David Welch's book is a welcome contribution to a field – the study of international relations – that all too often degenerates into scholasticism, or develops theories on far too slender an empirical base, or provokes heated debates on arcane questions. Here we have solid academic analysis that is also relevant to the world of policy.

Professor Welch's careful and lucid examination of the role played by considerations of justice in the outbreak of wars has many merits. One of the most obvious is that it is written without jargon, and therefore suggests that even complex issues can be presented in such a way that not only initiates understand them. Another merit is the empirical research he has done in a domain – history – that has been neglected far too often by model-builders on the one hand, and on the other by students who seem to believe that the world, and the study of its politics, began around 1945.

Even more important is Welch's criticism of the kind of “Realism” that has tended to dominate the field, and which rational choice analysis reinforces: a truncated and desiccated view of reality that singles out state interests and calculations of power, and leaves out all the passions and emotions that shape the definition of interests, the hierarchy of preferences, and the objectives of power.

It is difficult to read both the theoretical literature in political science on the causes of war and historians' case studies of the origins of particular wars without being struck by the difference in their respective evaluations of the importance of domestic political factors. Whereas historians devote considerable attention to these variables, most political scientists minimize their importance. Domestic political variables are not included in any of the leading theories of the causes of war; instead, they appear only in a number of isolated hypotheses and in some empirical studies that are generally atheoretical and noncumulative. This gap is troubling and suggests that political scientists and historians who study war have learned little from each other. A greater recognition of the role of domestic factors by political scientists would increase the explanatory power of their theories and provide more useful conceptual frameworks for the historical analysis of individual wars.

This study takes a first step toward bridging this gap by examining some of the disparate theoretical literature on domestic politics and war. It examines the relationship between national attributes and war behavior, the relative likelihood of democratic and non-democratic regimes going to war, Marxist and liberal theories regarding the impact of economic structure, the influence of nationalism and public opinion, and the scapegoat hypothesis. First, however, this article takes a closer look at the different treatment of domestic sources of war by political scientists and historians.

Are there really lessons of the past? The past is certainly a source of knowledge, our only source of knowledge given the flow of time, but, strictly speaking, it does not teach lessons. By lessons I mean maxims for attaining particular outcomes in the present or future: for example, Si vis pacem, para helium, or it is better for a prince to be feared than loved. Insofar as these maxims seem to offer guidance for policymakers, they are usually psychological, not historical. They supposedly summarize human traits that persist regardless of changing historical contexts. Other examples might include the notion that appeasement encourages aggression or that “military decision makers will tend to overestimate the feasibility of an operational plan if a realistic assessment would require forsaking fundamental beliefs or values.” Identification of such allegedly constant traits was the goal of philosophical history and may have seemed an appropriate program for historians during the Enlightenment and their policy-studies heirs. Subsequent historians, however, have usually sought to describe changing societal contexts or outcomes.

Explanations of the origins of World War II tend to emphasize either deliberate, if failed, choices or inexorable processes. The first view indicts Adolf Hitler's aggrandizing choices and preference for violence, and questions the judgment and strategy of the appeasers, personified, correctly or not, by Prime Minister Neville Chamberlain. The second view broadens the focus, pointing to secular changes in relative power between states; to the relation between states' commitments and their ability to uphold and protect them; and to domestic, economic, and cultural dynamics that individually, or in combination, predisposed the situation to conflict. Attention to both dimensions is necessary to appreciate Britain's strategy as the central axis of diplomacy and rivalry with Germany in the 1930s and to distill the “lessons” of the origins of the war.

In the 1930s, Britain took over the mantle of maintaining the status quo vis-à-vis Germany. Chamberlain and his associates faced the classic issue of judging the nature of its adversary's ambitions. Morgenthau captured the dilemma for Britain in this period: “While it would be fatal to counter imperialistic designs with measures appropriate to a policy of the status quo, it would be only a little less risky to deal with a policy [of an adversary] seeking adjustments within the status quo as though it were imperialistic.”

History is the study of events that have happened only once; political science is the effort to generalize about them. These caricatures sometimes seem an apt description of mutual reactions when members of the two professions discuss the origins and prevention of major wars. It might be amusing were it not that the next major war could be the last. Nuclear war is too serious to leave to either historians or political scientists alone.

As Waltz argues below, conflict may be endemic in human behavior, but war has its origins in social organization. Nonetheless, general theories of the causes of war can be misleading. “It is assumed, for instance, that there is a class of events involving human behavior that can be legitimately subsumed under a single term ‘war.’ True, the events have a common observable factor — organized violence perpetrated by groups of people upon each other. But that is near the extent of the commonality.” This volume does not search for a common set of causes of all violence from tribal vendettas to world wars. Instead, it focuses on the upper end of the scale.

Since the development of the modern state system in Europe some four centuries ago, there have been ten general wars involving a majority of the major powers and a high level of battle deaths. Some of these wars stand out in terms of their consequences for the hierarchy and structure of the system of states.

Among the major wars of modern European history, the Thirty Years' War stands out not only for its duration but also for its striking impact on the international system in which it took place. Before 1618, the Spanish Habsburgs were the central power in a Europe where religious differences were crucial. The war and its Franco-Spanish extension ended in 1659. By that time, France and other nations had increased their power, and religion played a much less important role in defining alliances. Moreover, Europe's center had moved east, as Russia, Prussia, and the Austrian Habsburgs became more powerful.

The war's origins are well known. Conflict in the Holy Roman Empire, especially in the Habsburg lands, over religion and over the power of the emperor provoked a civil war in Bohemia in 1618. The Bohemian war both resurrected and created a network of alliances which caused the conflict to continue into the 1620s. The opportunities offered by the disruption in Germany led the Danes to invade in 1625, and the Swedes and French to intervene in the 1630s, which continued the war by bringing in fresh combatants. The result was a conflict that could not be controlled by the Bohemians and the emperor, who had begun it. They were not allowed to extricate themselves until they had completely exhausted themselves and everyone else; as fitting compensation, the Bohemians and the emperor were important losers.

War has so many causes—in part because there are so many kinds of wars—and misperception has so many effects—again in part because there are so many kinds of misperceptions—that it is not possible to draw any definitive conclusions about the impact of misperception on war. But we can address some conceptual and methodological problems, note several patterns, and try to see how misperceptions might lead to World War III. In this article, I use the term misperception broadly, to include inaccurate inferences, miscalculations of consequences, and misjudgments about how others will react to one's policies.

Although war can occur even when both sides see each other accurately, misperception often plays a large role. Particularly interesting are judgments and misjudgments of another state's intentions. Both overestimates and underestimates of hostility have led to war in the past, and much of the current debate about policy toward the Soviet Union revolves around different judgments about how that country would respond to American policies that were either firm or conciliatory. Since statesmen know that a war between the United States and the Soviet Union would be incredibly destructive, however, it is hard to see how errors of judgment, even errors like those that have led to past wars, could have the same effect today. But perceptual dynamics could cause statesmen to see policies as safe when they actually were very dangerous or, in the final stages of deep conflict, to see war as inevitable and therefore to see striking first as the only way to limit destruction.