In the fall of 1962, two of the authors of this study were in school; the other was serving as an officer on the USS Canberra. On October 22 President John Kennedy reported to the nation and the world the presence of Soviet missiles in Cuba; he ordered a blockade of that island and demanded that the Soviets withdraw the missiles. The Organization of American States (OAS) gave its belated sanction to the American action.

The Canberra was the command ship of the blockading force, but we all sensed the universal danger: a threat had been issued; a line had been drawn. How would the Soviet leadership respond? Would the two powers find their way back from the brink? Or would the world be plunged into nuclear darkness?

Almost twenty years later, in a potentially less ominous but ultimately more tragic crisis, Argentine troops seized the Falkland (Malvinas) Islands, and Britain threatened to remove "the invaders" by force unless they withdrew speedily. Punctuating the threat, British warships, with troops embarked, made their way south in the Atlantic. The United Nations Security Council, the Organization of American States, and Britain's European allies condemned one or the other of the parties. Another threat had been issued.

The threat to use military force is a terribly grim and awesome business, raising, in our minds, several rather fundamental questions. Is the threat justified? Are the interests that the threatener feels to be endangered worth the risk of war? And what of the target? Are its interests or its prestige so at risk that it cannot acquiesce? If it calls the threatener's bluff, can the threatener find a face-saving way back from the brink? Will the target provide the way, or will its response be intransigent, its recourse war?

Threats are serious business and deserve attention. Some must inevitably be issued, for nations are still sovereign, disputes still arise, and no universal force exists that can resolve them all. We may not approve of this ineluctability, but we must acknowledge it. Threats, in any case, are to be preferred to preemptive attacks. Some threats must still be issued, but others should not, unless one is indifferent to the loss of life. Some should not be issued because they have little chance of success. That is one of the unanticipated findings of our study. We asked no more complex a question than why some threats succeed and others fail. But in concluding that some of those failures should have been foreseen by those who issued the threats, we found another reason for our study, even if we cannot recall specifically discussing it when we began.

Ours was little more than the social scientists' curiosity. To be sure, we sensed that our work was also that of the policy analyst—the search for better policy, as well as better theoretical understanding of policy. But for some time we thought of the relevance of our study only with apprehension, because we did not like to think of ourselves as servants of power, devising formulas to produce "better" threats. Our self-pride then amazes us now. No study of threats, however much more sophisticated or complete than ours, could succeed in giving to a national leadership any more than some ideas as to "what mattered" or how to proceed, given certain circumstances. None of these ideas, moreover, would supplant the particular conditions that prevail in any threat environment, independent of how deftly one manages the crisis. That is to say, if one does not threaten the right target for the right reasons, it may not matter how well one does it. But more of this within.

Direct military threats made by one power, or a coalition of powers, have for some time been the subject of attention of scholars and policy analysts. However, the existing literature dealing with direct military threats is either purely theoretical in character or is based on the analysis of a relatively small number of cases, and often the specific threat is subsumed in a broader study of policymaking or crisis management. (This literature is the subject of chapter 1.) Such studies can, of course, be insightful, and, as we think we have shown, some of them are. But we felt that one could not accept any of the hypotheses thereby generated regarding the success or failure of threats unless all such hypotheses (and, indeed, some others, implicit in the relevant variables) were subjected very specifically to verification through systematic analysis of a statistically sufficient number of actual cases. That is what we have attempted herein. Our methodology is described briefly in chapter 2 and in appendix 1.

In chapter 3 we address (and, on the basis of our evidence, answer) the following questions: What are the general characteristics of the "typical" direct military threat? What types of threats succeed, under what circumstances? What sorts of threats lead to war? Did threats issued in the pre-nuclear past differ in outcomes from those of the nuclear present? In that present, have threats made by the United States differed substantially from those made by the Soviets, Chinese, or British? And finally, can we say anything of the long-term consequences of these threats?

The concluding chapter summarizes our findings, compares them to the conventional wisdom, and tests those findings by asking how well they predict the outcome of six additional historical cases. We also attempt to apply our findings to a theoretical futuristic scenario, altering the character of several variables, in order to suggest the significance of what we have done. As the study drew to a close, events in Poland and in the Falkland Islands caught the attention of the world, and from the perspective of mid-1982, we added some observations on these crises, as they also seemed logical and telling tests of some of the key findings of this study.

Artis Frances Allen drafted the "Yugoslav scenario," described briefly in the final chapter (and available at nominal cost to interested readers). Peter Howell wrote the first draft of chapter 1 and is conducting a more specific, separate analysis, which is limited to "hard" data (allowing equal internal analysis, utilizing population figures, economic data, numbers of manpower and military hardware, and the like), referred to in chapter 2 and appendix 1. Peter Karsten conducted the analysis of the coded data cards and the questionnaire (appendix 2) and wrote the first draft of the preface, chapters 2, 3, and 4, and the appendices. But all three of us conducted the research, engaged in the coding of each case history, and prepared the select bibliography; and each read and suggested revisions to each chapter.

We are indebted to several colleagues and friends but want particularly to acknowledge the stimulation and support provided by Joseph Coffey, former Director of the Center for Arms Control and International Security Studies at the University of Pittsburgh, who set this study in motion, persuaded us to conduct it, and offered friendly advice and support throughout. We want also to thank Raymond Hinnebush of the College of St. Catharine's, who, while associated with us at an early stage of this project, had a substantial hand in the shaping of our research design. We are also grateful to Charles Gochman, Constance Rae, and Phil Sidel of the University of Pittsburgh, Paul Schroeder of the University of Illinois, Stephen Kaplan of the Brookings Institution, Patrick Morgan of Washington State University, and Timothy McKeown and Michael Solomon of Carnegie-Mellon University for their useful advice and criticism offered at various stages of the project. Of course, no one whose help we are hereby acknowledging is to be held responsible for any of our errors of omission or commission herein.