The Psychology of Nuclear Brinkmanship

The Psychology of
Nuclear Brinkmanship

Reid B. C. Pauly and
Rose McDermott

Although the role of
psychology and human emotions has increasingly made inroads into our under-
standing of the micro-foundations of decision-making in many areas of conºict
études, much less work has examined how they inºuence nuclear strategy. Ce
article highlights the critical role of these factors in nuclear brinkmanship. Comme
the ªeld of international relations seeks to update theories of nuclear strategy
and the utility of nuclear weapons in the wake of the notable return to great
power conºict, we highlight the centrality of the conscious choice that leaders
must undertake in deciding to use nuclear weapons. Human emotions can
introduce chance into bargaining in ways that contradict the expectations of
the rational cost-beneªt assumptions that undergird deterrence theory.

This article brings the study of psychology and emotion to bear on the
central puzzle of how “chance” can generate coercive leverage in nuclear cri-
ses while leaders still retain agency over the “choice” to escalate. Leverage is
the ability to generate power or inºuence over an adversary to get them to
bend to your will. Nuclear threats can produce leverage if made credibly. But a
rational decision-maker should never choose mutually assured destruction
(MAD); therefore canonical theories of brinkmanship assume that the “threat
that leaves something to chance” removes leaders from the process of escala-
tion.1 We argue instead that chance can coexist with choice. In a MAD world, it
is indeed irrational to carry out a nuclear threat if massive nuclear retaliation is
expected; but a human decision-maker acting on emotion or psychological
bias might do so anyway. Psychological factors explain how brinkmanship can
operate even when leaders retain control over their nuclear forces.

Reid B. C. Pauly is Assistant Professor of Political Science and the Dean’s Assistant Professor of Nuclear
Security and Policy at the Watson Institute for International and Public Affairs at Brown University. Rose
McDermott is the David and Mariana Fisher University Professor of International Relations at Brown
University.

For comments on drafts and useful discussions the authors thank Omar Afzaal, Dan Altman, Mat-
thew Bunn, Williamson Murray, Cullen Nutt, Beenish Pervaiz, Daniel Post, Scott Sagan, Paul
Slovic, the anonymous reviewers, and participants in a Peace Science Society (International) travail-
shop and a symposium at the U.S. Naval War College. The authors dedicate this article to Bob
Jervis, OMS, despite failing health, reºected deeply on their working paper. He is missed.

1. Thomas C. Schelling, The Strategy of Conºict (Cambridge, MA: Presse universitaire de Harvard, 1960);
Thomas C. Schelling, Arms and Inºuence (New Haven, CT: Yale University Press, 1966).

International Security, Vol. 47, Non. 3 (Hiver 2022/23), 9–51, https://doi.org/10.1162/isec_a_00451
© 2023 by the President and Fellows of Harvard College and the Massachusetts Institute of Technology.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 10

We identify three mechanisms that are the sources of risk in nuclear
brinkmanship—accidents, self-control, and control of others—by considering
research on the role of psychology and emotion in crisis decision-making. Chaque
of these mechanisms introduces an opportunity for decision-makers to make
choices that can turn chance into leverage over an adversary. Each mechanism
also allows for the possibility of bias to factor into how these choices are
made because of normal human decision-making tendencies. Accidents risk a
decision-maker overreacting to chance events in a way that unintentionally
raises the prospect of nuclear war. Self-control differs across individuals, et
although leaders may retain agency to act in a crisis, they may have greater or
lesser ability to control their own emotional responses to humiliation or provo-
cation. En outre, leaders may not correctly perceive their environment or
the nature of their adversary and may not be aware of their own biases.
Enfin, sometimes a leader cannot control the actions of others in crises but
can nonetheless choose how to respond. Ce faisant, they may fall prey to in-
ferential biases in interpreting the motivations or intentions of their ad-
versaries. Each of these mechanisms of brinkmanship can be inºuenced by
universal aspects of human decision-making architecture, including emotional
réponses. The ability to recognize and control these responses differs pro-
foundly across individuals.

Nuclear threats can be credible in the face of assured retaliation precisely
because of the uncertainty and unpredictability of normal human emotions.
Different people react differently to the same situation, just as they often react
similarly to different situations. One side’s leader can never be exactly sure
what the other leader is thinking or feeling. By focusing on how chance be-
comes leverage via individual differences in risk tolerance, emotional self-
control, and resilience, we offer a fuller account of how human psychology
affects nuclear strategy and crisis decision-making.

The theory of nuclear brinkmanship is core to the security studies canon. Il
explains how states can manipulate the risk of disaster to compete under the
shadow of a nuclear war that threatens their mutual survival. A nuclear war
cannot be won. Donc, Thomas Schelling’s “threat that leaves something to
chance” helps to explain both enduring great power competition and the ab-
sence of large-scale war among nuclear powers, providing strong evidence in
support of the theory of the nuclear revolution.2 Nuclear weapons did not

2. On the theory of the nuclear revolution, see Robert Jervis, The Meaning of the Nuclear Revolution:
Statecraft and the Prospect of Armageddon (Ithaca, New York: Cornell University Press, 1989); Bernard

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 11

eliminate war or the security dilemma, and nuclear powers were not typically
satisªed to build small arsenals to achieve minimum deterrence. Plutôt, what
nuclear weapons changed about international politics was how nuclear powers
competed. Nuclear-armed states still disagree over the status quo and chal-
lenge one another’s interests. But they typically do so without making credible
threats of nuclear annihilation. They use brinkmanship instead. And this the-
ory of nuclear brinkmanship helps to explain big and important cases of
nuclear crises, such as those over Cuba, Berlin, or Taiwan.

Brinkmanship is also, as a theory, poorly understood. The puzzle that brink-
manship is meant to solve is that nuclear powers cannot credibly threaten to
use their nuclear weapons against nuclear-armed rivals. Carrying out such a
threat would result in irrational, mutual annihilation. Donc, plutôt que
power or interests deciding the outcome of coercive bargains, it is the balance
of resolve that shapes the outcome via belligerents’ manipulation of the risk of
disaster. Those who can stomach more risk are understood to be more resolved
over the stakes. Those who are more resolved thus prove willing to take riskier
actions because they possess higher risk tolerance to get what they want. Comme
the overall threat of nuclear conºagration rises, less resolved belligerents will
call it quits when the chance of nuclear war is too great. According to the the-
ory, par exemple, President John F. Kennedy’s 1962 naval quarantine of Cuba
pressured the Soviets to remove missiles from the island by raising the possi-
bility that shots would be ªred.3 The near-simultaneous accidental incursion of
a confused U.S. U-2 pilot into Soviet air space, the actual shooting down of a
U-2 over Cuba without the permission of Moscow, and multiple false missile
warnings all made it seem as though the situation were slipping out of con-
trol.4 Soviet Premier Nikita Khrushchev proved less willing to tolerate the risk
of escalation than Kennedy in this circumstance.

Schelling called these signals of nuclear risk “threat[s] that leave something
to chance” and offered that they were a solution to the problem of agency in

Brodie, éd., The Absolute Weapon: Atomic Power and World Order (New York: Harcourt Brace, 1946).
For critiques, see Brendan Rittenhouse Green, The Revolution That Failed: Nuclear Competition, Arms
Contrôle, and the Cold War (Cambridge: la presse de l'Universite de Cambridge, 2020); Austin Long and
Brendan Rittenhouse Green, “Stalking the Secure Second Strike: Intelligence, Counterforce, et
Nuclear Strategy,” Journal of Strategic Studies 38, nos. 1–2 (2015): 38–73, https://doi.org/10.1080/
01402390.2014.958150; Keir A. Lieber and Daryl G. Presse, The Myth of the Nuclear Revolution: Power
Politics in the Atomic Age (Ithaca, New York: Cornell University Press, 2020).
3. See Michael Dobbs, One Minute to Midnight: Kennedy, Khrushchev, and Castro on the Brink of Nu-
clear War (New York: Alfred A. Knopf, 2008).
4. Scott D. Sagan, The Limits of Safety: Organizations, Accidents, and Nuclear Weapons (Princeton, New Jersey:
Princeton University Press, 1993).

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 12

coercion. If a leader cannot credibly threaten to start a nuclear war, peut-être
they can at least signal that the decision is out of their hands. The leader can
communicate that certain elements of the crisis incorporate chance beyond the
decision-maker’s control that could go the wrong way. Researchers subse-
quently studied the causes of nuclear risk, such as bureaucratic and organiza-
tional pathologies,5 accidents,6 or inadvertent escalation.7 We acknowledge
that these theories are crucially important and have contributed greatly to un-
derstanding nuclear risk. Encore, they have often minimized the role of the indi-
vidual decision-maker, a lens that is key to analyzing how a small group or an
individual can decide to authorize using nuclear weapons.

Ainsi, our central puzzle springs from the recognition that threats that leave
something to chance do not actually eliminate human decision-making from
crises. Plutôt, human emotions prove central to understanding the psycholog-
ical dynamics that underlie such events. Leaders do not control risky incidents
and near-misses, but even so they can choose whether and how to respond to
these chance events. Said another way, leaders retain the choice to further esca-
late a crisis even when they lose control over discrete escalatory incidents, comme
the Cuban missile crisis example demonstrates. Barring a preexisting dooms-
day machine,8 leaders still must make a conscious choice to use nuclear weap-
ons, even in response to an attack that is assumed to be so provocative as to
demand one.9 Jumping into the abyss is considered irrational, therefore the
canon assumes that chance cannot translate into leverage while agency re-
mains. Yet choices can remain even when chance events occur. Ainsi, nuclear
strategy scholarship does not fully answer the question: how does chance be-
come leverage in nuclear crises?

This analysis tackles the challenge posed by this theoretical gap between
chance and choice in threats that leave something to chance by bringing per-
ceptual and emotional variables into our understanding of nuclear brinkman-
ship. We argue that these psychological variables are central to achieving a

5. Graham T. Allison, “Conceptual Models and the Cuban Missile Crisis,” American Political Sci-
ence Review 63, Non. 3 (1969): 689–718, https://doi.org/10.2307/1954423.
6. Sagan, The Limits of Safety.
7. Barry R. Posen, Inadvertent Escalation: Conventional War and Nuclear Risks (Ithaca, New York: Cornell
Presse universitaire, 1991).
8. The Russians came closest to building such an automatic retaliation system. See David E.
Hoffman, The Dead Hand: The Untold Story of the Cold War Arms Race and Its Dangerous Legacy (Nouveau
York: Doubleday, 2009).
9. Even Barry Posen, in arguing that leaders are highly likely to respond to conventional attacks
that target their nuclear forces, acknowledges that some leaders might ignore the effect of these ac-
tions on their weapons. Posen, Inadvertent Escalation.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 13

more accurate understanding of threats that leave something to chance. A dis-
tinction between chance and choice is embedded in the idea of threats that
leave something to chance. Donc, one does not have to eliminate choice to
generate chance.

Psychological and emotional variables are sources of risk, uncertainty, et
ambiguity when leaders retain the choice to use nuclear weapons. In conºict
or crisis, a coercer may or may not do what they threaten, and a target may not
know whether to believe the threat. Either the coercer or the target could
also become swept away by emotions such as anger or fear and take an ac-
tion that they had not planned at the outset, especially under stress or in the
face of loss. These factors can make an already uncertain situation even more
unstable and unpredictable. Belligerents in nuclear crises can know this
and thus fear that the other will make an irrational choice. The greater the
fear, the more leverage the chance generates.10 In this way, our argument seeks
to provide a greater conceptual explication of brinkmanship by incorporating
the critical psychological elements of risk perception and emotion regulation
into the theory.11

The rest of this article proceeds as follows. Section two discusses the litera-
ture on brinkmanship in greater detail to provide a foundation for our work.
Section three divides a novel typology of threats that leave something to
chance into three distinct mechanisms of brinkmanship: accidents, self-control,
and control of others. These sources of chance together explain how it is possi-
ble to generate leverage in crises without eliminating leaders’ choices. Sections
four, ªve, and six unpack each mechanism’s emotional elements and pro-
cesses. We also distinguish these mechanisms from the alternative “madman
theory.” The ªnal section provides some conclusions about how our approach
improves the understanding of past, présent, and future nuclear crises and
opens new avenues for research.

Situating Our Work in a Literature That Discounts Humans

Brinkmanship purports to solve a problem of threat credibility. For a target to
acquiesce to a coercive demand, it must believe that it will actually be pun-

10. We thank Jacqueline Hazelton for helpful guidance on this point.
11. On “emotion regulation,” see James J. Gross, “Emotion Regulation: Current Status and Fu-
ture Prospects,” Psychological Inquiry 26, Non. 1 (2015): 1–26, https://doi.org/10.1080/1047840X
.2014.940781.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 14

ished for deªance and rewarded for submission. Coercers therefore take great
pains to generate leverage over their targets by demonstrating both the capa-
bility to hurt and the resolve to punish if their demands are not met.12 These
ideas are captured in several generations of theories about how states can sig-
nal their resolve and make their threats credible—costly signaling, such as ty-
ing hands and sunk costs,13 audience costs,14 reputation,15 and strategies of
commitment.16 To this list, and speciªc to nuclear coercion, Schelling added
brinkmanship as a strategy of communicating resolve via “the threat that
leaves something to chance.”17

Schelling’s analogy, which we return to throughout the article, was to two
mountaineers chained together at the edge of a cliff. One climber cannot credi-
bly threaten to push the other off the cliff, because that would doom them
les deux; just as if one state threatens nuclear war, unacceptable retaliation might
ensue. But each climber could still generate leverage over the other by taking
risks. One climber could take a step closer to the edge, stand on one foot, même
dance around on the smooth slope or loose gravel below their feet. In this way,

12. Alexander L. George, Forceful Persuasion: Coercive Diplomacy as an Alternative to War
(Washington, CC: United States Institute of Peace, 1991); Alexander L. George and William E.
Simons, éd., The Limits of Coercive Diplomacy: Laos, Cuba, Vietnam, 2nd ed. (Boulder, CO: Westview
Presse, 1994); Robert J. Art and Patrick M. Cronin, éd., The United States and Coercive Diplomacy
(Washington, CC: United States Institute of Peace, 2003); Barry M. Blechman and Stephen S.
Kaplan, Force without War: U.S. Armed Forces as a Political Instrument (Washington, CC: Brookings
Institution, 1978).
13. James D. Fearon, “Domestic Political Audiences and the Escalation of International Disputes,»
American Political Science Review 88, Non. 3 (Septembre 1994): 577–592, https://www.jstor.org/
stable/2944796; James D. Fearon, “Signaling Foreign Policy Interests: Tying Hands versus Sinking
Costs,” Journal of Conºict Resolution 41, Non. 1 (Février 1997): 68–90, https://www.jstor.org/stable/
174487; Branislav L. Slantchev, “Military Coercion in Interstate Crises,” American Political Science
Review 99, Non. 4 (Novembre 2005): 533–547, https://www.jstor.org/stable/30038963. On signaling
in general see Robert Jervis, The Logic of Images in International Relations (New York: Columbia Uni-
versity Press, 1970).
14. Fearon, “Domestic Political Audiences”; Fearon, “Signaling Foreign Policy Interests”; Kenneth
UN. Schultz, Democracy and Coercive Diplomacy (Cambridge: la presse de l'Universite de Cambridge, 2001);
Kristopher W. Ramsay, “Politics at the Water’s Edge: Crisis Bargaining and Electoral Competi-
tion,” Journal of Conºict Resolution 48, Non. 4 (Août 2004): 459–486, https://www.jstor.org/stable/
4149804; Jessica L. Weeks, “Autocratic Audience Costs: Regime Type and Signaling Resolve,” In-
ternational Organization 62, Non. 1 (Hiver 2008): 35–64, https://www.jstor.org/stable/40071874.
15. Daryl G. Presse, Calculating Credibility: How Leaders Assess Military Threats (Ithaca, New York: Cornell
Presse universitaire, 2005); Jonathan Mercer, Reputation and International Politics (Ithaca, New York: Cor-
nell University Press, 1996); Anne E. Sartori, Deterrence by Diplomacy (Princeton, New Jersey: Princeton Uni-
versity Press, 2005); Ryan Brutger and Joshua D. Kertzer, “A Dispositional Theory of Reputation
Costs,” International Organization 72, Non. 3 (2018): 693–724, https://doi.org/10.1017/S0020818318
000188.
16. Schelling, The Strategy of Conºict; Schelling, Arms and Inºuence.
17. Ibid..

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 15

the climbers are engaged in brinkmanship—a competition in risk-taking or
manipulating the chance of mutual disaster. The ground could always give
way underneath one of them, unintentionally plunging both to their deaths.
As both climbers move further down the path toward oblivion by engaging
in increasingly risky behavior, a rationalist theory predicts that whichever
actor cares more about the stakes of a nuclear crisis will have a higher risk tol-
erance and thus be willing to accept more risk of nuclear war.18 The less re-
solved actor will concede if the more resolved actor manipulates sufªcient
risk. Ainsi, even if one side or the other cannot credibly threaten to use nuclear
weapons in a crisis, each can still try to generate bargaining leverage by taking
steps that raise the risk that a crisis might unintentionally escalate to strategic
nuclear war.

Schelling’s theory of threats that leave something to chance, introduced in
The Strategy of Conºict and developed in Arms and Inºuence, is simple and pow-
erful. It purports to explain both how states could stumble into war without
intention and why actors might strategically seek to manipulate risk to achieve
advantage. It shows how there can be strength in vulnerability, since those
who take the greatest risk stand to gain the biggest reward. Et, like most of
Schelling’s work, the idea was policy relevant, prescribing and explaining ca-
nonical Cold War crisis behavior, such as the Cuban missile crisis.19 The theory
is also a crucial complement to the theory of the nuclear revolution, as it ex-
plains competition and coercion between nuclear powers despite the stalemate
that their arsenals are supposed to induce. Nuclear weapons still changed in-
ternational politics, they just did so by changing the way that states compete,
not the fact that they do.

Yet the threat that leaves something to chance, as Schelling presented it, a
a key ºaw: it discounts humans. It does not reckon with the role of decision-
makers in crisis and the psychological biases inherent in human behavior. Ad-
vances in the study of human psychology since Schelling’s writing have made
this missing perspective even more noticeable. En effet, debates about nuclear
strategy and coercion have mostly baked in this implicit original assumption
that for brinkmanship to work, humans must lose agency over the choice to
engage in nuclear war. Escalation in the aftermath of an accident or a false

18. Schelling, The Strategy of Conºict. On rationalist bargaining theory, see also Fearon, “Signaling
Foreign Policy Interests.”
19. Jeffry A. Frieden and David A. Lake, “International Relations as a Social Science: Rigor and
Relevance,” ANNALS of the American Academy of Political and Social Sciences 600, Non. 1 (2005): 136–
156, https://doi.org/10.1177/0002716205276732.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 16

warning is unnecessarily presumed to have a quality of automaticity. Encore, le
reality is exactly the opposite: for brinkmanship to work, humans must retain
agency to choose in the face of chance events.

The simplifying assumption comes from the seeming incompatibility of ra-
tional decision-making with nuclear conºict. If strategic nuclear war is irratio-
nal but decision-makers are assumed to be rational, then scholars must
discount the decision-making actor and assign to events a momentum of their
propre. The problem with this assumption, bien sûr, is that not all decision-
makers are rational, and it cannot be assumed that they are, or that they cannot
make disastrous mistakes even if they are largely rational in cost-beneªt terms.
Said in terms of Schelling’s analogy, the climber can still fall despite trying not
à, dragging the other climber to their mutual doom. An accurate understand-
ing of human psychology under conditions of threat requires proper con-
sideration of the role of emotion and the desire for revenge in the motivation
for attack.20 In the rest of this section we unpack how theories of nuclear
brinkmanship fail to account for imperfect human decision-making.

Robert Powell’s theoretical distinction between the risk model and the pun-
ishment model of nuclear strategy is one of the most important contribu-
tions to the literature on brinkmanship after Schelling.21 Powell argued that
Schelling’s brinkmanship model—the risk model—applied when escalation
was uncontrolled; and the punishment model applied when leaders remained
in control of escalation. Ainsi, for Powell, mutually invulnerable strategic
forces replaced the risk model with the punishment model, because limited
options would always remain limited if escalation were rational. “The threat
that leaves something to chance also relies on a particular irrational act,” he
wrote. “The implausibility of this particular act implies that there is nothing to
be left to chance.”22 For Powell, there is no bargaining advantage in the pre-
emptive use of nuclear force. The only threat that leaves something to chance
derives from accidents or irrationality. But Powell underappreciated how
much choice and chance can coexist, particularly regarding the uncertainty in-
troduced by human emotional responses to threat, perte, and risk. Decision-
makers must still choose to actually respond or escalate to nuclear use, même

20. Rose McDermott, Anthony C. Lopez, and Peter K. Hatemi, “‘Blunt Not the Heart, Enrage It’:
The Psychology of Revenge and Deterrence,” Texas National Security Review 1, Non. 1 (Novembre
2017): 69–89, http://hdl.handle.net/2152/63934.
21. Robert Powell, “The Theoretical Foundations of Strategic Nuclear Deterrence,” Political Science
Trimestriel 100, Non. 1 (Spring 1985): 75–96, https://www.jstor.org/stable/2150861.
22. Ibid., 85.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 17

when they face very strong incentives to do so. Leaders can still retain choice
in how to respond, including the choice to manipulate risk, even when con-
fronting external events that they cannot control.

Brinkmanship is also at the center of debates about the coercive utility of nu-
clear weapons.23 Nuclear coercionists are careful to articulate two different
schools of thought around the effects of nuclear weapons on crisis outcomes.24
The ªrst is passive: that the mere existence of nuclear weapons and the ability
to launch them pervades crises with a nuclear shadow. This nonzero probabil-
ity of nuclear war means that leaders must make choices in crisis attuned to
the risk of nuclear escalation. The second school sees the role of nuclear weap-
ons as more active, as Schelling did. Nuclear weapons affect the outcome of
the crisis when leaders choose to engage in brinkmanship tactics to raise the
risk of nuclear war (and the possession of nuclear weapons should make it
more likely that a leader would be emboldened to use these tactics). Neither
school provides a satisfying theory of nuclear leverage in crisis. Each assumes
that the risk of nuclear war is generated through the existence or operation of
military forces alone. War just sometimes gets out of hand. The human ele-
ment of choice is forgotten.

Matthew Kroenig, par exemple, attempts to resolve what we ªnd puzzling
about brinkmanship—the seeming paradox of choice and chance—but he does
so by introducing counterforce and damage limitation strategies to contest
Schelling’s original analogy.25 Actors with superior nuclear arsenals, he ar-
gues, are willing to risk more at the brink. In terms of the mountaintop exam-
ple, through counterforce and damage limitation, one climber aims to cut the
rope that tethers the two climbers together before pushing the other climber

23. Todd S. Sechser and Matthew Fuhrmann, Nuclear Weapons and Coercive Diplomacy (Cambridge:
la presse de l'Universite de Cambridge, 2017); Kyungwon Suh, “Nuclear Balance and the Initiation of
Nuclear Crises: Does Superiority Matter?,” Journal of Peace Research, online preprint, May 2022,
https://doi.org/10.1177/00223433211067899; Kyungwon Suh, “Bargaining with the Bomb: Milita-
rized Nuclear Signals and Crisis Bargaining” (working paper presented at the International Secu-
rity Research Colloquium, Centre for International Security, Hertie School, Berlin, Avril 14, 2022).
See also Robert A. Pape, Bombing to Win: Air Power and Coercion in War (Ithaca, New York: Cornell Univer-
sity Press, 1996), 20–21; Mark S. Bell and Julia Macdonald, “How to Think about Nuclear Crises,»
Texas National Security Review 2, Non. 2 (Février 2019): 41–64, http://dx.doi.org/10.26153/tsw/
1944; David C. Logan, “The Nuclear Balance Is What States Make of It,” International Security 46,
Non. 4 (Spring 2022): 172–215, https://doi.org/10.1162/isec_a_00434.
24. Marc Trachtenberg, “The Inºuence of Nuclear Weapons in the Cuban Missile Crisis,” Interna-
tional Security 10, Non. 1 (Été 1985): 137–163, https://doi.org/10.2307/2538793. Todd Sechser
and Matthew Fuhrmann make the same distinction in Nuclear Weapons and Coercive Diplomacy.
25. Matthew Kroenig, The Logic of American Nuclear Strategy: Why Strategic Superiority Matters
(New York: Presse universitaire d'Oxford, 2018).

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 18

off the cliff. Doing so successfully requires no brinkmanship. But what if the
knife is not sharp enough to sever the rope, or what if the other climber pre-
empts when they see the knife drawn? Both climbers can still fall to their fates.
We assume, like Schelling, that neither actor possesses a disarming ªrst-strike
counterforce capability, yet brinkmanship can still occur.

Other nuclear posture choices can also affect the magnitude of risk in brink-
manship, such as command-and-control structures, authorization and delega-
tion to use nuclear weapons, tactical nuclear options, as well as sheer capacity
for retaliation. To extend the analogy, one climber may have better balance,
better spikes or traction on their boots, or they may possess superior knowl-
edge regarding the structure and stability of the cliff itself, knowing where to
stand and what rocks are unstable. Yet the foremost studies either make pos-
ture a dependent variable or examine the effects of nuclear posture on only de-
terrence and not compellence.26 In his important work, Vipin Narang does
not theorize risk acceptance as an explanatory variable for posture choice
(although we infer some effects later in this article). Plutôt, Narang’s coding
of “civil-military arrangements” as “assertive or delegative” pertains to civil-
military relations and the balance between addressing internal and external
threats to the state. Negative and positive controls over nuclear forces are a
consequence of this choice.27 Moreover, in other work Matthew Fuhrmann and
Todd Sechser see forward deployed nuclear weapons as ex ante sunk cost sig-
nals that manipulate the risk of accident or unauthorized use. Par exemple, le
United States may deploy tactical nuclear weapons on the territories of North
Atlantic Treaty Organization (NATO) allies to make credible claims of ex-
tended deterrence.28 Thus, both approaches have similarly removed human
agency in crisis from their theorizing.

The literature on escalation does not do much better at accounting for the
role of human psychology in decision-making. We see the study of escalation
as being divided into two schools.29 One school sees conºict escalation as a se-

26. Vipin Narang, Nuclear Strategy in the Modern Era: Regional Powers and International Conºict
(Princeton, New Jersey: Princeton University Press, 2014).
27. Ibid., 36–39.
28. Fearon, “Domestic Political Audiences”; Matthew Fuhrmann and Todd S. Sechser, “Signaling
Alliance Commitments: Hand-Tying and Sunk Costs in Extended Nuclear Deterrence,” American
Journal of Political Science 58, Non. 4 (Octobre 2014): 919–935, https://www.jstor.org/stable/
24363534. Thomas Schelling thought of threats that leave something to chance as a much more dy-
namic phenomenon that could be rapidly manipulated in a crisis.
29. We thank Daniel R. Post for his clarifying work on escalation. Post, “Escalation and Coercion:
Exploring ‘Escalate to De-escalate’ Strategies in Past Wargames and Crisis Simulations” (unpub-
lished manuscript, 2022).

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 19

ries of deliberate decisions—to signal and bargain in limited war.30 Within this
school, scholars understand escalatory steps as calibrated to breach some sa-
lient threshold: vertical escalation to use new weapons or more violence, hori-
zontal to expand conºict to new geographies, or political if they breach a
norm. In nuclear strategy literature, deliberate escalation is chieºy understood
to be used either for warªghting or to signal resolve.31 A second school sees
escalation as a natural phenomenon involving accidental or inadvertent
pathways—war can get out of hand, “it has ªre-like properties that cause it to
continue once it begins,”32 or “war is like love, it always ªnds a way.”33
Richard Smoke developed it generally, arguing that smaller conºicts grow
larger bit by bit until leaders lose control over their ability to prevent escala-
tion.34 And Bernard Brodie applied the intuition to the nuclear age, offering
that “violence between great opponents is inherently difªcult to control.”35
Speciªc to limited nuclear war, Robert Jervis distinguished between two mech-
anisms of threats that leave something to chance: intended and unintended es-
calation;36 Morton Halperin uses similar categories of explicit decisions versus
unintended “explosions” of violence.37

To paraphrase Scott Sagan, one school sees escalation as a staircase and the
other as an escalator.38 Neither is fully satisfying when applied to nuclear
brinkmanship. The former school assumes that rational humans will make de-
cisions to gain advantage, so there should be no risk of strategic nuclear war
because the rational leader will stop short of annihilation. In terms of the

30. Forrest E. Morgan et coll., Dangerous Thresholds: Managing Escalation in the 21st Century (Santa
Monica, Californie: RAND, 2008); Herman Kahn, On Thermonuclear War (New Brunswick, New Jersey: Transac-
tion Publishers, 1960); Herman Kahn, Thinking about the Unthinkable (New York: Horizon Press,
1962).
31. J.. Michael Legge, Theater Nuclear Weapons and the NATO Strategy of Flexible Response, R-2964-FF
(Santa Monica, Californie: RAND, 1983), 19.
32. Michelle English, “3 Questions: Stephen Van Evera Revisits World War I,” MIT News, Novem-
ber 8, 2018, https://news.mit.edu/2018/3-questions-stephen-van-evera-revisits-world-war-i-100-
years-after-bitter-end-1108.
33. Bertolt Brecht, Mother Courage, quoted in Richard K. Betts, “Must War Find a Way? A Review
Essay,” International Security 24, Non. 2 (Fall 1999), 166, https://doi.org/10.1162/016228899560112.
34. Richard Smoke, War: Controlling Escalation (Cambridge, MA: Presse universitaire de Harvard, 1977).
35. Bernard Brodie, “What Price Conventional Capabilities in Europe?” Reporter, May 23, 1963, 32,
quoted in Robert Jervis, The Illogic of American Nuclear Strategy (Ithaca, New York: Cornell University
Presse, 1984), 137.
36. Jervis, The Illogic of American Nuclear Strategy.
37. Morton H. Halperin, Limited War in the Nuclear Age (New York: Wiley, 1963).
38. Eric Schlosser, “What If Russia Uses Nuclear Weapons in Ukraine?,” Atlantic, Juin 20,
2022, https://www.theatlantic.com/ideas/archive/2022/06/russia-ukraine-nuclear-weapon-us-
response/661315/.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 20

mountaineering analogy, every decision to jump down the cliff is more like a
choice to descend to a ledge some distance below. Leaping into the abyss is ir-
rational, but the climbers might survive the drop to each shelf on the way
down. The latter school tends to bury the element of human choice in the after-
math of accidents and inadvertence, denying human agency a role in deter-
mining the outcome. In some sense, the climbers are assumed to be clumsy,
and given no opportunity to catch their footing or react after they stumble.
Decision-making decisively inºuenced by powerful emotions suggests a far
less linear path to escalation. Rather than a series of steps, or ledges, emotional
dynamics are more likely to produce tipping points, whereby little changes
until some event precipitates a much larger shift—equivalent to a suicidal
jumping off the ledge entirely.39

Jervis and others articulated how worst-case assumptions and mispercep-
tions of adversary interests can lead actors to improperly calibrate military sig-
nals.40 But this approach only begins to explain the mechanisms underlying
brinkmanship. A leader might escalate because of a misperception of enemy
intention, while not believing themselves be an irrational or unpredictable
decision-maker. Bargaining leverage can also originate from such perceptions:
each side makes an assessment, however inaccurate, about the other’s willing-
ness to use nuclear weapons. Each side may simultaneously be convinced that
they are not irrational, but the other side might be prone to unwarranted esca-
lation. Brinkmanship can thus come down to each side’s ability to manipulate
the other side’s perception of its intentions. Generating sufªcient lever-
age to get the other side to back down can derive from just such an abil-
ity to manipulate perception on the other side. The black box of how
this “choice” to escalate is made deserves further investigation. What remains
missing is a theory that accounts for the role of individual decision-making
in escalating conºicts and crises. All types of escalation—deliberate, inadver-
tent, or accidental—can result from emotional or psychological causes as well
as effects.

39. P.. J.. Lamberson and Scott Page draw a distinction between direct tips and contextual tips:
“A direct tip occurs when a gradual change in the value of a variable leads to a large, c'est à dire. discontin-
uous, jump in that same variable in the future. A contextual tip occurs when a gradual change in
the value of one variable leads to a discontinuous jump in some other variable of interest.” P. J..
Lamberson and Scott E. Page, “Tipping Points,” Quarterly Journal of Political Science 7, Non. 2 (2012):
175–208, quote at 175, http://dx.doi.org/10.1561/100.00011061. From this perspective, these kinds
of contextual tipping points can precipitate escalation in a more sudden way than suggested by
linear models of change.
40. Robert Jervis, Perception and Misperception in International Politics, rev. éd. (Princeton, New Jersey:
Princeton University Press, 2017).

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 21

De plus, the valuable literature on organizations, nuclear accidents, et
accidental war does not entirely solve the puzzle of how threats that leave
something to chance might work in a crisis.41 The prospect of accidents intro-
duces risk into a crisis but is not in itself a necessary or sufªcient mechanism to
explain the onset of thermonuclear war. Powell, par exemple, models only “au-
tonomous” risk—just the fact of being in a crisis creates a risk that increases
throughout the crisis.42 But this operationalization is not quite right because
leaders may actually choose to do things to create those risks. And even in the
aftermath of accidents or inadvertence, decisions need to be made about how
to respond to speciªc events.

Ici, we seek to bring advances in the study of human psychology to bear
on nuclear strategy. Pathbreaking work on the role of psychology and deter-
rence examines the role of various psychological and emotional factors in en-
hancing or undercutting deterrence.43 But this work has rarely made the leap
to study the psychological underpinning of successful or failed compellence.44
Schelling was the ªrst to clearly explicate the difference between deterrence,
trying to prevent someone from doing something you do not like, and compel-
lence, trying to get someone to start doing something you want. Compellence
is widely recognized as being more difªcult and has historically received less
attention in the literature.45 Brinkmanship can be deterrent or compellent but

41. See Bruce G. Blair, The Logic of Accidental Nuclear War (Washington, CC: Brookings Institu-
tion, 1993); Paul J. Bracken, The Command and Control of Nuclear Forces (New Haven, CT:
Yale University Press, 1983); Peter D. Feaver, Guarding the Guardians: Civilian Control of Nu-
clear Weapons in the United States (Ithaca, New York: Cornell University Press, 1992); Sagan, The Limits
of Safety; Bradley A. Thayer, “The Risk of Nuclear Inadvertence: A Review Essay,” Security
Études 3, Non. 3 (Spring 1994): 428–493, https://doi.org/10.1080/09636419409347557; and other es-
says in this issue.
42. Robert Powell, Nuclear Deterrence Theory: The Search for Credibility (Cambridge: Cambridge
Presse universitaire, 1990).
43. Jervis, The Illogic of American Nuclear Strategy; Jervis, The Meaning of the Nuclear Revolution; Rob-
ert Jervis, Richard Ned Lebow, and Janice Gross Stein, Psychology and Deterrence (Baltimore: Johns
Hopkins University Press, 1989).
44. We build on the terriªc work of Roseanne McManus, who has investigated the coercive effects
of different forms of perceived “madness,” not speciªc to nuclear coercion. See Roseanne W.
McManus, “Revisiting the Madman Theory: Evaluating the Impact of Different Forms of Per-
ceived Madness in Coercive Bargaining,” Security Studies 28, Non. 5 (2019): 976–1009, https://
doi.org/10.1080/09636412.2019.1662482. On psychology and compellence, see also Kathleen E.
Powers and Dan Altman, “The Psychology of Coercion Failure: How Reactance Explains Resis-
tance to Threats,” American Journal of Political Science, online preprint, Juin 2022, https://est ce que je.org/
10.1111/ajps.12711.
45. For exceptions, see Todd S. Sechser and Matthew Fuhrmann, “Crisis Bargaining and Nuclear
Blackmail,” International Organization 67, Non. 1 (2013): 173–95, http://www.jstor.org/stable/
43282156; Matthew Kroenig, “Nuclear Superiority and the Balance of Resolve: Explaining Nuclear
Crisis Outcomes,” International Organization 67, Non. 1 (2013): 141–71, http://www.jstor.org/stable/
43282155; James W. Davis Jr., Threats and Promises: The Pursuit of International Inºuence (Baltimore:

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 22

is intimately connected to the concept of compellence, since two actors en-
gaged in brinkmanship disagree over the status quo, and at least one of
them seeks to change it. Néanmoins, we use Schelling’s deªnition of coercion
to include both deterrence and compellence. For all his innovative bril-
liance, Schelling underappreciated the role of psychology in coercion, which is
not surprising given his economic model of human decision-making.46 Be-
cause Schelling’s inºuence was both foundational and pervasive, and rightly
donc, the bulk of subsequent scholarship tended to overlook the role of psychol-
ogy in decision-making.47 Sometimes the role of individual leaders is dis-
missed entirely.

Ainsi, these questions persist: How does human decision-making affect
brinkmanship? How might leaders retain choice yet create chance in nu-
clear crises?

Bringing Humans Back In

The literature on nuclear brinkmanship has taken for granted that nuclear
weapons innately engender a risk of nuclear war and that crises peak that risk.
But if it is irrational to jump off the cliff, then why would brinkmanship work?
As long as a seemingly rational decision-maker retains a choice over whether
to escalate to strategic nuclear war, their threats to do so should not be credi-
ble. What is the real source of risk when agency remains? We locate it in the
architecture of human psychological decision-making.

Because we are proposing a theory, we must be clear about our scope condi-

Johns Hopkins University Press, 2000); Todd S. Sechser, “Militarized Compellent Threats, 1918–
2001,” Conºict Management and Peace Science 28, Non. 4 (2011): 377–401, https://doi.org/10.1177/
0738894211413066; Reid B. C. Pauly, “Deniability in the Nuclear Nonproliferation Regime: The Up-
side of the Dual-Use Dilemma,” International Studies Quarterly 66, Non. 1 (Mars 2022): sqab036,
https://doi.org/10.1093/isq/sqab036; Cullen G. Nutt and Reid B. C. Pauly, “Caught Red-Handed:
How States Wield Proof to Coerce Wrongdoers,” International Security 46, Non. 2 (Fall 2021): 7–50,
https://doi.org/10.1162/isec_a_00421. Much of this work became subsumed in the study of coer-
cive diplomacy; see footnote 12.
46. Benjamin Wilson, “Keynes Goes Nuclear: Thomas Schelling and the Macroeconomic Origins
of Strategic Stability,” Modern Intellectual History 18, Non. 1 (2019): 171–201, https://est ce que je.org/
10.1017/S1479244319000271. Benjamin Wilson argues that Schelling’s conªdence in the stability of
deterrence, even after shocks to the system, came not from the game theory to which it is most
often ascribed but from his training in Keynesian economic modeling. Wilson focuses on an early
papier, Thomas C. Schelling, Randomization of Threats and Promises, RAND P-1716 (Santa Monica,
Californie: RAND, 1959), which became chapter 7 of Schelling’s The Strategy of Conºict.
47. Robert Jervis’s work stands as an important exception, particularly Jervis, Perception et
Misperception in International Politics.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 23

tion. D'abord, our argument centers around the sources of risk in nuclear crises. Il
may not apply outside the high-stakes context of nuclear crises, but it should
be widely applicable to crises between states with secure second-strike nuclear
forces.48 We proceed from Mark Bell and Julia Macdonald’s deªnition of a nu-
clear crisis (based on the International Crisis Behavior project) as an interaction
between two nuclear-armed states in which there is a “heightened probability
of military hostilities” that “destabilizes their relationship” and begins with a
“disruptive act or event.”49 This initial disruptive act should be a discrete
escalatory event, but it need not involve a nuclear-speciªc risk. C'est, nuclear
accidents or close calls are not the only contexts in which these mechanisms
operate. They are also at play when leaders alert nuclear forces, make or imply
nuclear threats, operate their conventional forces to signal their resolve to esca-
late further, or otherwise engage in brinkmanship.

Deuxième, our unpacking of the mechanisms of brinkmanship is decision-
theoretical and not game-theoretical. We do not intend to model a strategic in-
teraction; cependant, each side may try to anticipate what the other will do. Notre
logic applies to both sides: the coercer and the target. Each side may misjudge
the other’s intentions, and each side may know that their judgments of the
other’s intentions may be wrong, introducing inherent uncertainty. Regardless,
even though nuclear brinkmanship is dyadic and dynamic, each mechanism
applies to a single decision-maker. In brinkmanship, there is no single coercer
and target since in theory both sides may simultaneously be engaged in bar-
gaining and generating leverage for deterrence or compellence.

Troisième, we focus on bringing individual psychology to bear on the study of
nuclear strategy. En tant que tel, we do not do justice to other promising levels
of analysis, such as bureaucracy as a source of risk. This is not to say that bu-
reaucracy has nothing to do with brinkmanship. En effet, it likely does; but our
primary aim is to explain the underappreciated risk-generating role of the in-
dividual decision-maker who has authority over the use of nuclear weapons
in a crisis.

In table 1, we summarize three distinct mechanisms by which chance can be
converted into leverage in nuclear crises: (1) accidents; (2) a lack of self-control;

48. No speciªc nuclear postures are scoped in or out of our theory, save for the assumption that
neither actor has a perfect counterforce capability against the other (c'est à dire., that the rope tying the two
climbers together is not severable). Néanmoins, we do discuss the effects of speciªc posture
choices.
49. Bell and Macdonald, “How to Think about Nuclear Crises,» 42. We comment in the conclusion
on the possibility of nuclear brinkmanship involving only one nuclear-armed actor.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 24

Tableau 1. Disentangling Chance and Choice in Nuclear Brinkmanship

Mechanisms

Concept mentioned in
Schelling’s Strategy of Conºict

Relationship to choice

accidents

accident

mechanical failure

self-control

panic

madness

false alarm

misapprehension of
enemy intentions

control of others

pre-delegation

mischief

limited war as a generator
of risk

The decider has agency in the
aftermath of the risk-generating
event.

The decider has agency but cannot
control themselves.

The decider has agency but acts on
misperception.

The decider does not have agency,
a third party or unauthorized
decider does.

The decider retains a choice, mais le
adversary also has agency over the
decision for nuclear war.

et (3) a loss of control over others. The categories may bleed together in the
real world, and they are not mutually exclusive, but they are analytically dis-
tinct. Each may apply to any actor on any side of a crisis.50

Our original contribution lies especially in the right-most column of table 1,
and in how the mechanisms interact. Mechanism 1—the chance of an
accident—is generally understood in the literature to be a source of risk in
threats that leave something to chance. But accidents do not remove the need
for either side to decide whether to use nuclear weapons. Plutôt, they gener-
ate risk because psychological and emotional factors—discussed in mecha-
nisms 2 and 3—make the outcomes of those decisions more uncertain than a
cost-beneªt calculus would suggest. We therefore spell out the relationship of
each mechanism to the “choice” to begin a strategic nuclear war—in other
words, making the decision to jump off the cliff—and unpack the dynamics of
that choice in the rest of the article.

To understand brinkmanship and the choices that surround threats that

50. For simplicity, each mechanism is conceived of—in this article and in Schelling’s presenta-
tion—as the behavior of just one belligerent in a crisis, even though, in reality, both sides may en-
gage in brinkmanship simultaneously. The ªnal mechanism (on limited war) comes closest to
capturing this interaction effect, which is core to other aspects of Schelling’s work.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 25

leave something to chance, we argue that it is critical to delineate the relation-
ship between choice and chance. Schelling wrote of brinkmanship: “Where
does the uncertain element in the decision come from? It must come from
somewhere outside of the threatener’s control. Whether we call it ‘chance,’ ac-
cident, third-party inºuence, imperfection in the machinery of decision, or just
processes that we do not entirely understand, it is an ingredient in the situa-
tion that neither we nor the party we threaten can entirely control.”51 The
choice over whether and when a nuclear war will break out is understood to
be a hindrance to effective brinkmanship, since it cannot be made credible.

But chance and choice are not as contradictory as they appear, or as
Schelling thought. Uncertainty can indeed emanate from within decision-
makers or from their inability to read their antagonists properly, both of which
can be, to some degree, within their awareness if not their control. A leader
does not have to lose control to threaten to make an irrational choice in crisis.
Leaders may also threaten or undertake actions that they do not consider to be
irrational, although they may be judged so by outsiders. Before the attack on
Pearl Harbor, Par exemple, Japanese leaders knew they were taking a tremen-
dous risk, but they also believed that the attack offered their best chance of
survival.52 Foreshadowing Schelling, they even spoke of leaps to potential
doom. “Sometimes a man has to jump, with his eyes closed, from the veranda
of Kiyomizu Temple,” argued General Hideki Tojo to Prince Fumimaro Konoe
dans 1941.53 Many wars start or fail to end in the ways that instigators anticipated
precisely because central players do not possess a shared sense of value. Mate-
rial cost-beneªt arguments are unlikely to dissuade suicide bombers who seek
their rewards in heaven.

Chance refers to an event or a cause that cannot be controlled. It introduces
some elements of uncertainty and unpredictability.54 Carl von Clausewitz long

51. Schelling, The Strategy of Conºict, 188. See also Daniel Kahneman, Olivier Sibony, and Cass R.
Soleilstein, Noise: A Flaw in Human Judgment (New York: Little, Brown Spark, 2021).
52. Scott D. Sagan, “Origins of the Paciªc War,” Journal of Interdisciplinary History 18, Non. 4 (Spring
1988): 893–922, https://doi.org/10.2307/204828.
53. Chihiro Hosoya, “Retrogression in Japan’s Foreign Policy Decision-Making Process,” in James
W. Morley, éd., Dilemmas of Growth in Prewar Japan (Princeton, New Jersey: Princeton University Press,
1971), 92.
54. Peter Katzenstein and Lucia Seybert argue that Schelling inappropriately conºated risk and
uncertainty. Risk can be assigned probabilities, whereas uncertainty is neither measurable nor
quantiªable. Pierre J. Katzenstein and Lucia A. Seybert, éd., Protean Power: Exploring the Uncertain
and Unexpected in World Politics (Cambridge: la presse de l'Universite de Cambridge, 2018), 11, 42–43. See also
Benoit Pelopidas, “The Book That Leaves Nothing to Chance: How The Strategy of Conºict and Its
Legacy Normalized the Practice of Nuclear Threats” (unpublished manuscript, Octobre 24, 2016);

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 26

ago appreciated that “no other human activity is so continuously or univer-
sally bound up with chance” as war.55 Choice on the other hand refers to the
extent to which a person or leader can direct events or believes themselves
able to do so. At their best, people should be able to control their responses.
When they retain choice, leaders can make decisions about the direction of a
crisis or conºict. The possibility, and indeed the likelihood, that the leader’s
beliefs about their own agency could be incorrect provides the area of overlap
between chance and choice—unpredictability uniªes both notions. Chance in-
troduces the possibility that events may not occur as planned; a leader’s false
sense of control or overconªdence in the ability to direct events allows for the
same outcome.

But the fact that these outcomes overlap does not mean that their causes
are the same. Plutôt, the causal component in chance derives from the proba-
bilistic nature of the universe, whereas the causal element in choice emerges
from the architecture of human decision-making, or the variation in individ-
uals’ skills and temperaments.56 Some leaders are better at controlling certain
types of encounters given their greater interpersonal skills or psychological re-
sources, such as cognitive intelligence. Some leaders are more perceptive and
intuitively accurate about their environment, the nature of their adversary, comme
well as their own relative strengths, weaknesses, and strategic positions. Some
also possess more capacity to perceive how plans may go off the rails. Impor-
tantly, some are much better at recognizing and controlling their own emo-
tional reactions to upsetting or threatening events or people. Arrogance can
cause others to overestimate their own skills and positions, failing to consider
others’ goals, perceptions, and capabilities. These leaders show less sophistica-
tion and ºexibility in their planning, refusing to accept the possibility that
things may not go as planned. Par exemple, General Douglas MacArthur’s
narcissism cost many men their lives. Less than a third of the 90,000 personnel
whom he abandoned on Corregidor Island in the Philippines on March 11,
1942, lived to see his return to Leyte on October 20, 1944.57

Our argument is that this distinction between chance and choice is elemen-

Mark Blyth, “Great Punctuations: Prediction, Randomness, and the Evolution of Comparative Po-
litical Science,” American Political Science Review 100, Non. 4 (2006): 493–498, https://www.jstor.org/
stable/27644375.
55. Carl von Clausewitz, On War, éd. and trans., Michael Howard and Peter Paret (Princeton, New Jersey:
Princeton University Press, 1976), 85.
56. Kahneman, Sibony, and Sunstein, Noise.
57. Uri Bar-Joseph and Rose McDermott, Intelligence Success and Failure: The Human Factor (Nouveau
York: Presse universitaire d'Oxford, 2017).

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 27

tal to whether leaders can generate leverage in nuclear crises. The most impor-
tant moments in brinkmanship when chance becomes leverage happen in the
war room, not on the battleªeld. It is those who retain agency, a choice over
what comes next, who are an unexplained source of risk. Their moment of
decision—to escalate or not—is key to illuminating the operative mechanisms
of threats that leave something to chance. There may be times when using
force is rational in the strictest sense, particularly if a leader believes that their
regime, or their state, confronts an existential peril. There are also many psy-
chological and emotional reasons to believe that choosing to use force might
not be rational in a particular crisis, and yet it is employed nonetheless in ser-
vice of a leader’s own bias, status, or ego. This occurs precisely because chance
and choice and decision-making itself are often not rational processes; in mo-
ments of crisis, they are as much psychological and emotional as strategic
in nature.

Threats that leave something to chance therefore generate leverage not by
removing the decision to employ nuclear weapons from humans on either
side, as Schelling argued, but rather precisely because of the psychological and
emotional factors that make the outcomes of those decisions more uncertain
than a cost-beneªt calculus would suggest. Extreme emotions, such as might
be expected during a crisis or war, can short-circuit more deliberate forms
of decision-making; Daniel Kahneman refers to the fast, intuitive form of
decision-making that characterizes so-called gut instinct as type I processing.58
The habits that humans rely on in these moments can be advantageous in
some circumstances and disastrous in others, especially under the shadow of
nuclear war. Donc, the unpredictability of type I fast, intuitive, emotional
decision-making makes the already uncertain context of brinkmanship even
more terrifying. Decision-makers are humans, and all humans have emotions,
and those emotions can be both idiosyncratic and unpredictable. Donc,
these psychological issues become part of the calculus and introduce inherent
uncertainty that Schelling overlooked. Whether the other side would act on
their emotions or not is unknown to all.

Emotion thus complicates the elements of chance and choice in brinkman-
ship because a coercer can choose to act on an emotion such as fear or anger, ou
pas. And most people are not able to control such strong emotions in moments
of stress. Most people also know that the same is true for others. En effet, même
the legal system has built in diminished responsibility for those acting in the

58. Daniel Kahneman, Pensée, Fast and Slow (New York: Farrar, Straus and Giroux, 2011).

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 28

heat of passion. So each side suspecting that the other may be unable to control
their emotions under pressure further heightens uncertainty surrounding any
signals sent by either side and intensiªes the role of chance in brinkmanship.

The rest of this article unpacks each of the mechanisms in sequence, ex-
plaining the psychological underpinning of the risks associated with each
type of chance and choice. We use illustrative examples from nuclear crises
et au-delà.

Mechanism 1: Accidents

The ªrst mechanism—accidents—is the category most familiar to the brink-
manship literature. Sometimes, escalation occurs despite the best intentions of
the belligerents—an accidental launch, a mechanical failure that causes two
planes to collide in the air, or a misaimed warning shot. The shooting down of
civilian airliners during moments of heightened tensions provide examples
that are all too common: Korean Air Lines Flight 007 shot down over the Soviet
Union in 1983; Iran Air Flight 655 shot down by a U.S. guided missile cruiser
in the Persian Gulf in 1988; Malaysia Airlines Flight 17 shot down over
Ukraine in 2014; Ukraine International Airlines Flight 752 shot down over Iran
dans 2020; and many more. The fog of war ensures imperfect target selection. Le
Bill Clinton administration’s 1999 accidental bombing of the Chinese Embassy
in Belgrade or myriad drone strikes in Afghanistan that killed innocent people
provide other examples. The longer a crisis goes on, the more likely an unex-
pected event could happen. Schelling conceived of this possibility as a kind of
iterated Russian roulette. However unlikely they may seem, as Sagan writes,
“things that have never happened before happen all the time.”59 Thus, all cri-
ses have some baseline level of risk.

De plus, leaders can manipulate the magnitude of these risks in crises.
D'abord, leaders may choose to augment the risk of accidents to signal their re-
solve by mobilizing more military forces, placing them on ever higher alert,
forward deploying them, or ordering them to operate near the enemy. A leader
who approves “buzzing” tactics by pilots, par exemple, is delegating to agents
the risk of accidents occurring in the ªeld. A ªghter jet buzzing another aircraft
was for Schelling “the purest real-life example I can think of in international
affairs” of a threat that leaves something to chance.60

59. Sagan, The Limits of Safety.
60. Schelling, Arms and Inºuence, 104.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 29

The second way that leaders can manipulate the magnitude of these risks is
through decisions about how much risk to bake into the nuclear postures
around which states have organized their nuclear forces. Nuclear command-
and-control arrangements can be “delegative” to military operators or can as-
sert centralized control over launch procedures.61 Using Narang’s typology of
nuclear postures, an asymmetric escalation posture with delegative control
should be more suitable for brinkmanship than an assured retaliation posture
with assertive control, and catalytic postures work precisely by signaling the
risk of nuclear escalation to a third-party intervener.62 In general, leaders who
delegate more authority to use nuclear weapons, have more mobile and nu-
merous nuclear forces, and do not institute negative controls, such as permis-
sive action links, raise the risk of accidents in a crisis.63 (Mechanism 3 discusses
how devolving or pre-delegating authorities on the use of force down the
chain of command also raises the risks of escalation by others.)

Troisième, bureaucratic and organizational structures also affect the risk of acci-
bosses. Organization theory on “normal accidents” suggests that accidents
and mechanical failures will always happen eventually in complex, tightly
coupled systems like the military organizations that manage nuclear weap-
ons.64 The Cuban missile crisis is also replete with examples of how military
standard operating procedures raised the level of risk beyond that which lead-
ers intended.65 For instance, because of standard nuclear alerting procedures,
the ªghter aircraft that scrambled to intercept a U-2 that strayed into Soviet
airspace were armed with nuclear-tipped air-to-air missiles.66 All these factors
can augment the risk of accidents or close calls. But they do not remove choice

61. Peter D. Feaver and David Arceneaux, “The Fulcrum of Fragility: Command and Control in
Regional Nuclear Powers,” in Scott D. Sagan and Vipin Narang, éd., The Fragile Balance of Terror:
Deterrence in the New Nuclear Age (Ithaca, New York: Cornell University Press, 2023); David Arceneaux,
“Beyond the Rubicon: Command and Control in Regional Nuclear Powers” (PhD diss., Syracuse
University, 2019).
62. Narang, Nuclear Strategy in the Modern Era.
63. Leaders have historically resisted devolving control of nuclear weapons in this fashion, comment-
jamais. The United States, which consistently planned to ªght and win a nuclear war against the
Soviet Union, resisted introducing limited options into its war plans. Long and Green, “Stalking
the Secure Second Strike”; Lieber and Press, The Myth of the Nuclear Revolution; Scott D. Sagan,
“SIOP-62: The Nuclear War Plan Brieªng to President Kennedy,” International Security 12, Non. 1
(Été 1987): 22–51, https://doi.org/10.2307/2538916. Today it is debated how much Pakistan
is prepared to loosen control over its tactical nuclear forces in crisis. Arceneaux, “Beyond the
Rubicon.”
64. Charles Perrow, Normal Accidents: Living with High-Risk Technologies (Princeton, New Jersey: Princeton
Presse universitaire, 1999); Sagan, The Limits of Safety.
65. Allison, “Conceptual Models and the Cuban Missile Crisis.”
66. Sagan, The Limits of Safety, 137.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 30

entirely. Et, most importantly, choice does not have to be removed for brink-
manship to function.

the explanatory limits of accidents

Accidents can only go so far in explaining the sources of bargaining leverage
in brinkmanship. They do not dispatch the problem of choice. After an acci-
dent or mechanical failure causes a crisis to escalate, for a nuclear war to begin,
one leader must decide to be the ªrst to launch deliberately. Even if the
chance event involves the accidental use of a nuclear weapon, a nuclear war
does not start without a nuclear response. Accidents themselves do not elimi-
nate leaders, their credibility, or their deªciencies from the equation.67 Leaders
can increase the odds of an accident, and they can choose not to respond to
provocations that result from transgressions.

In his original climbing analogy, accidents are what made Schelling describe
the brink as a curved slope. But it is more accurate to think of these chance
events as ledges on the cliff. Falling all the way down the abyss at once is un-
likely. If the climber ever fell, she might ªnd herself on a lower ledge, peut-être
injured but still with the acumen to decide whether to plummet another ledge
down or try to climb back up to the top. Néanmoins, the abyss never ceases to
exister, so it remains irrational at any ledge to simply leap into the bottomless
chasm. (In limited war, the climber intentionally leaps to a lower ledge, lequel
we discuss as a subcategory of mechanism 3.)

Two well-established psychological biases suggest that leaders will down-
play the signiªcance of the risk of accidents in their crisis decision-making.
D'abord, the “illusion of control” may plague leaders in crisis. Actors atop hierar-
chies tend to overestimate their own control over events and outcomes.68
Leaders also often suffer from overconªdence in ways that risk expanding
conºict.69 Robert Trivers, Par exemple, points out that human males have a
strong tendency toward overconªdence and illusions of control. He notes that
this tendency offers a strong evolutionary advantage, particularly in combat,
since those who can bring more people to their side by projecting a strong like-

67. We build here on Katzenstein and Seybert, Protean Power, 42.
68. E. J.. Langer, “The Illusion of Control,” Journal of Personality and Social Psychology 32, Non. 2
(1975): 311–328, https://doi.org/10.1037/0022-3514.32.2.311; Robert Jervis, Perception and Misper-
ception in International Politics (Princeton, New Jersey: Princeton University Press, 1976).
69. Dominic D. P.. Johnson, Overconªdence and War: The Havoc and Glory of Positive Illusions (Cam-
bridge, MA: Presse universitaire de Harvard, 2004); Dominic D. P.. Johnson and James H. Fowler, “The
Evolution of Overconªdence,” Nature 477 (2011): 317–320, https://doi.org/10.1038/nature10384.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 31

lihood of victory are more likely to win. They may even win without having to
ªght by getting the other side to back down in fear of overwhelming odds.
People can often tell when enthusiasm is lacking, cependant. Donc, blufªng
or lying to oneself about one’s prospects for victory is the surest way to de-
ceive others, recruit more followers, and thus be more likely to actually win a
ªght. But such illusions also have negative consequences. Illusions of control
can result in illusory pattern recognition; c'est, perceiving patterns that do
not exist, including exaggerating the prospect that others would actually sub-
mit.70 Those who argued that the Iraqis would greet U.S. forces as liberators,
like so many French citizens had done at the end of World War II, fell prey to
this misconception. The problem of overconªdence can plague both sides in a
context of brinkmanship. Both leaders may feel like they are more likely to
emerge victorious from a conºict, and thus may prove more risk acceptant and
willing to escalate than may be wise given the stakes.

The second psychological bias is reºected in Jervis’s argument that leaders
tend to overestimate the adversary’s unity and control. They see others’ behav-
iors as more centralized, disciplined, and coordinated than they actually may
être, while simultaneously recognizing their own system as being more frac-
tured, and expecting the other side to accurately perceive as much.71 During
the Cold War, many U.S. leaders believed that the Soviet Union was a united
monolith, and treated it as such, failing to recognize the importance of divi-
sions about future directions without the Politburo. Those same ofªcials re-
mained fully aware of the fractious nature of the U.S. government and
assumed the Soviets were fully aware that decisions and statements were sub-
ject to revision based on Congressional oversight. Autrement dit, in any situa-
tion, leaders assume that opponents are more united than they are, yet also
believe that those same opponents recognize their own internal divisions. Both
these tendencies reºect failures of imagination and an inability to see outside
one’s own experience. Jervis also warned against conªdent expectations in
the case of nuclear war, writing that “decision makers can never be absolutely

70. Robert Trivers, The Folly of Fools: The Logic of Deceit and Self-Deception in Human Life (New York:
Basic Books, 2011), 22–23. Trivers’s main argument is that humans have evolved to lie to ourselves
in order to more effectively lie to others.
71. Hypothesis 9 in Robert Jervis, “Hypotheses on Misperception,” World Politics 20, Non. 3 (Avril
1968): 454–479, https://doi.org/10.2307/2009777. Hypothesis 10 is related: because a state gets
most of its information about another state via the other state’s foreign ofªce, it tends to assume
that the foreign ofªce’s position is exactly that of the other state. C'est, we assume no rifts or dis-
agreements within the other state.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 32

sure how they or the citizens of their countries will react if nuclear bombs were
to explode even as demonstrations.”72

In sum, in the aftermath of accidents, choices remain. Accidents are insuf-
ªcient as a sole mechanism of brinkmanship. They beg another solution to the
puzzle of agency or choice and must work in tandem with the other mecha-
nisms. In terms of Schelling’s analogy, we want to know: what makes one
climber respond to events that are slipping out of control with an irrational de-
cision to step further toward the abyss while another strives to pull back from
the edge? This is where individual variance in risk acceptance and tempera-
mental variables can come into play in decisive ways.

Mechanism 2: Self-Control

The second mechanism is that of losing self-control. The risk of escalation
comes from within the system, not outside it. Here we see two subcategories.
D'abord, a decider may choose to escalate out of panic or madness. We explore the
possibility of these irrational choices below. Deuxième, Schelling cited false
alarms and the misapprehension of enemy intentions (or a correct apprehen-
sion of the enemy’s misapprehension of its adversary) as sources of biased
decision-making in crisis. In such circumstances, the decider still chooses war,
but this decision is based on incorrect information or misperception about the
enemy’s intentions and actions. The reciprocal fear of surprise attack would be
pernicious in crisis and incentivize a ªrst strike, especially if a state’s warning
systems were postured to accept the risk of type I errors (false positives) être-
cause leaders are worried more about type II errors (false negatives). U.S. early
warning radars, hastily set up to face south during the Cuban missile crisis,
were apt to deliver exactly such false alarms of incoming ballistic missiles.73
This mechanism also seems most affected by Schelling’s notion of “the imper-
fect process of decision” whereby governments, as groups of imperfect units,
fail to move information to where it needs to be on time. We build more on
Jervis’s psychological interpretation of that “imperfect process” of policy-
making, driven by the fact that “the workings of machines and the reaction of
humans in time of stress cannot be predicted with high conªdence.”74

To restate the puzzle, chance is Schelling’s proposed rational solution to the

72. Jervis, Logic of Images, 239.
73. Sagan, The Limits of Safety, chap. 3.
74. Jervis, The Illogic of American Nuclear Strategy.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 33

problem of nuclear threat credibility because it removes the actor from
the equation. But if the actor cannot be entirely removed, then individuals re-
tain some degree of choice that must be explained. To square this circle, we ex-
plain how the emotions of the decision-makers on each side inºuence chance
and risk in brinkmanship.

Most leaders control their emotions most of the time, and some people cer-
tainly have greater emotional awareness and emotional regulation.75 But many
people are not able to completely or consistently control their feelings. In times
of stress and crisis, one of the most destabilizing of these pressures is the desire
for vengeance in the face of an attack.76 The desire for revenge, as universal
as it may be, is not the only emotion that might decisively affect a leader’s
decision-making calculus. Pride, shame, envy, status-seeking, or a desire to de-
fend one’s own or one’s family’s honor might ignite aggression as well. Docu-
mented psychological processes—such as “psychic numbing,”77 which makes
individuals relatively less sensitive to the suffering of larger numbers of peo-
ple affected by a tragedy, and “security prominence,” whereby most people
tend to privilege small increases in security over even much larger beneªts in
other domains—constitute meaningful ways in which thoughts and feelings
systematically deceive most people.78 These notable dynamics cause individu-
als to behave in ways that often contradict expressed values. En outre, feel-
ings and emotions can be wildly unstable, buffeted by the vagaries of limited
attention spans, time pressures, and other immediate situational factors that
people do not consciously control.79

There are at least two areas in which Schelling’s original economically in-
spired notions neglected some critically important and universal aspects of hu-
man psychological architecture. D'abord, he wrote that “to inºict suffering gains
nothing and saves nothing directly; it can only make people behave to avoid
it.”80 This was, après tout, his basic premise in theorizing about the diplomacy
of violence. Notwithstanding that such a construction neglects the existence of

75. Gross, “Emotion Regulation.”
76. McDermott, Lopez, and Hatemi, “‘Blunt Not the Heart, Enrage It.’”
77. Paul Slovic et al., “Psychic Numbing and Mass Atrocity,” in Eldar Shaªr, éd., The Behavioral
Foundations of Public Policy (Princeton, New Jersey: Princeton University Press, 2013), 126–142.
78. Paul Slovic et al., “Virtuous Violence from the War Room to Death Row,” Proceedings of the Na-
tional Academy of Sciences 117, Non. 34 (2020): 20474–20482, https://doi.org/10.1073/pnas.200158
3117.
79. Stephan Dickert and Paul Slovic, “Unstable Values in Lifesaving Decisions,” Frontiers in Psy-
cologie 2, Non. 294 (2011), https://doi.org/10.3389/fpsyg.2011.00294.
80. Schelling, Arms and Inºuence, 2.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 34

sociopaths who revel in inºicting harm, and who may inhabit high ofªces
more frequently than we might like to acknowledge, there are a few things that
are psychologically and biologically naive about this contention. Anyone who
revels in another’s misfortune or who feels the hormonal rush that follows
winning a physical ªght or intense competition knows that inºicting suffering
on others can generate endogenous pleasure, particularly if those others have
caused them harm. Almost all violence is perceived by the perpetrators to be
virtuous, and individuals thus feel quite justiªed in attacking others whom
they consider bad and blameworthy.81 The German word for this feeling is
schadenfreude, which is well intuited by most. What is more, the physical rush
that accompanies a victorious ªght—even if verbal and not physical—is auto-
matic, effortless, immediate, and endogenous.82 The feeling cannot be bought,
nor can it be tamped down easily. Achieving dominance can trigger this strong
biological rush. There are precious few other ways to produce it. En effet, là
is a long evolutionary history that sets people up for the cascade of positive in-
ternal feelings that follow successful combat. Victory feels great; defeat feels
awful. Those who win like the feeling and want to ªght again; those who lose
are less inclined to try again.83 Those who have obtained victory often learn to
like the taste of blood; one need only look at the difªculty of successfully rein-
corporating military personnel with long combat histories back into civilian
society to see the downside of such reinforcement.84 Many who succeed at
inºicting suffering become addicted to the experience.85

A rationalist approach to brinkmanship also fails to account for the per-

81. Slovic et al. “Virtuous Violence from the War Room to Death Row.”
82. Pranjal H. Mehta, Amanda C. Jones, and Robert A. Joseph, “The Social Endocrinology of
Dominance: Basal Testosterone Predicts Cortisol Changes and Behavior Following Victory and De-
feat,” Journal of Personality and Social Psychology 94, Non. 6 (2008): 1078, https://doi.org/10.1037/
0022-3514.94.6.1078.
83. Brian A. Gladue, Michael Boechler, and Kevin D. McCaul, “Hormonal Response to Competi-
tion in Human Males,” Aggressive Behavior 15, Non. 6 (1989): 409–422, https://psycnet.apa.org/doi/
10.1002/1098-2337(1989)15:6%3C409::AID-AB2480150602%3E3.0.CO;2-P.; Michael Elias, “Serum
Cortisol, Testosterone, and Testosterone-Binding Globulin Responses to Competitive Fighting
in Human Males,” Aggressive Behavior 7, Non. 3 (1981): 215–224, https://doi.org/10.1002/1098-2337
(1981)7:3%3C215::AID-AB2480070305%3E3.0.CO;2-M.; Pranjal H. Mehta and Robert A. Joseph,
“Testosterone and Cortisol Jointly Regulate Dominance: Evidence for a Dual-Hormone Hypothe-
sis,” Hormones and Behavior 58, Non. 5 (2010): 898–906, https://doi.org/10.1016/j.yhbeh.2010.08.020;
Pranjal H. Mehta and Robert A. Joseph, “Testosterone Change after Losing Predicts the Decision
to Compete Again,” Hormones and Behavior 50, Non. 5 (2006): 684–692, https://est ce que je.org/10.1016/
j.yhbeh.2006.07.001.
84. This manifests in tragic suicides and high rates of violence.
85. Lionel P. Solursh, Charles A. Meyer Jr., and William P. Nolan, “Addiction to Violence in the
United States Vietnam Combat Veteran,” Med. & Loi. 10, Non. 375 (1991): 1775010.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 35

ceived emotional and psychological, as well as material, values of retaliation,
and the strong and instinctual drive for revenge against those who cause us
harm. Rose McDermott, Anthony Lopez, and Peter Hatemi argue that “the hu-
man psychology of revenge explains why and when policymakers readily
commit to otherwise apparently ‘irrational’ retaliation” such as that envi-
sioned by second-strike nuclear forces.86 The psychology of vengeance under-
lies the stability of nuclear deterrence far more than a rational theory of the
nuclear revolution appreciates.

Policymakers have sometimes intuited the value of such revenge and anger
in coercive bargaining. During the 1973 Yom Kippur War, as the two Cold
War superpowers traded threats of intervention and escalation, in reply to
Secretary of State Henry Kissinger’s warning that “[Washington] would not
accept Soviet troops in any guise” in the conºict, Soviet Ambassador Anatoly
Dobrynin rejoined that in Moscow “they have become so angry, they want
troops.”87 It was not material interest driving policy, he was suggesting, mais
emotion. Secretary of State James Baker also explicitly invoked emotion to de-
ter Saddam Hussein from using chemical weapons in a 1991 meeting with
Iraqi Minister of Foreign Affairs Tariq Aziz. “If the conºict starts, God forbid,
and chemical or biological weapons are used against our forces, the American
people would demand vengeance,” Baker threatened.88 Whether or not they
travail, nuclear strategists have indeed often made these appeals to emotion in
their threat-making. To bolster U.S. deterrence of biological warfare or
cyberattacks against critical infrastructure (“signiªcant non-nuclear strategic
attacks”), John Harvey et al. write that “the American people would not seek
to take any military option off the table in responding to such a catastrophic
attack.”89 The anticipated emotions of domestic publics can tie the hands of
leaders, in some cases to stabilizing effect.

86. McDermott, Lopez, and Hatemi, “‘Blunt Not the Heart, Enrage It.’”
87. Henry Kissinger, Years of Upheaval: The Second Volume of His Classic Memoirs (New York: Simon
and Schuster, 2011), 582, quoted in Barry M. Blechman and Douglas M. Hart, “The Political Utility
of Nuclear Weapons: Le 1973 Middle East Crisis,” International Security 7, Non. 1 (Été 1982):
138, https://doi.org/10.2307/2538692.
88. Quoted in Scott D. Sagan, “The Commitment Trap: Why the United States Should Not Use
Nuclear Threats to Deter Biological and Chemical Weapons Attacks.” International Security 24,
Non. 4 (Spring 2000): 85–115, https://doi.org/10.1162/016228800560318. Scott Sagan and Allan
Weiner also see invoking public pressure to carry out a threat as a common strategy when leaders
would rather bluff. See Christopher A. Ford et al., “Are Belligerent Reprisals against Civilians Le-
gal?,” International Security 46, Non. 2 (Fall 2021): 166–172, https://doi.org/10.1162/isec_c_00422.
89. John A. Harvey et al., “Continuity and Change in U.S. Nuclear Policy,” Real Clear Defense, Fév-
ruary 7, 2018, https://www.realcleardefense.com/articles/2018/02/07/continuity_and_change
_in_us_nuclear_policy_113025.html. See also Nuclear Posture Review (Washington, CC: Office of the

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 36

Once force is used, cependant, emotions can quickly exhaust their strategic
utility. Dans 1898, despite two successive presidential administrations attempting
to avoid conºict with Spain, the destruction of the USS Maine in Havana
harbor so angered the U.S. Congress that it chose war. President William
McKinley later reºected that “but for the inºamed state of public opinion, et
the fact that Congress could no longer be held in check, a peaceful solution
might have been held.”90

In war itself, these emotions can be taken to extremes. Wanton cruelty is a la-
mentably common occurrence in warfare, from Alexander the Great’s siege of
Tyre to the Paciªc Islands of World War II.91 Alexander III of Macedon was
said to have grown so angry at the duration of the siege of Tyre in 332 BC that
he took vengeance on its occupants by allowing his Macedonian troops to
“vent their fury at such a brutal siege upon those who survived;”92 Roman
historian Quintus Curtius Rufus called Alexander’s choice “an irrational tem-
per tantrum.”93 In the words of Schelling, “any use of force tends to be brutal,
thoughtless, vengeful, or plain obstinate.”94 For instance, depictions of the
Crusaders breaching the walls of Jerusalem in 1099 describe how “pent up
emotions found an outlet in murder, rape and plunder, which discipline [était]
powerless to prevent.”95 In the nuclear era, these psychological patterns bear
on the question of whether limited nuclear war can be terminated or whether,
as we suggest, leaders are unlikely to succeed in escalating to deescalate and
instead continue the use of nuclear weapons once instigated.

A second way in which the theory of brinkmanship must better account for
human psychology is in its observation of the value of irrationality. Schelling
wrote, “Another paradox of deterrence is that it does not always help to be, ou
to be believed to be, fully rational, cool-headed, and in control of oneself or of

Secretary of Defense, 2018), https://media.defense.gov/2018/Feb/02/2001872886/-1/-1/1/2018-
NUCLEAR-POSTURE-REVIEW-FINAL-REPORT.PDF; and Ford et al., “Are Belligerent Reprisals
against Civilians Legal?,” in which Sagan and Weiner call this approach “a common strategem,
when making veiled threats of illegal or incredible action, to place the onus on others for executing
those threats.”
90. Richard F. Hamilton, President McKinley, War and Empire, vol. 1, President McKinley and the Com-
ing of War, 1898 (New Brunswick, New Jersey: Transaction Books, 2007). This example is used in
McDermott, Lopez, and Hatemi, “‘Blunt Not the Heart, Enrage It.’”
91. Schelling, Arms and Inºuence, 5. On the brutality of the Paciªc War, see John W. Dower, War
without Mercy: Race and Power in the Paciªc War (New York: Pantheon Books, 1986).
92. Stephen English, The Army of Alexander the Great (Barnsley, ROYAUME-UNI: Pen and Sword, 2009), 148.
93. Stephen English, The Sieges of Alexander the Great (Barnsley, ROYAUME-UNI: Pen and Sword, 2010), 57.
94. Ibid., 5.
95. Schelling, Arms and Inºuence, 9.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 37

one’s country.”96 Nevertheless, Schelling overall indicated that leaders are
more inºuenced by a kind of cost-beneªt analysis that privileges economic
forms of decision-making; we argue instead that leaders tend to favor more
psychological and emotional forces in their decision-making. Not all actors,
even leaders, are always rational. Adolf Hitler provides the classic example;
his hatred of the Jewish people led him to keep using the railroads to ship
them to concentration camps, from the outskirts to the center, when it would
have been much better for his military prospects, especially at the end of the
war, to use that transport to ship men to the front, from the center out.97 Hitler
is usually understood as an atypical leader but other examples abound.
Vladimir Putin’s 2022 invasion of Ukraine, especially his doubling down after
the unprecedented sanctions imposed in its wake and the underperformance
of the Russian military, would hardly seem rational from a cost-beneªt per-
spective. But his actions appear more comprehensible when explained in
terms of his view of his own role in Russian history.

Additional work in psychology validates the notion that people weigh un-
predictable threats more heavily in making decisions. Par exemple, Paul Slovic
notes the importance of risk perception in understanding the nature of the
threats that humans confront. Speciªcally, he notes that people are most scared
of so-called “dread” risks, those that feel uncontrollable or that appear to be
harbingers of unpredictably worse things yet to come. The threat of climate
change might easily fall into this category. Encore, Slovic speciªcally uses nuclear
war as an example of an existential fear outside most people’s control, helping
to explain, at least in part, the dread associated with the risk.98 Next we ex-
plore these psychological consequences of unpredictability by distinguishing
the mechanism of self-control from President Richard Nixon’s infamous “mad-
man theory.”99

comment (and how not) to lose self-control: nixon the “madman”

Schelling’s notion of the rationality of irrationality rests on the idea that get-
ting others to believe that one is irrational and might do something crazy or

96. Ibid., 37.
97. Yaron Pasher, Holocaust versus Wehrmacht: How Hitler’s “Final Solution” Undermined the German
War Effort (Lawrence: University Press of Kansas, 2014).
98. Paul Slovic, “Perception of Risk,” Science 236, Non. 4799 (1987): 280–285, https://est ce que je.org/
10.1126/science.3563507.
99. H. R.. Haldeman and Joseph DiMona, The Ends of Power (New York: W. H. Allen, 1978), 83.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 38

unpredictable may sometimes prove to be a rational strategy if it gets the ad-
versary to back off or back down without a ªght. President Donald Trump
tried, unsuccessfully, to play this game with Kim Jung Un.100 It is a strategy de-
signed to win without incurring the costs of ªghting, if a leader can pull it off.
In this way, irrationality operates as a kind of strategic manipulation.

There are various ways that such a strategy might be pursued, and one of
the most noteworthy is the so-called madman strategy. The madman approach
can be understood as a subtype of Schelling’s larger rationality of irrationality
notion. The idea became famous because Nixon used it to try to drive the
North Vietnamese leadership to the negotiating table during the Vietnam War.
The most authoritative account comes from Nixon’s White House Chief of
Staff H. R.. “Bob” Haldeman’s memoirs:

The threat was the key, and Nixon coined a phrase for his theory which I’m
sure will bring smiles of delight to Nixon-haters everywhere. We were walking
along a foggy beach after a long day of speech writing. He said, “I call it the
Madman theory, Bob. I want the North Vietnamese to believe I’ve reached the
point where I might do anything to stop the war. We’ll just slip the word to
them that, ‘for God’s sake, you know Nixon is obsessed about Communism.
We can’t restrain him when he’s angry—and he has his hand on the nuclear
button’—and Ho Chi Minh himself will be in Paris in two days begging
for peace.”101

Nixon complemented the strategy with a bomber alert meant to look like
the United States was preparing for nuclear war. Operation Giant Lance
tasked eighteen thermonuclear-armed B-52 bombers to ºy toward Soviet
airspace, turn at the last minute, and loiter in an oval pattern nearby.102 The
mission also unleashed its share of accident risks, as one base used untrained
crews to load the bombers while two B-52s “accidently ºew so close to each

100. Daniel W. Drezner, “Trump Likes to Be ‘Unpredictable.’ That Won’t Work So Well in Diplo-
macy,” Washington Post, Novembre 23, 2016, https://www.washingtonpost.com/posteverything/
wp/2016/11/23/trump-likes-to-be-unpredictable-that-wont-work-in-diplomacy/.
101. Haldeman and DiMona, The Ends of Power, 83. See also Schelling, The Strategy of Conºict; Scott
D. Sagan and Jeremi Suri, “The Madman Nuclear Alert: Secrecy, Signaling, and Safety in October
1969,” International Security 27, Non. 4 (Spring 2003): 150–183, https://doi.org/10.1162/016228803
321951126.
102. Sagan and Suri, “The Madman Nuclear Alert,» 174. See also William Burr and Jeffrey P.
Kimball, “Nixon’s Secret Nuclear Alert: Vietnam War Diplomacy and the Joint Chiefs of Staff
Readiness Test, October 1969,” Cold War History 3, Non. 2 (2003): 113–156, http://dx.doi.org/
10.1080/713999987; William Burr and Jeffrey P. Kimball, Nixon’s Nuclear Specter: The Secret Alert of
1969, Madman Diplomacy, and the Vietnam War (Lawrence: University Press of Kansas, 2015); Todd
S. Sechser and Matthew Fuhrmann, “The Madman Myth: Trump and the Bomb,” H-Diplo, ISSF
Policy Series, Mars 22, 2017, https://issforum.org/policy/1-5w-madman.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 39

other that a classiªed Strategic Air Command report later called the opera-
tion unsafe,” writes Sagan.103 Nixon in fact instructed Kissinger to “shake
his head” in a meeting with the Soviet ambassador and say, “I am sorry
Mr. Ambassador, but Nixon is out of control.”104 In this way, Nixon thought
that he could game the Soviets and North Vietnamese by getting them to be-
lieve he was more irrational than he really was, or at least more than he
believed himself to be.

The strategy failed, cependant; and it failed in a way that illuminates the im-
portant role of perceptions of rationality or irrationality in brinkmanship.
Nixon’s threat failed because the North Vietnamese understood Nixon to be a
manipulative, strategic actor. To them, Nixon was being crazy like a fox more
than crazy like a loon. As Nguyen Co Thach—vice minister of foreign affairs
and a top aide to Foreign Minister Le Duc Tho in his negotiations with
Kissinger in Paris—later reported in an interview with Jeffrey Kimball, “he
would like to show to the Vietnamese that he was a changeable [unpredict-
capable] person, that he can surprise—how to say, a big stick surprise. But this
backªred on Nixon, because we saw that Nixon could not have a big stick, être-
cause of the step-by-step withdrawal of American forces. That means the
stick becomes smaller and smaller.”105 Meanwhile, Ambassador Dobrynin
derived a similar inference from Nixon’s attempt to appear unpredictable.
Dobrynin took Nixon’s behavior to indicate that Nixon was not, in fact, un
strong leader.106

Despite its lack of success, Nixon employed this strategy widely, never fully
backing down from it and continuing to believe that it would produce the dip-
lomatic results that he desired. As Kimball writes:

The madman theory lay at the heart of the president’s strategy for dealing with
foreign adversaries, such as North Vietnam and the Soviet Union. This striking
phrase, Haldeman reported, was Nixon’s alternative name for the “principle
of a threat of excessive force.” Nixon thought that military force was an essen-
tial component of diplomacy because of its coercive power, but its coer-

103. Scott D. Sagan, “The World’s Most Dangerous Man: Putin’s Unconstrained Power over Rus-
sia’s Nuclear Arsenal,” Foreign Affairs, Mars 16, 2022, https://www.foreignaffairs.com/articles/
russian-federation/2022-03-16/worlds-most-dangerous-man.
104. Quoted in Sechser and Fuhrmann, “The Madman Myth.” U.S. Department of State, Soviet-
American Relations: The Detente Years, 1969–1972 (Washington, CC: U.S. Government Printing
Ofªce, 2007), 87.
105. Nguyen Co Thach, interview by Jeffrey Kimball, Hanoi, Septembre 24, 1994, in Jeffrey P.
Kimball, The Vietnam War Files: Uncovering the Secret History of Nixon-Era Strategy (Lawrence: Uni-
versity Press of Kansas, 2003), 286.
106. Burr and Kimball, Nixon’s Nuclear Specter, 291.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 40

cive power, he believed, could be enhanced if his opponents could be
convinced that he was capable of or intent upon using extreme force, since this
would suggest that he possessed one or more of the interrelated qualities of
madness. . . . He meant to convey his supposed madness as irrationality, et-
predictability, unorthodoxy, reckless risk-taking, obsession, and fury.107

Dans ces cas, Nixon incorrectly believed that he could control the other
side’s perception of him and thus manipulate chance itself. Et, in the most
crucial case of the North Vietnamese, he was wrong. His belief that he could
trick the enemy into believing that he was unpredictable and irrational pro-
duced the exact opposite effect: it merely convinced them that he was com-
pletely in control and trying to manipulate them, causing resentment and
blowback and making their opposition to him more entrenched. Ironically, être-
cause of his serious drinking, Nixon probably had less control of himself than
either he or the North Vietnamese believed.108

A leader’s sense of his own agency in manipulating chance constitutes a
central aspect of the mechanism of self-control. As Nixon feigned in his origi-
nal madman bluff, he assumed that if the North Vietnamese believed he were
“crazy,” they would fear that he could do anything, including to use nuclear
weapons against them. He thought that this strategy would drive them to ac-
quiesce to U.S. demands. He incorrectly believed that the North Vietnamese
would give in if they thought that he had no reason to spare them.

This assurance argument has been used to explain why a “madman’s”
threats, often expressed by personalist dictators, échouer: such leaders are unpre-
dictable, so they may punish their adversaries regardless of their behavior.109 If
one side genuinely believes that the other is “crazy,” then it would have no
reason to give in because it would not believe that the other side would cease
hostilities if it were to submit. After all, one of the deªning features of mental
illness is its imperviousness to environmental input.110 If a target believes that

107. Kimball, The Vietnam War Files, 15.
108. Rose McDermott, Presidential Leadership, Illness, and Decision Making (Cambridge: Cambridge
Presse universitaire, 2007).
109. Roseanne W. McManus, “Crazy Like a Fox? Are Leaders with Reputations for Madness More
Successful at International Coercion?,” British Journal of Political Science 51, Non. 1 (2021): 275–293,
https://doi.org/10.1017/S0007123419000401; McManus, “Revisiting the Madman Theory,” 976–
1009; Reid B. C. Pauly, “Stop or I’ll Shoot, Comply and I Won’t: Coercive Assurance in Interna-
tional Politics” (PhD diss., Massachusetts Institute of Technology, 2019).
110. We note that there exists a true difference between those who might feign madness for strate-
gic leverage, real or imagined, and those who may have a mental illness or who may be unable to
control or manage their feelings or behavior. An additional complication arises when leaders who
have a mental illness or other forms of neurocognitive compromise may nonetheless believe that

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 41

no beneªt can reliably follow from concession, but costs are certain, then there
is no clear incentive to acquiescence. It stands to reason that successful threats
that leave something to chance must function by a different mechanism than
unpredictability because the target must believe that compliance would lead to
beneªt. Autrement dit, the target must trust that the coercer is predictable
enough to follow through on their commitment to cease hostilities if the target
were to comply. That mechanism likely lies in the emotional domain of trust
underappreciated by Schelling and ignored by most rational models of choice.
Speciªcally, trust can secure what cannot be guaranteed, but it depends on a
relationship that builds over time through iterated interactions of increasing
valeur. This process takes time. En effet, trust can take a long time to develop
and can vanish in an instant.111 The trope suggests that when there is trust,
agreements are unnecessary, and yet agreements cannot exist without some
trust. Negotiations to revive a 2015 nuclear agreement to limit the Iranian nu-
clear program centered on the question of whether the United States would
withdraw again and reimpose sanctions. But here again the key to success lies
in the inversion: it is not merely the opponent’s trustworthiness that matters
for success but one’s own as well. And this is where self-deception, as outlined
by Trivers, can undermine decision-making.

how emotion complicates self-control

Nixon’s frequently quoted madman theory is explicitly emotional, suggérant
that anger could trigger a vitriolic attack. Encore, Haldeman says the mes-
sage that Nixon wished to send was, “We can’t restrain him [Nixon] when he’s
angry—and he has his hand on the nuclear button.”112 But this conception of
the role of emotion in motivation and self-control conforms to the view that
emotions are erratic, unpredictable, harmful, and detrimental. While this may
sometimes be the case, more recent examinations of neuroscience based on
brain lesion studies in particular demonstrate that emotions are necessary for
any form of rational decision-making to take place.113

they are completely rational; whether others see their limitations may depend, at least in part, sur
the skill of their advisers in recognizing and hiding the worst manifestations of illness. And it may
not always be obvious to the target if they are confronting someone who is genuinely mentally ill,
or simply a leader pretending to be beyond reason.
111. For empirical veriªcation of trust asymmetry see Paul Slovic, “Perceived Risk, Trust, and De-
mocratie,” Risk Analysis 13, Non. 6 (Décembre 1993): 675–682, https://doi.org/10.1111/j.1539-6924
.1993.tb01329.x.
112. Haldeman and DiMona, The Ends of Power, 83.
113. Antoine Bechara et al., “Deciding Advantageously before Knowing the Advantageous Strat-

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 42

Psychology offers two potential ways of squaring this circle between the role
of emotion in arousal and the role of emotion in rationality. One way is along
the lines posited by Daniel Kahneman’s dual-process theory of cognition, dans
which he contrasts fast and slow forms of decision-making. In a crisis, because
of time pressure, leaders are likely to disproportionately face decisions that
require fast rather than slow thinking processes. Leaders do not have weeks
or months to study and reºect on an issue before deciding on a response;
instead, a leader under threat or attack must decide quickly, using intuition
and instinct more than studied reºection. Fast processes rely dispropor-
tionately on emotion, although both types are prone to speciªc kinds of errors
and biases.114

The second answer relies on Antonio Damasio’s somatic marker hypothesis,
which argues that through developmental and other experiences, people learn
to attach experiences to feelings without necessitating the consequent physio-
logical arousal that typically accompanies strong emotions. In this way,
emotion can convey meaning without the stress inherent in physiological
arousal.115 Here, aussi, emotion guides decision-making, but it can lead individ-
uals astray if the original association between experience and emotion rests on
a dysfunctional experience. Consider the trauma of child abuse. Given how
unfortunately common adverse childhood events are, the resulting and idio-
syncratic biases are likely endemic. Abused children are much more likely to
grow up to be in abusive relationships and abuse their own children; ils
have unfortunately come to associate certain forms of violence with love.

Ainsi, the effect of emotion on decision-making is nuanced and sometimes
idiosyncratic. It does not operate in the way that a rational theory of brink-
manship would favor. Emotion is necessary to generate motivation;116 without
it, people tend not to possess the will to expend energy and act. People only
spend precious resources on things that they care about, and emotions dictate
what people should care about if they are to maximize their chances for sur-
vival. Those instincts may not align with what classical economics suggest is
“rational,” but they do rest on millennia of natural selection instilling instinc-
tual tendencies that maximize the chances for survival over time. In this way,

egy,” Science 275, Non. 5304 (1997): 1293–1295, https://doi.org/10.1126/science.275.5304.1293; Anto-
nio Damasio, Descartes’ Error: Emotion, Reason, and the Human Brain (New York: Manchot, 2006).
114. Kahneman, Pensée, Fast and Slow.
115. Damasio, Descartes’ Error.
116. Leda Cosmides and John Tooby, “Evolutionary Psychology and the Emotions,” in Michael
Lewis and Jeannette M. Haviland-Jones, éd., Handbook of Emotions, 2nd ed. (New York: Guilford,
2000): 91–115.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 43

emotions that look irrational from a modern standpoint, such as the drive for
status over wealth, may serve humans well in terms of survival and reproduc-
tion but may not be rational from an economic standpoint. Par exemple, le
best way to protect oneself from an enemy may in fact be to annihilate the en-
tire group; that assures that future trouble from that adversary will not be
forthcoming. This strategy indeed appears to have been common even in
the early modern era.117 In the age of nuclear war, such a strategy becomes
more dangerous, and threats that leave something to chance risk more wide-
spread annihilation.

Enfin, a satisfying theory of brinkmanship must account for the fact that
different emotions trigger different perceptions of risk. Par exemple, anger
makes people more risk-seeking, and thus more likely to downplay the nature
of the risks that they confront. It also makes them much more likely to be-
lieve that they will be victorious in a conºict, even if such a belief rests on
pure overconªdence. Consider those who breached the U.S. Capitol during the
insurrection of January 6, 2021. Many rioters became increasingly inºamed
by the contagious anger apparent in the crowd, resulting in violent attacks
against the Capitol Police, threats against members of Congress, and wide-
spread destruction of property. One rioter, a former police ofªcer himself who
told authorities that he arrived at the Capitol with no intention of hurting his
“brothers in blue,” nevertheless had “his angered state further inºamed by the
crowds that day” and ended up committing some of the most violent attacks
against the Capitol Police. His actions were so violent that he was one of
the few arrested who were denied bond.118 Conversely, fearful people have
more pessimistic risk assessments and thus prove more risk averse in
their choices and behaviors.119 Gender differences emerge in these tenden-
cies as well. Par exemple, men are more prone to anger, which predicts more
support for punitive political policies. In contrast, some studies ªnd that
women tend to be more fearful and are thus much more likely to support reha-
bilitative policies.120

117. Surtout, just because a given strategy promoted survival at the margins across human
history does not mean that it serves people well now. Par exemple, modern humans experience
high rates of diabetes as a result of strong intrinsic preferences for sugar.
118. Michael Wilson, “How a Respected N.Y.P.D. Ofªcer Became the Accused Capitol Riot
#EyeGouger,” New York Times, Juillet 27, 2021, https://www.nytimes.com/2021/07/27/nyregion/
capitol-riot-january-6.html.
119. Jennifer S. Lerner and Dacher Keltner, “Fear, Anger, and Risk,” Journal of Personality and Social
Psychologie 81, Non. 1 (2001): 146, https://psycnet.apa.org/doi/10.1037/0022-3514.81.1.146.
120. Barbara A. Gault and John Sabini, “The Roles of Empathy, Anger, and Gender in Predicting
Attitudes toward Punitive, Reparative, and Preventative Public Policies,” Cognition and Emotion
14, Non. 4 (2000): 495–520, https://doi.org/10.1080/026999300402772.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 44

Bien sûr, leaders do not necessarily constitute representative members of
the public. But the processes by which leaders are selected and self-select make
them more, rather than less, prone to the kinds of biases that might affect their
decision-making in negative ways. Traits such as narcissism, Par exemple, sont
more likely to emerge in this population than in other groups.121 These indi-
viduals are much more likely to exaggerate their own skill and sense of con-
trol. Such effects vary by person and some leaders are more susceptible to
certain kinds of biases than others.122 Importantly, different kinds of leadership
may call for different kinds of characteristics. Par exemple, traits that might
serve a military leader well might undermine the success of a political leader.
General George S. Patton provides a good example of a successful military
leader who could command loyalty from his personnel in ways that pro-
duced unprecedented feats of success, as when he marched the Third Army
over 100 miles in the dead of winter to reinforce the Allies during the German
assault on Bastogne. Yet Patton continually got into trouble with command be-
cause he failed to respect the political necessities of working in concert with
the Soviets. En effet, Dwight D. Eisenhower relieved Patton in part because
of his aggressive statements about preventive war against the Soviet Union
after the defeat of Germany.123

These emotional mechanisms serve clear functions that often help facilitate
productive and even enjoyable day-to-day lives. But they may also put leaders
at risk for overestimating their self-control, underestimating the probability
that chance might go against them, misperceiving their own and others’ abili-
ties and intentions, and displaying overconªdence in their choices. Leaders
may believe, like Nixon did with the North Vietnamese, that emotions can
be used to generate leverage in bargaining. But such assessments are often
wrong. This inability to recognize the effect of one’s own behavior on another
encourages leaders to underestimate the level of unpredictability that they cre-
ate in generating chance and leverage. This misperception results precisely
from their belief that they have more control than they do, and an overestima-
tion of the effect of their behavior on others and their likelihood of victory in a
contest of wills or armaments.

121. Zoltán Fazekas and Peter K. Hatemi, “Narcissism in Political Participation,” Personality and
Social Psychology Bulletin 47, Non. 3 (2021): 347–361, https://psycnet.apa.org/doi/10.1177/014616
7220919212.
122. While there may be a distribution of such effects, the process by which leaders emerge tends
to select for certain traits over others.
123. Carlo D’Este, Patton: A Genius for War (New York: HarperCollins, 1995), 736.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 45

Emotions underpin motivation in useful ways. Encore, when leaders are narcis-
sistic, lack emotional self-awareness or resilience, or fail to link their behavior
to their best interests, certain emotions such as pride, shame, envy, anger, rejec-
tion, or hurt diminish the quality of decision-making. Deªcits in emotional
self-awareness and control result in the inability to make choices consistent
with best interests. The consequence is not only numerous lost opportunities
but increased risk of devastating outcomes.

Mechanism 3: Control of Others

The third mechanism of brinkmanship is that of a decider losing control be-
cause others make the choice to escalate without the decider’s input or knowl-
bord. Here again there are two subcategories. D'abord, Schelling identiªed
mischief as a means of war onset. Someone else (perhaps irrational but he left
that unsaid) makes the choice for war instead of the leader. When the authority
or ability to escalate is shared among multiple actors, these risks increase.
A leader may or may not be aware of having created these risks. Pour dans-
position, nuclear command-and-control arrangements that pre-delegate to mili-
tary commanders the authority to use nuclear weapons exacerbate these risks
by introducing more vectors for others’ individual choices. And those individ-
uals are just as susceptible to the irrational decision-making described in the
prior section.

Some might conceive of this risk as a principal-agent problem, whereby the
principal loses control of an agent. In the military domain it has been more
speciªcally theorized as a lack of civilian control over military forces.124 An
overzealous local commander with the ability to use nuclear weapons ªts the
bill—a Dr. Strangelove scenario. But short of nuclear war, military command-
ers have routinely stretched the purview of their offensive missions beyond
the authorization of civilian leaders—the 1982 Israeli invasion of Lebanon,
General MacArthur’s Yalu River bombing campaign in the Korean War, ou
the German Navy’s submarine campaign during World War I all exceeded to
some extent their operational intent.125 Indian commanders along the Line of

124. Military organizations prefer offensive operations, sometimes without the approval of civil-
ians. See Posen, Inadvertent Escalation; Stephen Van Evera, “The Cult of the Offensive and the Ori-
gins of the First World War,” International Security 9, Non. 1 (Été 1984): 58–107, https://est ce que je.org/
10.2307/2538636; Jack Snyder, “Civil-Military Relations and the Cult of the Offensive, 1914 et
1984,” International Security 9, Non. 1 (Été 1984): 108–146, https://doi.org/10.2307/2538637.
125. Posen, Inadvertent Escalation, 18–19. On the German U-boat campaign, see Fred Charles Iklé,
Every War Must End (New York: Columbia University Press, 1971). Standard operating procedures

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 46

Control in Kashmir have provoked skirmishes with Pakistani forces without
approval or instruction from New Delhi.126 The extent of Prime Minister
Nawaz Sharif’s knowledge of military plans in advance of Pakistan’s invasion
of Kargil in 1999 is also debated.127

Even within military organizations, signiªcant commands have been dis-
obeyed or countermanded by individual leaders. Japanese General Tomoyuki
Yamashita did not wish to commit troops to the defense of Manila in 1944, pre-
ferring instead to husband dwindling military resources for delaying actions
in more favorable terrain. But local commander Rear Admiral Sanji Iwabach,
whose sense of honor made him “determined to ªght for Manila,” disobeyed
the order to abandon the city and demolish the bridges.128 Superior Japanese
ofªcers had no choice but to support as best they could the doomed defense.
Air Marshal Sir Charles Portal also gave clear instructions to Sir Arthur
“Bomber” Harris in September 1944 to concentrate Royal Air Force bombers
on oil and transportation targets in Nazi Germany. Harris refused and contin-
ued with his effort to turn German cities into rubble.129

The second subcategory is limited war as a generator of risk. Although it
stretches our scope conditions,130 we place this mechanism into the category of
controlling others because deciders may be fully in control of their own forces,
while the element of chance lies in the possibility that they could unknowingly
cross one of the adversary’s red lines; and deciders may not even give suf-
ªcient weight to the probability that their actions may do so. The choice to
escalate further in response now rests with the enemy. En effet, this last subcat-
egory of limited war overlaps with all the others, since each mechanism dis-

also famously exceeded civilian intent during the Cuban missile crisis. Allison, “Conceptual
Models and the Cuban Missile Crisis.”
126. Saikat Datta, “DNA Exclusive: Uri Commander’s Forceful Retaliation Led to Beheadings?,»
DNA, Novembre 21, 2013, https://www.dnaindia.com/india/report-dna-exclusive-uri-command-
ers-forceful-retaliation-led-to-beheadings-1787448; Happymon Jacob, Line on Fire: Ceaseªre Viola-
tions and India-Pakistan Escalation Dynamics (New Delhi: Presse universitaire d'Oxford, 2018).
127. S. Paul Kapur, “Nuclear Proliferation, the Kargil Conºict, and South Asian Security,” Security
Études 13, Non. 1 (2003): 84n22, https://doi.org/10.1080/09636410490493868.
128. Ronald H. Spector, Eagle against the Sun (New York: Free Press, 1985), 523.
129. Only 6 percent of Sir Arthur “Bomber” Harris’s bomb tonnage in the next month (Octobre)
was devoted to oil targets. Charles Webster and Noble Frankland, The Strategic Air Offensive against
Germany 1939–1945, vol. 3, Victory, Part 5 (Londres: Her Majesty’s Stationary Ofªce, 1961), 58, 67.
We thank Williamson Murray for this example.
130. Our focus has been on one side’s decision-making rather than strategic interaction. Encore, lim-
ited war was so core to how Schelling and others thought about brinkmanship that it is impossible
to overlook. We ªnd it within our scope to explain the decisions of either side to engage in limited
war, rather than the consequences of their choices in combination.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 47

cussed in this article might operate in the context of limited war and increase
the chances of an irrational choice to escalate.

Schelling considered limited nuclear war as brinkmanship in this way.131
Deliberately crossing the nuclear threshold demonstrates resolve, even if not
enough resolve to jump off the cliff completely. The initial use of limited nu-
clear force is conceived of by the escalator as rational. In terms of the moun-
taineering analogy, one climber intentionally jumps down to a lower ledge (comme
opposed to falling by accident), dragging the other climber with them. But just
because one climber jumped to a lower ledge and dragged the other down
does not necessarily make inevitable the second jump into the abyss to doom
them both. The ªrst use of a nuclear weapon does not automatically lead to
strategic nuclear exchange; it is merely an action that is deliberately risky.

Limited war can generate leverage by giving the adversary the “last clear
chance” to avoid disaster. It surrenders choice to others.132 President Kennedy
conceived of limited war in this way during the 1961 Berlin crisis. In October,
he penned instructions to Supreme Allied Commander Lauris Norstad in
which he explained, “It seems evident to me that our nuclear deterrent will
not be credible to the Soviets unless they are convinced of NATO’s readiness
to become engaged on a lesser level of violence and are thereby made to real-
ize the great risks of escalation to nuclear war.”133 Actions could include con-
ventional air operations to “show the Soviets visibly higher risks of nuclear
war,” and “selective nuclear attacks for the primary purpose of demonstrating
the will to use nuclear weapons.” In giving these instructions, Kennedy
fully acknowledged that “the Allies only partially control the timing and scale
of nuclear weapons use,” since brinkmanship might “prompt unrestrained
pre-emptive attack.”134

De plus, limited war as brinkmanship need not exclusively involve the use
of nuclear weapons. Chinese brinkmanship strategies, par exemple, have ex-
plicitly relied on nonnuclear assets—conventional missiles, espace, and cyber

131. Schelling, Arms and Inºuence, 108–109; Schelling, The Strategy of Conºict, 193. See also Robert
Powell, “Nuclear Brinkmanship, Limited War, and Military Power,” International Organization 69,
Non. 3 (Été 2015): 589–626, https://www.jstor.org/stable/24758314.
132. Schelling, The Strategy of Conºict, 37.
133. “Letter from President Kennedy to the Supreme Commander, Allied Powers Europe
(Norstad),” Washington, CC, Octobre 20, 1961, in Charles S. Sampson, éd., Foreign Relations of the
États-Unis, 1961–1963, vol. 14 (Washington, CC: U.S. Government Printing Ofªce, 1993), 521. Nous
thank Matthew Bunn for this source.
134. “Letter from President Kennedy to the Supreme Commander,» 523.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 48

assets—to posture to ªght limited wars and place the “last clear chance” to
avoid nuclear escalation onto the adversary.135 Inadvertent escalation also ªts
into this category, as conventional force can risk a nuclear response if it de-
grades the adversary’s nuclear forces.136

Within the mechanism of control over others, again, the role of emotion can
calibrate motivation differently for those at the center versus those at the pe-
riphery of a ªght, as indeed occurred with Nixon, the Soviet Union, et le
North and South Vietnamese governments. Sometimes the enemy gets a vote
that leaders did not plan for. The Battle of Navarino in 1827, par exemple, être-
gan when an allied ºeet comprised of British, French, and Russian vessels in-
tended to conduct a naval demonstration in Navarino Bay, but “the enraged
Turks” ªred upon it.137 Other times leaders lose control of their own militaries.
Dans 1576, angry Spanish soldiers who had not been paid by their bankrupt king,
Philip II, pillaged the city of Antwerp, a massacre that so enraged the Low
Countries that they put aside their divisions to unite against the invaders, et-
dermining a decade of Spanish strategy in the Eighty Years’ War.138 And mili-
tary commanders sometimes even ªnd it useful to create such risks by stirring
up emotions in combat and crisis, as Henry V does in Schelling’s favorite pas-
sage from Shakespeare’s play, rallying his soldiers to “conjure up the blood”
and “disguise fair nature with hard-favour’d rage.”139

The British also suffered this lesson in the context of alliance politics at the
1757 surrender of Fort William Henry during the Seven Years’ War.
Commanding British ofªcers responsible for the 2,300-person garrison reached
terms of surrender with French General Louis-Joseph de Montcalm and read-
ied a retreat to nearby Fort Edward. Outside the walls, cependant, British col-
umns were harassed by France’s multi-tribal indigenous allies, resulting in

135. Since 2014, China has reined in these risks and adopted a “calibrated escalation” posture.
Fiona S. Cunningham, “Strategic Substitution: China’s Search for Coercive Leverage in the Infor-
mation Age,” International Security 47, Non. 1 (Été 2022): 46–92, https://est ce que je.org/10.1162/
isec_a_00438. In laying out the elements of a brinkmanship posture, distinguished from a cali-
brated escalation posture, Fiona Cunningham’s primary brinkmanship mechanism is that of lim-
ited war: the chance that the other side will respond with nuclear force.
136. Posen, Inadvertent Escalation; Caitlin Talmadge, “Would China Go Nuclear? Assessing the
Risk of Chinese Nuclear Escalation in a Conventional War with the United States,” International Se-
curity 41, Non. 4 (Spring 2017): 50–92, https://doi.org/10.1162/ISEC_a_00274.
137. Jervis, Logic of Images, 237.
138. The Eighty Years’ War (c.1566–1648), also called the Dutch Revolt, was a conºict in which the
Dutch Republic successfully fought for independence from Spain. Geoffrey Parker, The Dutch Re-
volt (New York: Penguin Books, 1985), 178.
139. William Shakespeare, Henry V, acte 3, scene i, in Stephen Greenblatt, éd., The Norton Shake-
speare: Based on the Oxford Edition: Histories (New York: W. W. Norton, 1997).

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 49

dozens of casualties. The “massacre,” thereafter embellished in British and
then U.S. lore,140 was later used to justify vicious retaliations against indige-
nous peoples. But rather than a coordinated deception, the bloody morning of
Août 10, 1757, was the result of a chaotic disagreement between the French
and their allies over the legitimacy of British surrender. Indigenous warriors,
who had canoed hundreds of miles to join the ªght for no other compensation
than the plunders of war, felt betrayed and robbed of their earned spoils.141
“They could not fathom French behavior in wasting their victory and protect-
ing their enemies from their allies,” writes historian Ian Steele. “They resented
the European conspiracy, which had defrauded them of their agreed share
of the loot in the fort.”142 Coordination between French ofªcers and tribal
chiefs had always been ªtful.143 Indeed, Montcalm knew the risks and sought
to use them to his advantage in bargaining over the fort’s surrender. “Once
[the French] batteries were in place and the cannon ªred, perhaps there would
not be time, nor would it be in [mon] power to restrain the cruelties of the mob
of Indians of so many different nations,” wrote Montcalm to the British de-
fenders of the fort.144 Still, the British resisted for one more week. When they
ªnally surrendered, they believed Montcalm’s assurances of safe passage. Mais
Montcalm’s attempt to convey the terms of surrender to assembled chiefs and
translators clearly failed. During the British retreat, a chaotic scufºe culmi-
nated in short-lived but brutal violence after someone let loose a “dreaded war
whoop that was an intertribal signal to attack.”145 The French lost control.

Conclusion

By distinguishing between “chance” and “choice,” this article elevates the
long-hidden psychological and emotional elements that remain central to a
more comprehensive understanding of brinkmanship and the “threat that
leaves something to chance” in nuclear crises. We have brought to the study of
nuclear strategy what decades of psychological research have to say about

140. James Fenimore Cooper, The Last of the Mohicans: A Narrative of 1757 (Londres: John Miller,
New Bridge Street, 1826).
141. Ian K. Steele, Betrayals: Fort William Henry and the “Massacre” (Oxford: Oxford University
Presse, 1990), 82, 113.
142. Ibid., 113.
143. Ibid., 104.
144. Fred Anderson, Crucible of War: The Seven Years’ War and the Fate of Empire in British North
America, 1754–1766 (New York: Vintage Books, 2000), 192.
145. Steele, Betrayals, 117.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

International Security 47:3 50

how humans are likely to behave during a “war of nerves,” as Schelling explic-
itly called brinkmanship.146

We distilled three mechanisms for how chance becomes leverage in crisis:
accidents, self-control, and control of others. Critique, these sources of chance
do not have to wrest choice away from leaders in crises to generate leverage.
Psychological and emotional variables are sources of risk and uncertainty,
even when leaders retain the option to use nuclear weapons. The brink is not
equally abhorrent to all leaders in all crises. Said another way, the point of de-
cision does not have to be eliminated for risk of catastrophic destruction to re-
main. This is because leaders cannot always accurately assess the chances that
they take, the extent to which they can control outcomes, or their ability to
force concessions on the part of opponents. They are uncertain if the other side
will actually do what it says, and they cannot know if they properly under-
stand the signals that they receive. That both sides do not, and cannot, know if
the other side is telling the truth raises uncertainty for everyone. Heightened
emotions exacerbate this uncertainty. The structure of the situation imposes
uncertainty and instability and inevitable human emotional responses to
threat enhance and strengthen these effects. In short, crises are even more un-
stable than the traditional theory of brinkmanship posits.

A focus on psychological as opposed to economic models of decision-
making offers an important rebalancing in the literature on nuclear strategy
and deterrence. If scholars misidentify the mechanisms by which brinkman-
ship works, and therefore the risks of escalation, then any proposed solutions
may be incomplete at best and inaccurate at worst. An economic approach
might encourage leaders and scholars to focus on material solutions to crisis
instability rather than decision-making solutions. Par exemple, arms control
agreements that eliminate ªrst-strike weapons systems or increase the ºight
times of missiles are worthy goals but should be paired with interventions
that constrain decision-makers from having unilateral authority over nuclear
weapon use. These might include, Par exemple, expanding two-person rules to
every point in the decision chain, implementing lengthier protocols to check
and recheck orders, and maintaining hotlines for crisis communications be-
tween adversaries. Reducing haste is important but the augmented decision-
making time must also be used to disrupt irrational processes. As research
on crisis stability continues to assess the effects of new technology, new net-
work effects, and new vulnerabilities of the information age,147 scholars should

146. Schelling, The Strategy of Conºict, 196.

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3

The Psychology of Nuclear Brinkmanship 51

not overlook how crisis events are interpreted, digested, and acted upon
by humans.

By illuminating the mechanisms and relative roles of uncertainty and
decision-making in threats that leave something to chance, we also open sev-
eral avenues of further research. D'abord, scholars might further stipulate the
relationship between speciªc dispositional characteristics of leaders and brink-
manship. Certain kinds of leaders may possess different risk-tolerance thresh-
olds and hold different beliefs about how much choice they have, how much
choice is necessary, and how much can be left to chance. Some leaders may
think that they can simply walk back from the brink at any point. Second and
related, the renaissance of studies on nuclear command and control needs to
confront decision-maker psychology more fully.148 Speciªc types of leaders
may embrace (ou non) changes to nuclear doctrine to manipulate risk—
deploying tactical nuclear weapons, preparing ºexible response options, et
pre-delegating launch authority.

A third avenue for future work might examine how the concept of brink-
manship extends to limited contexts outside the nuclear domain or at least
beyond nuclear-armed dyads. South Korea, par exemple, is embracing a revo-
lution in military affairs that integrates intelligence, surveillance, and recon-
naissance assets with long-range precision weapons to push the bounds of
what was once thought possible with conventional counterforce.149 How these
capabilities might permit nonnuclear Seoul to manipulate the risk of disaster
in a crisis with North Korea is well worth exploring.

Enfin, some studies of brinkmanship conclude that threats that leave some-
thing to chance are empirically rare because leaders seek to maintain control
in crisis. Fuhrmann and Sechser go so far as to call it “the brinkmanship
myth.”150 Our approach suggests that scholars are looking for evidence in the
wrong place. Chance can be all around us, even when we try to exercise choice
over its consequences.

147. Cunningham, “Strategic Substitution.”
148. Feaver and Arceneaux, “The Fulcrum of Fragility”; Arceneaux, “Beyond the Rubicon.” On
digitization risks, see Herbert Lin, Cyber Threats and Nuclear Weapons (Stanford, Californie: Stanford Uni-
versity Press, 2021).
149. Ian Bowers and Henrik Stålhane Hiim, “Conventional Counterforce Dilemmas: South Ko-
rea’s Deterrence Strategy and Stability on the Korean Peninsula,” International Security 45, Non. 3
(Hiver 2020/21): 7–39, https://doi.org/10.1162/isec_a_00399.
150. Sechser and Fuhrmann, “The Madman Myth.”

D
o
w
n
o
un
d
e
d

F
r
o
m
h

t
t

:
/
/

d
je
r
e
c
t
.

je
t
.

e
d
toi

/
je
s
e
c
/
un
r
t
je
c
e
–
p
d

F
/

4
7
3
9
2
0
6
9
2
7
0

/
je
s
e
c
_
un
_
0
0
4
5
1
p
d

b
oui
g
toi
e
s
t

o
n
0
7
S
e
p
e
m
b
e
r
2
0
2
3
Télécharger le PDF