The Perceptions of Robert Jervis: An Appreciation

Ken Booth and Nicholas J. Wheeler
Ken Booth and Nicholas J. Wheeler are the authors of The Security Dilemma: Fear, Cooperation and Trust in World Politics (Houndmills and New York: Palgrave Macmillan, 2008)


As we write this appreciation of Robert Jervis, weeks after his death at the end of 2021, media outlets are full of stories of major powers readying themselves for the expansion of a war across the Russia-Ukraine border. TV news reports soldiers, tanks, warships, and aircraft carrying out their drills, while out of sight we know that cyber warfare specialists will be preparing for the worst. On the airwaves, we hear the obscene sounds of rockets firing, fighter planes taking to the air, and the rat-tat-tat of rifles and machine guns. History is reawakening, and major war in Europe is once again thinkable if – we hope – unlikely. Meanwhile, as this nightmare invades our senses, diplomats parade concern, ratchet up threats and counter-threats, play the mutual blame game, offer conciliatory moves, plead innocence, and flex what they hope are the right muscles.

In the Introduction to his second major book, Perception and Misperception in International Politics (1976), Robert Jervis emphasised the causes and characteristics of misperception among decision-makers, and he demonstrated through numerous illustrations why this really matters. It was vital work, he argued, because specialists in the discipline of International Relations (IR) tended to assume that ‘decision-makers usually perceive the world quite accurately and that those misperceptions that do occur can only be treated as random accidents.’ In Jervis’s book, and indeed for the rest of his academic career, he sought to show that this perception was ‘incorrect’.In all this uncertainty, two things are for sure: first, misperception will be rife, as signals and counter-signals will not be interpreted accurately; and second, with the passing of Robert Jervis, we have lost one of the very best guides in understanding the relationships between signalling and misperceptions in relations between states. Since the 1970s Jervis taught his students, the profession, and sometimes his government, how to think more clearly about situations such as the one facing us in the far east of Europe: the dangers in the methods by which President Putin is manipulating fear; the problems with the ways Western decision-makers are seeking to dampen things down by what they hope is deterrence combined with reassurance; the spiralling of mutual mistrust and distrust; the undertaking of tactical and strategic moves that are open to misperception; the dynamics of ‘the other mind’s problem’ (trying to get inside the heads of others); the ambiguous meaning of weapons systems and deployment patterns in relation to whether they convey offensive or defensive motives and intentions; the challenge of accurate signalling, by word and by action, when their meaning is ultimately determined by the possible target not by the sender; and the unpredictable outcomes of pursuing interests through military moves in an environment of interlocking and escalating fear.

Jervis’s intellectual canvas was huge. It spanned the foundational concept of the ‘security dilemma’ (originated by John Herz, who introduced it into the literature in 1950), ‘security regimes’ (a concept Jervis himself invented), ‘security communities’ (developed by Karl Deutsch and his co-researchers in the late 1950s), and the ‘nuclear revolution’ (where Jervis continued the pioneering work of Bernard Brodie, Glenn Snyder, and Thomas Schelling). Jervis’s first published volume (which he subsequently referred to as ‘the signalling book’) was The Logic of Images in International Relations (1970). It was based on his PhD, and its ambitious ‘driving idea’, in his own words, was ‘why should we believe anything?’

**********

As a result of Jervis’s stellar career at the heart of the study of IR in the United States, he knew that he was a successful academic. We are less sure whether he fully realised how important he and his work has been in the intellectual and indeed personal lives of so many other scholars, from students at the start of their careers to long-established professors. The outpouring of warm and deeply-felt tributes to ‘Bob’ on social media and elsewhere since his death, is testimony to his inspiration as an outstanding teacher and mentor.

Yet Jervis had much wider impact and renown than that treasured by his closest friends, colleagues, and students. In the very first conversation the authors of this appreciation had with each other following the news of his death, we recalled that we had been in an imaginary conversation with ‘RJ’ -through his writing – almost since the moment we first met in 1985. In our subsequent 30-plus years of conversations and collaborations, still ongoing, Robert Jervis sat on our shoulders. He will remain there. He was also there even before our first meeting as our independent academic interests had led each of us to have read Perception and Misperception with great care: for one of us (NJW) this was the result of student enthusiasm, while for the other (KB) it was in the course of writing a book on strategy and ethnocentrism.

At the core of our shared interest, so long ago, was the phenomenon of the security dilemma. From early in our teaching and research careers we recognised it as what we came to call the ‘quintessential dilemma’ in relations between decision-makers at the international level of world politics. In Jervis’s work we had discovered a kindred spirit, and one who was already very far ahead in his journey.

**********

From the mid-1970s onwards, Jervis became the towering figure writing about the security dilemma in IR. In our view – as well as that of some others – the concept itself had not then made the impact it should have done after its first theorisation by John Herz and Herbert Butterfield in the 1950s. In the decades since it got better, and theorists in the United States sometimes made a passing reference to Herz, and even Thucydides, but never to Butterfield. But with Jervis’s intervention, the concept did become harder for the US discipline to ignore, though it still often was. An exception was one international conference held in the United States at the start of the millennium, where a panel was organised on the work of John Herz. Butterfield’s contribution for once got a mention. We recall that in the subsequent discussion one US scholar frankly admitted that until that point, he had believed it was Jervis who had invented ‘the security dilemma’ – an anecdote that speaks both to Jervis’s influence and to the insularity in the discipline in parts of the US academy.

Jervis’s influence on thinking about the security dilemma has been colossal. He brought theoretical rigour to the pioneering ideas of Herz and Butterfield and did so by embracing an interdisciplinary approach. In particular, his research in IR was immersed in the latest thinking in political psychology. The result was that his explorations into the perceptual dynamics of political relations under anarchy were carried out with a sophistication that had not been seen before.

The crux of Jervis’s building on the work of Herz and Butterfield was the formulation of what he called the ‘spiral’ and ‘deterrence’ models. Through them, he sought to explain how decision-makers in one state tried and often failed in navigating the uncertainties and risks about the current and future intentions of those states with the military capability to inflict harm against them. The ‘spiral model’ was largely a sophisticated elaboration of Butterfield’s earlier notion of ‘Hobbesian fear’, resting on the assumption that escalating insecurity could result from decision-makers failing to understand the true nature of their situation. In particular, he pointed out that decision-makers were apt to interpret each other’s behaviour as indicating aggressive intent, when the actions being taken may well have been initiated for defensive purposes. As Jervis told us in an interview in 2014, the spiral model and the security dilemma were synonymous in his own thinking.

The spiral model was driven by the mutual misperceptions between adversaries of each other’s intentions, and at its root was the insecurity and fear arising from ‘the anarchic setting’ of international relations. Crucially, he wrote that ‘neither party appreciates how their actions contribute to mutual fear’. In such circumstances, better signalling through words and action was the challenge for decision-makers seeking to wind down the potential escalation of mutual distrust: but first the parties involved had to appreciate that they were indeed potentially trapped in a ‘spiral’.

If the spiral model developed Butterfield’s argument about the ubiquity of the security dilemma, the ‘deterrence’ model built on Herz’s conclusion that the security dilemma did not explain all conflicts. Using the example of Nazi Germany, Herz argued that ambition and not fear might be the driver of aggressive behaviour; if this were the case, deterrence was the required response on the part of the threatened party or parties. This was because, according to the model’s assumption, aggressor or revisionist powers cannot be converted into ‘status quo’ states through concessions or conciliatory signalling: deterrence alone has the potential to contain. Importantly in this view – and particularly prominent during some phases of the Cold War – was the lesson many took from the 1930s, namely that ‘appeasement’ of any kind, notably concessions to dictators, only fed their appetites.

The basic challenge for decision-makers, as posed by Jervis, was therefore to determine accurately whether they are in a spiral or deterrent situation with potential adversaries, and then to adopt the appropriate response. The two models structured his thinking, but like all models he acknowledged that they simplified reality.

The spiral model (‘the security dilemma’ in Jervis’s view) was predicated on the assumption that conflict may be driven by mutual misperceptions, but that these are potentially correctable through a more subtle understanding of security dilemma dynamics. In particular, he argued that decision-makers need to appreciate how their own actions might contribute to spirals of insecurity as a result of unwittingly provoking fear in the minds of others. Having such an appreciation is what we call ‘security dilemma sensibility’.

Despite being a major step forward in understanding security dilemma dynamics, the spiral and deterrence models have always been open to the criticism that they are too dichotomous; they risk falling into the temptation of seeing states (in Charles Glaser’s terminology) as either ‘security-seekers’ or ‘greedy’. Critics asked: what about the possibility that states believe they can only be secure if they expand at the expense of others? In other words, what if each state in a dyad believes its security requires the insecurity of others?

Jervis himself explored these complex questions in great depth over the decades. His position was that an adversary could be a ‘security-seeker’ or a ‘greedy’ state or both. The latter might be the case, for example, where a state had different intentions in different issue-areas, or different intentions at different points in time. The United States and the Soviet Union during the Cold War were complex cases in this regard.

The ‘adversary partners’ in the Cold War might have been security-seekers in relation to certain issue-areas (notably nuclear weapons) where policy-makers on both sides shared and on occasion recognised a mutual interest in arms control: but at the same time they might have considered themselves to be in a global ideological confrontation in which there could be no predictable stability or path to mutual security. When the latter was the case both adversaries would seek to try and undermine the other, and in ways that would make long-term cooperation impossible. Such a relationship is what Jervis came to call the ‘deep security dilemma’ (2001).

Jervis characterised the Cold War as a ‘deep security dilemma’, with one of its defining features being the ideological fundamentalism generally shown (in words if not action) by the leaders of both superpowers. The corollary of decision-makers seeing their own behaviour as peaceful in intent, brimming with defensive self-images, has often been a failure to understand how others might see them as ‘enemies’ and ‘aggressive’. Appreciating this dynamic is why security dilemma sensibility is so important on the part of leaders if cooperative moves are to make any progress.

A group of mostly US scholars built on Jervis’s work and explored the practicalities of successfully signalling peaceful/defensive intentions in a context where conflict was believed to be driven solely by spiral model dynamics. Ideas included ‘normal methods’ of cooperation-building such as dividing up a large transaction into a series of small ones; bolstering the weight of non-provocative defence capacity; encouraging transparency; and so on. The result, it was hoped, would be to alter the mindset of the adversary, and encourage cooperation rather than conflict.

**********

Jervis was ready to admit that his upbringing and education during the Cold War had predisposed him towards the cautious end of the spectrum on the scope for harmony in international security. This was sometimes evident in his thinking about ‘regime theory’.

First developed in the United States in the late 1970s in relation to political economy, Jervis led the application of regime theory to the field of international security. His chief contribution was in a reference-point article in 1982 entitled ‘Security Regimes’. In it he defined a ‘security regime’ as ‘those principles, rules, and norms that permit nations to be restrained in their behaviour in the belief that others will reciprocate’. In other words, it is ‘a form of cooperation that is more than the following of short-run self-interest’. Jervis’s words became the standard formulation, and over the next decade his ideas were built upon by a range of other international security theorists.

In discussing the preconditions for the growth of a security regime, Jervis foregrounded the scope for misperceptions when interpreting the offence/defence ambiguities of the weapons and strategies of a potential rival state. Even if regime formation is achievable, however, he pointed out that a variety of contingent and structural factors might conspire to set in motion a spiral of mutual distrust, resulting in the eventual collapse of the cooperative edifice. Based in part on his study of the decline of the Concert of Europe in the first half of the nineteenth-century, he warned that it is not enough to control the risks of war: by failing to become institutionalised and not developing supranational loyalties, he concluded that ‘the Concert may have contained the seeds of its own destruction’. A century later he did not regard US-Soviet relations, even during periods of détente, to have met the criteria of being a security regime.

For those drawn to conservative understandings of international politics, awareness of the potential for regime breakdown is always likely. Today, such pessimism is difficult to escape even on the part of those with more open perspectives on international security. All must ask, looking at the present crisis in eastern Europe, whether the security order that developed at the end of the Cold War, and lasted 30 years, is now suffering from having failed to eradicate the seeds of its own potential destruction. This refers to the policy-makers of the leading states, especially in the 1990s, falling short in embedding trust. This was evident in the ostensible ‘humiliation’ of Russian leaders and their new state, and the apparent complacency if not hubris of the West. Are we therefore now witnessing a desire to make gains at each other’s expense in an environment where the restraints of what was once trumpeted as a ‘new post-Cold War’ order are losing whatever traction they once had?

**********

In considering how the conflictual pressures of life under anarchy might be further dampened down, Jervis emphasised the need for a comprehensive understanding of the ‘nuclear revolution’ – the focus of a book he published in 1989. Like many of his generation, nuclear weapons had been a pressing interest and concern from his youth, and in a series of publications he discussed the complex issues relating to their stabilizing potential. Above all, he thought Mutual Assured Destruction (MAD) ruled out major wars between major powers. More originally, he claimed that MAD threatened such catastrophic consequences that it escaped the ambiguity of offence-defence differentiation in security dilemma thinking at the strategic nuclear level. He concluded in a 1978 article that ‘as long as states believe that all that is needed is a second- strike capability, then the differentiation between offensive and defensive forces that is provided by reliance on SLBM’s [submarine-launched ballistic missiles] allows each side to increase its security without menacing the other’. These views were opposed by those identified with ‘offensive realism’ and ‘nuclear war-fighting’ positions, who continued to claim that there was advantage to be had by securing dominance at higher levels of nuclear escalation. If the logic of anarchy compelled the superpowers to compete in this way, they argued, nuclear weapons developed and deployed (and potentially used) with discrimination, could still have strategic leverage.

Below the balance at the strategic nuclear level there remained the apparently unresolvable uncertainty of the security dilemma at the level of conventional forces. Here, in the late 1970s, Jervis accepted the security dilemma still existed: ‘On issues other than defense of the homeland, there would still be security dilemmas and security problems.’ But, he added, with the stability he believed MAD ensured, ‘the world would nevertheless be safer than it has usually been’. These views firmed up. He came to think that even military asymmetries at these lower levels were not too worrying because decision-makers could expect to be deterred from using them by the fear of escalation at the nuclear level. This argument was first set out in his book The Illogic of American Nuclear Strategy (1984).

**********

Given the competitive pressures of anarchy – a factor that ran through Jervis’s security regime thinking – it is not surprising that his ambitions for security cooperation were constrained during the Cold War. Following its collapse, interestingly, his interest grew in ‘security communities’ – a development showing that his thinking, like that of Herz and Butterfield before him, could not be branded by one label. Across his career his ideas embraced ‘fatalist’, ‘mitigator’, and ‘transcender’ logics of international security; such an open approach was related to his ambivalence as to how far the psychological can trump the competitive pressures associated with the anarchic structure of international politics.

In an article in 2002 he announced a significant rethink, focused on the trajectory of ‘the Community’ comprising the United States, the European Union, Japan, Australia, Canada, and New Zealand. Their trajectory, Jervis argued, represented a significant shift from a cardinal assumption in his earlier writings, namely the idea that there could be no escape from security competition under anarchy. He now wrote: ‘For most scholars, the fundamental cause of war is international anarchy, compounded by the security dilemma. These forces press hardest on the leading powers because while they may be able to guarantee the security of others, no one can provide this escape from the state of nature for them…what is most important is that the Community constitutes a proof by existence of the possibility of uncoerced peace without central authority…the Community poses a fundamental challenge to our understanding of world politics and our expectations of future possibilities’ [emphasis added].

Having powerfully argued in the 1980s that regimes always collapse under anarchy, he argued at the start of the new millennium that this wider Western/liberal security community ‘does not have within it the seeds of its own destruction’. Not surprisingly, his view on the embeddedness of the process of bonding within the ‘Community’ was strongly opposed by proponents of ‘structural’ and especially ‘offensive’ realism. Even for those drawn to Jervis’s argument, it left two lacunae: the different trajectories of Russia and China, and the nature of relations between the Community and the rest of the world, and notably these two excluded great powers.

Today, the puzzles thrown up by the concepts of anarchy, regimes, and community remain as central preoccupations in the discipline of IR; and the issues involved, as indicated in our introduction, are being played out militarily across the Russia-Ukraine border. How stabilizing is nuclear overkill? Does leverage in the end come down to the balance of boots on the ground? What is being misperceived by whom? Can cooperation grow out of the crisis? If so, how far can it go? And on and on. In the Cold War the cost of overestimating structural factors were the risks associated with fatalistic assumptions about what is achievable in international security, and particularly excessive military hedging in ways that an adversary was likely to misread as aggressive intent. In the post-Cold War era the cost of underestimating structural factors has been the risk that decision-makers (and academics) might be drawn into believing that efforts at mitigating or transcending security competition might have better prospects than realistically exist.

**********

As the words above indicate, we need not search far for evidence of the enduring relevance of Robert Jervis’s work: it focused on big questions, sophisticated theorising, and rich historical analyses of the enduring puzzles of international politics. We have several times mentioned his immediate relevance to the issues swirling around the Russia-Ukraine border. Shortly before this particular crisis became headline-catching, it had been the situation across the Taiwan Straits that was being touted as the site for the next major crisis and possible war involving great powers. At issue here are Beijing’s ambitions to incorporate Taiwan into the Chinese state, and the crucial matter of ‘power transition’ between the rise of China and the supposed waning power of the United States. Such a state of affairs is widely characterised in the discipline as a manifestation of ‘The Thucydides Trap’, recalling the much-quoted sentence of the great historian from Ancient Greece, and his famous words: ‘It was the rise of Athens and the fear that this instilled in Sparta that made war inevitable’.

Robert Jervis knew brute power matters, including the psychological factors involved in the problematics of sending and receiving signals. He argued that the nuclear revolution had made power transition by all-out war to be highly irrational: but he knew that ‘irrational’ is certainly not the same as impossible. This may be the case even if decision-makers of adversarial states want to avoid calamity. He knew this because his research on the security dilemma had shown him the frequency, power, and negative consequences of misperception. In his closing remarks in Perception and Misperception, nearly a half century ago, he warned: ‘I strongly suspect that decision-makers have not accurately assessed the costs of various kinds of misperceptions and would be wise to correct for the tendency to be excessively vigilant’.

Jervis’s death is a sad yet needed reminder to all of us in academic life that what we do matters, not only because of what we might contribute to the body of influential ideas about IR, but also because of the potential impact of our attitudes and behaviour in our working lives as individual human beings. In writing this appreciation, pointing to Robert Jervis’s many ideas and achievements, we hope in particular to encourage students and early career academics who may not be familiar with his body of work to engage with the rich legacy of a truly exceptional scholar.


Robert Jervis was born in New York City on 30 April 1940, and died on 9 December 2021. His academic positions included the University of California and Harvard University, and after 1980 Columbia University. In 1978 he began consulting for the CIA. Among his academic honours, he received the Grawemeyer Award for Ideas Improving World Order; he was elected to the American Philosophical Society and the National Academy of Sciences; and he served as the President of the American Political Science Association. A selective list of most influential books and articles include: The Logic of Images (1970); Perception and Misperception in International Politics (1976); ‘Cooperation under the Security Dilemma’ (World Politics 1978); ‘Security Regimes’ (International Organization 1982); The Illogic of American Nuclear Strategy (1984);  ‘From Balance to Concert: a Study of International Security Cooperation’ (World Politics 1985); ‘Realism, Game Theory and Cooperation’ (World Politics 1988); The Meaning of the Nuclear Revolution: Statecraft and the Prospect of Armageddon (1989); Realism, Neoliberalism and Cooperation: Understanding the Debate (International Security 1999); ‘Was the Cold War a Security Dilemma’ (Journal of Cold War Studies 2001); ‘Why Intelligence fails: Lessons from the Iranian Revolution and the Iraq War (2010); and How Statesmen Think: The Psychology of International Politics (2017).


Previous post Kashmir Seeks the Hope

Leave a Reply

Your email address will not be published. Required fields are marked *