaged scholarship in a complex world
functional forms of the models used to run the simulations—while refusing to
assign probabilities to those possible futures.
A central element of DMDU is the emphasis placed on incorporating stake-
holders—those who will be affected by adopted policies—directly and meaning-
fully into policy deliberations. Stakeholders are involved with DMDU experts
in consultations over possible policy solutions to pressing problems, providing
expertise about what are taken to be the relevant systems that will bear on a policy’s
effects, assessing policy options for their apparent robustness, and choosing which
risks to take in pursuit of which social goals. In a world of adequate knowledge,
this engagement would be inefficient and unnecessary. But in a world character-
ized by deep uncertainty, stakeholder engagement elicits the tacit knowledge that
diverse actors carry, without which decision-makers are apt to make avoidable,
damaging mistakes. Stakeholder engagement also conveys respect for those who
stand to be harmed by any decisions taken. Rather than the economist deciding
for the community which risks of harm are appropriate to impose on some for the
benefit of others, here the stakeholders themselves make the critical judgements.
Once a chosen policy is implemented, the same community of experts and
stakeholders then monitors the world as it evolves and looks to adjust policy to
enhance policy robustness. Here there is no ‘end’ to policy engagement in confron-
tation with a wicked problem. Instead, a diverse community of actors remains
engaged in monitoring and adjustment. Finally, the approach involves responsi-
bility on the part of the DMDU experts. Unlike in arms-length, one-off engage-
ment, such as when an economist is hired to produce a cost–benefit analysis of a
particular policy dilemma, here the expert remains engaged. Experts have what
Nassim Taleb calls ‘skin in the game’, where they must confront the consequences
of their decision and engage directly with stakeholders suffering the consequences
of prior decisions.46 That kind of engagement is apt to overcome the problem of
disengagement that too often leads to the construction of the ‘blasé’ who sleep
well at night while others face harm.47
DMDU is not a panacea—it cannot tame wicked problems and it cannot
eliminate risk of severe damage from decisions taken when confronting them. Its
chief virtue is ethical. DMDU practice confronts honestly, with stakeholders and
decision-makers, the limits to expert knowledge; and it replaces a paternalistic
ethic in which the expert is thought to know best with an ethic that emphasizes
the integrity and agency of those whom experts purport to serve.

Conclusion—fourth-world pedagogy
We are advocating here for the fourth world, in which scholars and decision-
makers recognize that they confront deep uncertainty. As our interviews with
engaged scholars and our discussion of recent literatures and strategies reveal, we

46
Nassim Nicholas Taleb, Skin in the game (New York: Random House, 2018).
47
The concept of the blasé self appears in Andrew Linklater, The problem of harm in world politics: theoretical inves-
tigations (Cambridge, UK: Cambridge University Press, 2011), p. 226.
179
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 179 12/18/23 9:53 AM

 Deborah Avant, Naazneen H. Barma, George F. Demartino and Ilene Grabel
have many ingredients for better understanding the limits to academic (and other)
expertise and the bounds of our ability to control a complex world. We find many
epistemically aware scholars and practitioners shifting gears towards new kinds
of engagement that foreground uncertainty and that seek to engage stakeholders
in partnerships where they can help navigate policy decision-making in a world
where even robust decisions might cause harm.
But we have been on this precipice before. Pragmatic scholars like John Dewey,
Jane Addams, William James and Charles Peirce made many of these arguments
at the beginning of the twentieth-century. Hirschman’s work, too, was epistemi-
cally and ethically self-aware.48 And yet, our professional practice and public
discourse often presume adequate scholarly knowledge of a knowable world. Our
textbooks, our methods training, our professional literatures together conspire
to deliver a message that we know the unknowable and we are rewarded richly
for this illusion—in policy influence, privileges of professional self-governance,
income and status. Each of these factors represents a powerful obstacle to the
transformation from our third world, in which scholars confront an opaque world
but think they confront a knowable world, to the fourth world we advocate here.
Where are we to begin? Moving towards this fourth world requires shifts in
how we train and reward new scholars and, equally, how we train students heading
for non-academic careers who will use the language of scholarship in their profes-
sional lives. Fourth-world pedagogy would teach what Wendell Berry describes as
the ‘way of ignorance’.49 The pedagogy would foreground questions concerning
what we do not know—and perhaps cannot ever know—despite our best efforts
and the best research techniques. That pedagogy would tease out the implications
of irreparable ignorance for what scholars can and cannot deliver in pursuit of
social betterment.50 Equally important, that pedagogy would instruct current and
future decision-makers and other stakeholders about what should and should not
be asked of scholars. Moving to this kind of pedagogy entails the unwelcome but
necessary task of pushing back against the intoxicating presumption of knowing
the unknowable and controlling the uncontrollable. For us, and for many other
scholars today, making that transition is a pressing ethical duty as we seek to act
responsibly in an inscrutable world.

48
It is striking how much Hirschman anticipates the recent turn in economics away from theorizing the econ-
omy as an essentially simple, self-contained system towards recognizing it as an adaptive, complex system. See,
for example, Wolfram Elsner, ‘Complexity economics as heterodoxy: theory and policy’, Journal of Economic
Issues 51: 4, 2017, pp. 939–78, https://doi.org/10.1080/00213624.2017.1391570.
49
Wendell Berry, The way of ignorance: and other essays (Emeryville, CA: Shoemaker & Hoard, 2005).
50
DeMartino, The tragic science.
180
International Affairs 100: 1, 2024

INTA100_1_FullIssue.indb 180 12/18/23 9:53 AM