War is increasingly a legitimate object of study in sociology, a discipline in which it used to be considered as a rather marginal subjectFootnote 1. Conversely, in disciplines in which war was a central of object of study from their inception, such as international relations (IR), researchers are increasingly turning towards more sociological perspectives, including in the study of war and organized violenceFootnote 2. In both cases, the main inspiration has often been a historical macro-sociology, in a Weberian vain, but in which a healthy dose of Marxism was often presentFootnote 3. More recently, micro-sociological approaches have also developed an interest in organized violence and armed combatFootnote 4.
In this context, On wars, by renowned historical sociologist Michael Mann, is an important book. Indeed, it brings the two abovementioned strands of research together. One the one hand, it draws on Weberian and Marxian traditions in sociology, both infusing the macro-sociological theoretical framework that Mann developed in his previous booksFootnote 5. In this book, the usefulness of this framework to the study of war is convincingly shown. On the other hand, the author also engages in cross-disciplinary debates with, for example, International Relations (IR) theory. It is in this context mostly the neorealist “school of IR” that Mann focuses upon. Other theoretical schools are mentioned but hardly debated in any systematic way and democratic peace theory is resolutely rejected. Moreover, the author does not oversee the significant contributions made by other disciplines, such as behavioral psychology, to the understanding of armed combat at the micro-level.
Michael Mann has (once again) here produced an impressive and ambitious work on a topic that was important in most of his previous books, but never as the main subject Footnote 6. In this review, I will first try to give a general overview of the book, then discuss the main arguments that are developed, and finally discuss the question of rationality of (and in) war that Mann grapples with throughout the book.
A historical sociology of war: a global overview of the book
For the reader looking for a parsimonious and mono-causal theory of war, Michael Mann’s book is not to be recommended. There is a unified analytical framework and many recurring themes, but much of the pleasure of reading On Wars lies in this erudite author’s ability to explore war in very different spatio-temporal contexts: the Roman Republic and Empire, Ancient and imperial China, medieval and modern Japan, European wars over the last millennium, South and Central America over the last seven hundred years and the contemporary conflicts in the wider Middle East, with a particular focus on the recent US interventions. He devolves a chapter to each one of these contexts. Transhistorical and transregional comparisons are often made but specificities and contingencies are never forsaken to the identification of regularities and commonalities. One could regret that the place of Africa is downplayed, and that the Indian subcontinent is hardly mentioned. Otherwise, the book nearly reads as a universal history of war.
While the author adopts a mostly macro-sociological approach, there are three chapters on the micro-sociological dynamics of war, delving into the experience of fighting for soldiers, irregular combatants and sometimes civilians. Michael Mann here offers a very useful overview of the existing literature but also defends his own perspective. He for example severely criticizes the main conclusions of S.L.A “Slam” Marshall’s influential studies of the US Army in the 1940s, 50s and 60s. In particular, Mann criticizes the notion that moral qualms about killing enemy soldiers would play a crucial role in hampering accurate firing in the context of close-range combat as claimed by many following MarshallFootnote 7. He also highlights the irrationality of war with regards to the extreme sufferance that it causes for those, the soldiers and civilians, that are generally neither involved nor consulted when political decisions on war and peace are made.
There are moreover three theoretical chapters at the beginning of the book and a chapter criticizing the literature on the decline of war. In the latter, Mann shows that there are no clear long-term historical trends in the frequency or intensity of war. He also sees no reason to think that the future will be more peaceful than the past. He however notes that wars of conquest, a historically frequent form of war, have virtually disappeared (with the possible exception of the Russian invasion of Ukraine). Finally there is a chapter on possible futures. In the latter, the author quite daringly makes policy proposals, mostly targeted at the American foreign policy establishment, concerning ongoing conflicts.
Like many other sociologistsFootnote 8, Mann sees war as a phenomenon that is contingent upon social context rather than linked to human nature. Although collective violence does not seem to have been absent from the Paleolithic, or from later societies of hunter-gatherers, it is really with the Neolithic that war becomes a recurrent phenomenonFootnote 9. Since sedentarization and agriculture allow for the extraction of an economic surplus, a class of warriors and rulers emerges during this period. The increased stratification of societies and the rise of war are hence linked. Until here, the argument could seem quite classically Marxian. In fact, throughout the book, the importance of the articulation between political institutions and class-structure remain central to the understanding of war.
The book’s main originality lies in its analytical framework. Indeed, Michael Mann here once again demonstrates the usefulness of his theory of power developed amongst others in the first volume of The Sources of Social Power Footnote 10. In the latter, Mann shows that, while we tend to think of “societies” as distinct spaces, they are the result of imperfectly overlapping and intersecting networks of power. These networks involve distinct types of power and forms of organization: ideological, economic, military and political (hence, he calls it the IEMP model in this first volume). In On Wars, this framework is put to good use. One would expect military power to be the quasi-exclusive focus, but Mann persuasively shows that economic, ideological and political power networks all are involved in the various paths to war. Although the sociologist uses the concept of state (and of international order) in a very broad sense, including for the “pristine states” of the late neolithic period, this analytical framework also allows underlining that each polity is shaped by a unique articulation of the four types of power-networks.
Mann’s framework proves particularly helpful in underscoring the role of militarism in paving the way to war. Militarism, in the sociologist’s understanding, combines “the dominance of military elites in society, the ideological exaltation of military virtues above those of peace and extensive and aggressive military preparedness” [303]. It implies that ideological, political, and sometimes economic networks, align with military priorities. It plays a central role in his analysis of war under the Roman Empire, Ancient China, Genghis Khan’s Mongol empire, Imperial Germany or to a lesser extent the precolonial Inca and Aztec empires. Militarism is the propensity to wage war “baked into” (a recurrent metaphor in the book) political institutions and cultures. In fact the whole book can be read as a scathing critique of militarism. Mann’s definition of militarism is however difficult to apply to contemporary liberal democracies by his own admission.
In On Wars, Michael Mann also refines some of his central concepts. For example, with regards to the now classical distinction between “despotic” and “infrastructural power”, previous publications gave the impression that infrastructural power is foremost linked to the bureaucratization of the modern stateFootnote 11. This bureaucratization allows for the state to penetrate deep into society and transform it. Infrastructural power seemed, if not distinctly modern, then at least as linked to what Sinisa Malesevic has called the long-term “cumulative bureaucratization” of coercionFootnote 12. Despotic power, on the other hand, is the seemingly timeless power to impose decisions by wielding force and violence. However, Mann in this book introduces the idea that polities with small bureaucracies can also leverage significant infrastructural power provided there is a mutually strengthening articulation between class-structure and political institutions. About the Roman Empire, he for example writes that it was “run by its militaristic class structure, defined by nobility, wealth and military service, whose combination of collective solidarity and hierarchy of rank conferred considerable infrastructural power” [68]. By manoeuvring large legionary armies, the Roman empire made up for the small size of its bureaucracies in transforming the societies over which it ruled.
A theory of the path dependency of war
It is difficult to sum up the book beyond the wider analytical framework that has just been outlined. Wars are each time the contingent result of the interplay between the different power-networks. These networks are themselves transformed in the process. There is however a recurrent pattern that characterize the multiple paths that can lead to war throughout time and space.
Let us first clarify that war is here both explanandum and explanans. It both is what needs to be explained and what explains the course of history. In Mann’s historical narrative, explanandum and explanans are linked: one of the strongest predictors of wars of conquest are previously victorious wars of conquest. War creates the conditions for further wars, especially on the side of the victorious party. In neo-institutionalist (and in Mann’s) terms, there is a “path-dependency” of war and militarism. The paths are never defined in deterministic or mono-causal terms, yet they create a predisposition to war.
Warring states (especially if they previously have been deemed “successful”) relapse into militaristic behaviour because of the way in which previous wars have shaped ideologies, political institutions and economies. Emotional predispositions are also created in the process: the quest for vengeance, honour and systematic “over-confidence” that war will be short and victorious become common motivations of armed conflict. When the authority of a ruler resides in his ability to redistribute land to his vassals, to allow his soldiers to take spoils of war, to shore up his domestic legitimacy as “Son of Heaven”, to accommodate the interests of the military aristocracy…each war entrenches militarism further and pushes for more conquests. The imperial conquests of China under the Qin and Qing dynasties, or of Imperial Rome, were the result of such processes in a context in which both China and Rome were able to mobilize more military resources than their adversaries. Mann often evokes the “tyranny of history” to describe the path-dependent nature of war. As he writes in the conclusion to the book: “[…] wars are historical sequences in which the experience of past generations lies heavily on the brains of the living” [515].
Interestingly, Michael Mann himself sometimes describes the abovementioned pattern in terms of “Marxian circular processes”. Crises (in this case military crises), and the ways they are overcome, create the structural conditions for further, and potentially more damaging, crises in the future. Successful conquerors are ultimately trapped by their own conquests, victims of a self-reinforcing mechanism that ultimately leads to defeat. All factors do however not conspire towards wars of conquest. The tendency to “coup-proof” political institutions, for example by setting up praetorian guards or manipulating struggles between military elites, has been a perennial feature of many kingdoms and empires. This is still the case today in the Middle East and other countries in which the role of the military is less “to fight and win the nation’s wars”, than to keep the rulers in place while enjoying the latter’s lavish generosity.
Michael Mann recognizes the role played by what realists in IR theory call “international anarchy”, the “security-dilemma” and strategic rationalization in wars between polities. He however convincingly shows that it is a very thin theory of war if one does not account for the social dynamics of the polities involved. The domestic/ international distinction, that realist scholars cherish, is resolutely contradicted by Mann’s analysis. That being said, the sociologist foremost discusses and criticizes the two most famous neorealists, Kenneth Waltz and John Mearsheimer. Many aspects of his analyses are however much more consistent with what another neorealist author, John Vasquez, has written about warFootnote 13. Vasquez has shown that many wars occur because political leaders imprison themselves in their own logic of escalation, originally meant to deter their adversary, and hence to avoid war. When deterrence fails, it is difficult to back down from previously proffered threats and the conflict is consequently likely to escalate. Wars occur because militarized interstate disputes (MIDs) spiral out of control. Decision-makers get trapped by the wider strategic interactions that they feed. Many of Michael Mann’s analyses go in the same direction.
The question of rationality
Michael Mann is highly critical of the realist hypothesis of the rationality of state-behaviour in matters of war and peace. While I agree with the critique, the postulate of rationality indeed raises more questions than it solves, the book’s discussion of rationality/ irrationality is not fully convincing from a neorealist point of view. One of the points of departure for Mann’s reflections on rationality are the numerous “vanishing kingdoms” of world history: the polities like the Kingdom of Naples, the state of Chu or the Duchy of Burgundy, that have disappeared from the world map. They have been destroyed. swallowed by neighbouring polities or integrated in wider polities, generally through war. If so many polities have disappeared because of war, Michael Mann reasons, can the hypothesis of rationality of war be maintained?
In fact, neorealist (or structural realist) scholars like Kenneth Waltz believe states are compelled by the structure of the international system to adopt rational decisions, that is to maximize their chances of survivalFootnote 14. If they do not, they risk disappearing. It is the survival of the most powerful but also of the most rational. It is therefore likely but not certain that states will adopt rational foreign policies. As long as “vanishing kingdoms” can be explained by irrationality or relative lack of power, their vanishing does not contradict neorealism. Even when the “vanishing kingdom” initiates the war that makes it vanish, this can in neorealist terms be explained as “high-risk/ high gain” behavior on the part of a state that sees its long-term chances of survival compromised and seeks to make a last (and retrospectively unsuccessful) attempt to prevent the inevitable. Moreover, since Mann criticizes the realist hypothesis in the context of a study of several thousand years of history, one can truly wonder whether Waltz’s conception of the international system really makes sense with regards to, for example, feudalism in Europe or JapanFootnote 15. In fact, to be fair, the more fundamental critiques that can be addressed to “rationality” as conceived of by neorealists and rational choice theorists are also pointed out by Michael Mann throughout the book: rationalities are multiple, preferences cannot be simply assumed, the nature of the actor is not given (individuals, groups, organizations)… However, by concluding that wars are mostly irrational, and hence by buying into the simple distinction between the rational and the irrational (more than into the distinction between multiple rationalities for example), Mann’s framework exposes itself to some of the same criticisms as the rational actor model that he critiques.
At several instances, Mann highlights that with hindsight war can only be rational for the few. This is however hardly something rational choice theorists and neorealists would disagree with. Their point is rather that as long as nobody knows in advance who the winners and losers will be, war might be explained as an attempt to be part of these “few” in a context of uncertainty in which states pursue competitive advantages and long-term cooperation is unlikely. Let us also not forget that the realist argument of the security-dilemma generally implies that wars occur in a context in which the available foreign policy options are already severely restricted by previous militarized interstate disputes (MIDs). At the outbreak of war, the parties might still consider peace to be a more rational option than war. However if they find launching an attack (and hence seizing the initiative of war) more rational than risking being targeted by a surprise attack, and if they think such a surprise attack is imminent, they might still opt for war instead of for peace, without this being irrational per se. Michael Mann sees very little rationality in war with the exception of defensive wars. The main concrete example he gives in this regard is the second world war on the side of the allies. Realists would however point out that in war both sides might credibly consider that they are engaged in a defensive war.
What Michael Mann calls the “tyranny of history” is to be discussed in relation to what realists consider to be the “tragedy” of international relations: state-leaders can end up doing rationally, in a context of uncertainty, and after several cycles of escalation have reduced their foreign policy options, what they initially or ideally wanted to avoidFootnote 16.
The reason the irrationality of war is not more frequently exposed, according to Mann, is that the victors have written history. On the same account, one could argue that the reason a historical sociologist can highlight the irrationality of war is that he knows how it all ended. It would indeed have been irrational to initiate a losing war had the attacking party known what Michael Mann, as a historian, knows retrospectively. Mann can “write away” the virtual histories that never came about but that at the time were still seen as possible, plausible or even probable because of uncertainty and lack of information. In this regard, Mann rarely analyses the wars or military interventions that nearly occurred, or were massively advocated for, but that never occurred. Was it rational for them not to occur? It is often impossible to tell. For example, some people still argue that Obama’s decision not to intervene massively against Bashar al Assad’s Syria after the chemical attack on Eastern Ghouta in August 2013 was a strategic error and an irrational US foreign-policy decision, given the reality of the Syrian regime at the time and today, and all the massacres that happened in between. Even after the fact it is sometimes difficult to know what the most “rational” option was or would have been for a given actor. One can only conjecture what history would have been if other decisions had been taken.
Michael Mann himself points out that in many cases, war was not a choice, nor an instrument, but rather “the name of the game”, the way in which militaristic societies, from the Roman empire to feudal Japan, functioned. War and militarism were “baked” into societies and cultures, and hence could not be chosen away. This is one more reason for which the question of rationality/ irrationality, when raised in explicitly instrumental terms, has a limited analytical reach in the context of Mann’s long-term historical analysis. It does not fully do justice to the quality of the book. The question, from Michael Mann’s partly neo-Weberian point of view should maybe not be whether war is irrational or rational but rather what type of rationalities it serves and for whom. To put it differently, irrationality can hardly be an explanation since it basically means that we are unable to understand and interpret the thought-processes (be they value-based, instrumental, rooted in habit…) that led to a given decision.
Michael Mann does not need the concepts of rationality/ irrationality for his analyses to be convincing and compelling. It is as if rationality is here used to criticize war and militarism on seemingly “scientific” grounds, while the author’s condemnation of war and militarism is fundamentally moral, normative and political (in the noblest sense of the word). The US invasions of Iraq, the NATO war in Afghanistan, the one-sided US-support to Israel in the Israeli-Palestinian conflict, have all catastrophic effects in the “wider Middle East” and in the world. Is this not reason enough to criticize these policies? Do they additionally need to be irrational (for whom? In what regards?), given that such a claim – just as the neorealist postulate of the rationality of states in international relations - raises more questions than it answers?
Conclusions
This is undoubtedly a major work that will have a lasting impact on the historical sociology of war. It can be read as a “repository” of contemporary knowledge about war, but it also proposes original and thought-provoking analyses of the historical development of war over several millennia and continents. In this book Michael Mann once again shows his erudition and encyclopaedic knowledge. The breadth of the empirical/ historical chapters however sometimes raises questions about the limits of transhistorical comparisons. Is there for example a transhistorical unity to the concept of war? Are raids between hunter-gather communities during the Neolithic and the Second World War not fundamentally different phenomena? Can one really talk about MIDs when it comes to ancient China for which the very notion of an “international order” might be questioned? Such criticism are however inevitable given the ambitions of the book and in no way question its brilliance.
