|
|
|
Email this citationCIAO DATE: 10/00
Contingency, Catalysts and System Change
Richard Ned Lebow
April 2000
"Very few things happen at the right time, and others do not happen at all."
— Herodotus (ca. 450 B.C.E.)
Wars, revolutions and depressions change the world and the way in which we think about it. World War I was a seminal event in both respects. It ushered in a profound transformation of the international system, and is described by many historians as the crucible in which the twentieth century was shaped. It has also been a critical case for the generation and testing of theories about conflict and international relations more generally.
Many historians contend that World War I — or something like it — would have been very hard, if not impossible. The distinguished British historian, F. H. Hinsley, insisted that “If the Sarajevo crisis had not precipitated a particular great war, some other crisis would have precipitated a great war at no distant date.” 1 Neorealists and power transition theorists make similar claims, albeit for different reasons. 2 I do not doubt that many, perhaps most, of the causes of war in 1914 that historians and political scientists have identified created a conflict-prone environment. But underlying causes, no matter how numerous or deep-seated, do make an event inevitable. Their consequences may depend on fortuitous coincidences in timing and on the presence of catalysts that are independent of any of the underlying causes.
World War I was “overdetermined” and highly contingent. It was contingent in both its underlying and immediate causes. Historians have proposed a variety of underlying causes for World War I, including social Darwinism, nationalism, the alliance structure and shifts in the balance of power. I argue that what made Europe ripe for war was not the multitude of alleged causes, but rather the interactions among them. World War I is best understood as a non-linear confluence of three largely independent chains of causation. These chains produced independent but more or less simultaneous gestalt shifts in Vienna and Berlin, and a slightly earlier one in Russia. Had the timing of the Austrian and German shifts been off by as little as two years, Austrian leaders would not have felt so intent on destroying Serbia or German leaders would not have been so willing to encourage them to do so. For this reason alone, World War I was overdetermined and highly contingent.
Theoretical explanations for war take catalysts for granted. If the right underlying conditions are present, some incident will sooner or later set armies on the march the way the twin assassinations at Sarajevo did in 1914. But Sarajevo was not just any provocation; it met a diverse set of political and psychological requirements that were essential for Austrian and German leaders to risk war. It is possible, but extremely unlikely, that some other provocation would have met these conditions, or that some other combination of great powers would have started a war for different reasons. In the absence of a catalyst, several more years of peace could have altered the strategic and domestic contexts of the great powers to have made war less likely. There was a two year window when the leaders of at least two great powers thought their national or dynastic interests were better served by war than peace. Social scientists often assume that major social and political developments are specific instances of strong, or even weak, regularities in social behavior. But these developments may be the result of accidental conjunctures. Conversely, events that seem highly likely may never happen. The concatenation of particular leaders with particular contexts, and of particular events with other events is always a matter of chance, never of necessity. 3
My findings have important implications of my findings for the study of international system change – by which I mean a change in the polarity of the system or the rules by which it operates. They suggest that system transformations — and many other kinds of international events — are unpredictable because their underlying causes do nothing more than create the possibility of change. Actual change depends upon contingency, catalysts and actors. Neither contingency nor catalysts have been analyzed systematically by social scientists, and I offer some thoughts about how this might be done. There is a large literature on actors, most of it based on the premise that they are instrumentally rational. A striking finding of the World War I case, and of the two other system transformations of the twentieth century — World War II and the end of the Cold War — is the extent to which the behavior that brought about these transformations was based on extreme miscalculations of its likely consequences. Such behavior may not be the norm in everyday foreign policy decisionmaking, but it may be characteristic of the decisions that unwittingly usher in system transformations.
The first two stages of my inquiry make use of counterfactual thought experiments. 4 Counterfactuals are past conditionals, or more colloquially, “what if” statements about the past. They alter some aspect of the past (e.g., doing away with a person or event, changing a critical decision or outcome, inserting an event or development that never happened or making it take place sooner or later than it did), to set the stage for a “what might have been” argument. I use only “minimal rewrite” counterfactuals. They entail small, plausible changes in reality that do not violate our understanding of what was technologically, culturally, temporally or otherwise possible. 5 A world in which Archduke Franz Ferdinand and his wife returned alive from their visit to Sarajevo is an example. Princip’s accomplice missed the royals en route to city hall, and Princip was lamenting his failure in a bar when the touring car carrying Franz Ferdinand and his wife came to a stop in front to allow the cars at the head of the procession to back up because they had made a wrong turn. With only a minimal rewrite of history — the procession stays on the planned route — the assassination might have been averted. Such a rewrite does not strain our understanding of the world because most twentieth century royal processions follow their intended routes. The archduke’s was an exception.
Case Selection
Case studies are often described as ill-suited for testing propositions and theories. One confirming or infirming case — even if the latter meets the criteria of a critical case — permits no definitive conclusions. Unless the case can be shown to be representative of the phenomenon under study, there is always the possibility that it is an outlier. But the laws of statistics are not the sole criterion or justification for case selection. Interesting cases, especially anomalous ones, are a primary source of theory generation. Careful analysis of an infirming case can suggest reasons why the outcome was contrary to predictions and reveal more general problems with a theory. Infirming cases can also improve a theory. By demonstrating why it fails in a single case, a researcher may identity the scope conditions of the theory and the variables or processes it must take into account to expand the domain in which it is applicable. 6
I do not use the 1914 case to evaluate a proposition or theory, but to critique a more general approach to understanding social phenomena based on the determining role of so-called structures. — e.g., system polarity, balance of power, alliances, regime types, I argue that structural theories cannot adequately account for World War I, and identify the reasons for this failure and the additional processes these theories must take into account to offer a better account of international politics. I do not claim that the 1914 case falsifies structural approaches; approaches, like paradigms, cannot be falsified. Rather, I contend that structural approaches, while valuable, provide only part of the analytical purchase needed to understand key foreign policy decisions, especially those responsible for system transformations.
There are two ways to generalize from a single case. The first strategy is to add additional cases in the hope of making the finding statistically significant. In this connection, I discuss two additional cases of system transformations: World War II and the end of the Cold War. I review the relevant literature on these cases which suggests the presence of the same phenomena that confound structural explanations in 1914. Three cases do not represent a large “n,” but they do constitute the universe of twentieth century system transformations and thereby lend a certain weight to my findings. The second strategy is conceptual. A researcher can argue that a problem, process or phenomenon that confounds a theory in a single case is prima facie universal, or so widespread, to render the theory inadequate or invalid. I make two such claims: that social systems are open and can be transformed through their internal operation as well as through external stimuli, and that all wars and system transformations require independent catalysts.
Statistical logic is not the only criterion of case selection. Not all cases are equal. Some have greater visibility and impact because of their real world or theoretical consequences. World War I is non pareil in both respects. Many historians contend that it was the crucible in which the twentieth century was formed. Its origins and consequences are also the basis for many of our major theories in domains as diverse as political psychology, war and peace, democratization and state structure. If World War I can be shown to be highly contingent, with the corollary that a very different twentieth century was possible — in its absence, imagine the gradual evolution of Wilhelminian Germany toward a constitutional monarchy, no Russian revolution, or at least a non-communist Russia, no World II and no Holocaust — then so too are the social realities that structural theories not only attempt to explain but make appear foreordained.
Underlying Causes of War
To use counterfactual experiments to explore alternative worlds, it helps to have as a starting point a generally accepted interpretation of why the world we live in has come about. Historians rarely agree about causes, but even for a contentious profession the degree of controversy surrounding the origins of World War I is extreme. Scholars disagree about the causes of war and the appropriate level of analysis at which to search for them. They also differ in their judgments of which state or states were most responsible for the war and the reasons why their leaders acted as they did. If the nature of the explanations and limitations of evidence make it difficult to discriminate among competing interpretations, limitations of space make it impossible to address all, or even most, of these interpretations.
Fortunately, something of a dominant interpretation has emerged in recent years. Historians associated this interpretation disagree on specific points (e.g., Chancellor Theobald von Bethmann-Hollweg’s motives for risking war, the relative responsibility of Austria and Germany), but agree that both powers set in motion the chain of events that led to war; Austria-Hungary exploited the assassination of Franz Ferdinand as the pretext for war with Serbia, and Germany encouraged — even pushed — Austria toward decisive action. 7 Students of Austria-Hungary argue that its leaders acted for a combination of closely related foreign and domestic concerns. Since the publication of the Riezler Diaries, a near consensus has emerged among German historians that von Bethmann-Hollweg did not seek to provoke a European war but recognized that an Austrian conflict with Serbia would be difficult to contain. He was willing to run the risk of a continental war in the belief that such a war was sooner or later inevitable and that Germany’s chance of winning it declined with every year that passed. 8 Given the primary responsibility of Austria and Germany for initiating the chain of events that led to war, I will focus on their decisions, and secondarily on Russia’s, but in full recognition that it is only part of the story.
Counterfactual experiments are wonderful means of exploring relationships among hypothesized explanations. Even a superficial counterfactual examination of underlying causes of World War I reveals that many theorized causes are tightly coupled. Counterfactual experiments also help us probe contingency. The most straight-forward way to do this is would be to use “minimal rewrite” counterfactuals to remove putative causes of an outcome. To apply minimal rewrites at levels of analysis other than leaders it would be necessary to go back to a time and place where the structures, ideas or institutions in question were more malleable or still in the process of taking shape The fewer counterfactuals necessary to remove the causes, the more counterfactuals that can accomplish this end, and the more proximate they are to the outcome, the more contingent it is. Limitations of space preclude such an exercise, and I employ a different strategy to address the problem of contingency. I look at the extent to which the underlying causes of war can be understood as a confluence, and use minimal rewrite counterfactuals to examine the implications of this causal metaphor for contingency.
A confluence envisages a multiple stream of independent causes that come together to produce an outcome. A house goes up in smoke. Investigation reveals that the fire spread from a lighted candle that was left unattended on a window sill. The window was not completely sealed, and a draft blew one of curtains close enough to the flame for it to catch fire. The smoke alarm, connected to the house security system, did not function because its battery was dead, and the fire department failed to receive the timely warning that might have permitted it to save the dwelling. What caused the house to burn down? The insurance investigator pointed to the candle, but it would not have been lit or placed on the window sill if it had not been the holiday season and had its owners not been following a neighborhood custom. If the window had not been warped, or the insulation around it had provided a better seal, the candle would not have started a fire. If the owners had been home, or if smoke alarm had a charged battery, the house would not have burned down. No single factor was responsible for this disaster; it took a combination of them interacting in a particular way. 9
An outcome that requires the confluence of many independent causes, but could be prevented by removing any one of them with a minimal rewrite — like the fire in the house — is highly contingent. An equifinal outcome to which multiple pathways lead, any one of which could bring it about requires multiple interventions to prevent. Its contingency would depend on how many minimal rewrites were necessary to halt or deflect each possible pathway. Some outcomes might be so highly redundant as to make them all but inevitable. Sooner or later, we will all die no matter how careful our diet, how much we exercise or how many diseases modern medicine is able to prevent or cure.
World War I is probably best understood as a non-linear confluence in which multiple, inter-related “causes” had unanticipated interactions and unpredictable consequences. Three causal chains were critically important. First and foremost was Germany’s security dilemma, caused by the prospect of a two-front war in which the general staff believed it would soon no longer be possible to defeat one adversary at a time. The second causal chain consisted of all the Balkan developments that threatened the external security and internal stability of Austria-Hungary, and encouraged influential opinion in Vienna to consider war with Serbia as a possible solution to these threats. The third chain centered on St. Petersburg, and was itself a confluence of external setbacks (e.g., defeat in Russo-Japanese War of 1904-05, humiliation in the Bosnian Annexation crisis of 1909) and internal weaknesses (e.g., the revolution of 1905, growing alienation of the middle classes, rise of a powerful revolutionary movement) that made Russian leaders fearful of the foreign and domestic costs of another foreign policy defeat.
Each of these three chains of causation was contingent; they were the result of decisions, bad and generally avoidable ones, that had the unintended consequence of dividing Europe into two armed and hostile camps. If Bismarck had been able to dissuade Wilhelm I from annexing Alsace-Lorraine, there might have been no enduring Franco-German rivalry. If Bismarck’s successors had managed relations with Russia better, St. Petersburg would have had no incentive to ally with France. The Anglo-French Entente might have been prevented by a kaiser who know how to keep his mouth shut and recognized the unnecessary expense and counterproductive nature of a naval race with England.
The point of no-return in Austro-Serb relations was the Empire’s annexation of Bosnia and Herzogovina in 1908. Austrian chief-of-staff, Conrad von Hötzendorff, had been pushing for annexation for some time, and convinced Alois Lexa von Aehrenthal, who became foreign minister in 1906, to do this as part of a new, assertive policy in the Balkans. Aehrenthal’s poorly conceived initiative provoked a war-threatening crisis, humiliated Russia and deeply embittered the Serbs. The annexation crisis destroyed a decade of Austro-Russian cooperation and put the two empires on a collision course. With more far-sighted foreign ministers in Vienna and St. Petersburg, this clash could have been avoided and Austro-Russian rivalry managed more effectively. In this environment, Russia would have been more restrained, and probably would not have violated the tacit agreement between the empires that neither would support dissident groups within the other’s empire. If St. Petersburg had not encouraged anti-Habsburg factions in Belgrade or stoked the fires of Rumanian nationalism in Transylvania, the Austrian foreign office would have been much less threatened by the likely defection of Rumania from its secret alliance with Austria. In the long run, Slavic nationalism would almost certainly have asserted itself, but that threat could have been managed for some time; as late as 1914 there were relatively few voices for independence within the Habsburg empire. The division of Europe into two militarily powerful but insecure alliance systems contributed to the outbreak of war, but did not make it inevitable.
Even more important than these chains of causation was the synergistic interaction among them. Two features of non-linear systems came into play here. The first is that the effect of one variable (or cause) often depends on the state of another variable (or cause). In such cases, the consequences of either cannot be predicted or understood independently. 10 This phenomenon is well-illustrated by the relationship between Russian armaments and railway construction and the Schlieffen Plan. German generals planned to solve the problem of a two-front war by committing most of their forces to a invasion of France in the expectation that they could occupy Paris before Russian forces penetrated too deeply into eastern Prussia. The general staff worried that Russian mobilization and railway reforms would render the Schlieffen plan obsolete, and this fear shaped the response of Germany’s political leaders to Austria’s request for support in its confrontation with Serbia. They reasoned that it was better to fight a war while there was still a chance to win. France had underwritten the expense of the Russian railways in the hope of restraining Germany; French leaders reasoned that Berlin would become more cautious in proportion to its fear of the consequences of a two-front war. The French strategy of deterrence actually encouraged the German invasion it had been intended to prevent. Something similar occurred during the July crisis when Russia mobilized in the hope that it might restrain Germany, not prove a casus belli . In each instance, the intensity and effect of particular actions or policies were determined by other actions and policies. In their absence, the leaders in question might have made very different choices, or the same choices might have had different consequences.
Game theorists who model strategic interactions have long recognized the need for actors to share a common framework, or at the very least to agree on the kind of game they are playing. They assume that actors use a Bayesian process to update estimates of one another's preferences and that such learning will allow them to establish a common framework. 11 In practice, new information is commonly assimilated to existing frameworks, and actors can continue to communicate for prolonged periods of time without realizing that they are playing different “games.” “Signals” may be missed, or their intended import only grasped after it is too late to respond appropriately. Frameworks also change in the course of interactions, and these changes can have profound consequences for behavior. The ability of actors to transform themselves and their understandings of their strategic interactions in the course of those interactions is a second fundamental characteristic of non-linear, open systems, and central to understanding the events of July and August 1914.
Key Austrian and German leaders underwent independent gestalt shifts in 1913-1914 that prompted dramatic reversals in their foreign policy preferences. Moltke and Bethmann-Hollweg had been troubled by Russia’s seemingly growing military and mobilization capability for some time, but only 1914 — and probably as a result of the assassination of Franz Ferdinand — did the chancellor’s concern reach the level where he was willing to do something he had consistently rejected in the past: risk war in the hope of achieving a diplomatic triumph that might break up the Franco-Russe. Austrian leaders worried about a Balkan League directed against them that would constitute an external threat and fan the fissiparous tendencies within their empire. Conrad wanted to exploit the assassination of Franz Ferdinand as a pretext to attack Serbia, Berchtold and Franz Josef saw not alternatives and Berlin was now willing to offer its support. Serbs and Russians knew nothing about these gestalt shifts, and behaved in ways that exacerbated Austro-German insecurities and provoked a war that neither country wanted. Wilhelm, Bethmann-Hollweg and Foreign Secretary, Gottlieb von Jagow were in turn victims of a Russian gestalt shift. They deluded themselves into thinking that they might repeat the success of 1909 and compel Russia to remain on the sidelines of an Austro-Serb war. But that humiliation had led to a commitment not to back down again, something the Russians believed would undermine their status as a great power. This commitment was strengthened by cabinet shifts in 1914. These several gestalt shifts entirely changed the nature and outcome of great power interactions.
One of the most remarkable features of 1914 was the coincidental timing of Germany and Austria’s security problems and gestalt shifts. Although Russia was a common threat, each ally’s security problems had largely independent causes, and there was no particular reason why they should have become acute at the same time. Germany’s security dilemma was the result of its geographic position and prior policies that had encouraged its two most powerful neighbors to ally against it. Russia’s improved military and mobilization capability, the development that drove Germany to brinkmanship in 1914, was the result of Russian industrialization and access to French capital markets. German willingness to risk war was also the result of the perceived decline in capability and will of Austria-Hungary, Germany’s principal military ally. German political and military authorities worried that failure to support Austria in 1914 would accelerate its decline and leave Germany at the mercy of Russia and France.
Austria-Hungary’s insecurity was the consequence of a precipitous decline in Ottoman power. That decline had many internal causes, but it was dramatically hastened by the Italian occupation of Tripoli in September 1911 and the war this unexpected triggered with the Ottoman Empire. The war provided the opportunity for Serbia, Bulgaria and Greece to take up arms, and to almost everyone’s surprise, they succeeded in all but expelling Turkey from Europe. Serbia doubled its population and territory, and backed by Russia, sought to organize an anti-Austrian alliance in the Balkans and to incite national unrest within the Dual Monarchy.
German leaders did not feel so threatened before 1914; German chancellors rejected military demands for war in 1905 and 1912, and supported diplomatic resolutions to the 1912 and 1913 crises that threatened war between Austria and Serbia and Austria and Montenegro. If Italy had not occupied Tripoli — and there had been considerable opposition to this ill-conceived venture — no Balkan Wars in would have started in its wake, and Serbia would not have constituted a threat, if it did at all, until some later date. Alternatively, if the Balkan events that Austria-Hungary found so threatening had occurred a few years earlier, the German kaiser and chancellor would probably not have encouraged Austria to draw its sword. Timing was everything in 1914, and that timing, so to speak, was fortuitous, if that is the right word. For this reason alone, World War I was highly contingent
Immediate Causes of War
Wars, like fires, need catalysts. Most structural theories assume an unproblematic relationship between underlying and immediate causes. If the underlying causes are present, an appropriate catalyst will come along — or it be manufactured by leaders. 12 In February 1965, in the aftermath of a Vietcong attack on the American advisors’ barracks at Pleiku, National Security Advisor McGeorge Bundy wrote a memorandum to President Lyndon Johnson urging the sustained bombing of North Vietnam in response. Bundy later acknowledged that he conceived of Pleikus as “streetcars.” He could count on repeated Viet Cong attacks against South Vietnamese forces or their American advisors to provide him with the pretext he needed at the opportune moment to sell escalation to the president. 13
Pretexts do not always resemble streetcars. They may be infrequent, inappropriate, or may fail to materialize, and without a catalyst, the predicted or intended behavior may not occur. In a matter of months or years, the underlying conditions may evolve as to make war less likely even if an otherwise appropriate catalyst ultimately comes along. The “window of opportunity” for war may be temporally narrow or broad, depending on the nature and rate of underlying changes leadership, security conceptions domestic and international structure. War, like many other kind of events, requires a conjunction of underlying pressures and appropriate catalysts.
One of the most common metaphors used to describe Europe in 1914 is that of dry kindling waiting for a spark to set it alight. As sparks are frequent in acute international conflicts, this metaphor is well-chosen by structuralists because it emphasizes the underlying causes of war. Metaphors are no substitute for careful analysis, and historians and political scientists need to develop more precise conceptions of catalysts. We can begin by asking what about Sarajevo made it a successful catalyst, and what other provocations or events might have served the same end? Answers to these questions will provide a second perspective from which to estimate the contingency of war in 1914.
Joachim Remak insists that “Sarajevo was more than an excuse for war, it was one of its major causes.” 14 Although he does not elaborate on his claim, many reasons can be adduced in support. Arguably the most important was the assassination itself and the political challenge it constituted for Austria-Hungary. In June 1914, Berchtold, with Franz Josef’s support, began a diplomatic offensive to arrest the decline of the Empire’s position in the Balkans, and to frustrate Entente efforts to build a new Balkan League. There was no talk of war. 15 The assassination transformed the situation. Not only Conrad pushed for war, but other officials in the foreign office and military argued that failure to respond forcibly to this provocation would undermine, if not destroy, the Empire’s standing as a great power and embolden its domestic and foreign enemies. 16 For Franz Josef there was an additional, personal dimension to Sarajevo; he was outraged by the assassination of a member of the royal family. Kaiser Wilhelm also grieved over the loss over Franz Ferdinand, whom he considered a friend, and had spent time with only two weeks before. 17 He wanted Austria to take action against Serbia to show that actions against legitimate rulers would not be tolerated. 18
Sarajevo shifted the balance of power in Vienna. Franz Ferdinand’s views on defense matters were almost as important as the emperor. His influence derived from his official status as successor to the throne ( Thronfolger), from his interest and knowledge about military affairs and the extensive network of contacts he had cultivated throughout the armed forces. His decidedly peaceful orientation evolved during the course of the Balkan wars. 19 The Thronfolger was intent on extending Austrian influence in the Balkans, but not at the risk of war with Russia. He warned that “A war between Austria and Russia would end either with the overthrow of the Romanovs or with the overthrow of the Habsburgs — or perhaps the overthrow of both.” 20 He cherished the unrealistic idea of monarchial unity in the form of some revival of the Holy Alliance, and had made a point of cultivating good relations with Nicolas II. On a more practical level, he took Russian military capability more seriously than either the war minister or chief-of-staff, and was convinced that war against Serbia would draw in Russia. He did not believe that the Austro-Hungarian army was ready for war, worried that Italy would defect from the Triple Alliance and that Germany would find some reason to stand aside More fundamentally, Franz Ferdinand opposed war because it would make it impossible for him to impose fundamental changes in the structure of the empire upon his accession to the throne. 21
Samuel R. Williamson, Jr. offers an intriguing counterfactual: If Governor General of Bosnia-Herzogovina, Oskar Potiorek, had been killed at Sarajevo instead of Franz Ferdinand, Vienna would have responded differently. 22 Like Conrad, Potiorek was a charter member of the war party and his death would have removed another supporter of military action from the scene. More importantly, Franz Ferdinand would have been influential in shaping Austria’s response. His opposition to war, combined with that of Hungarian prime minister, István Tisza, the senior voice against war in June and early July 1914, would have carried considerable weight because the two men were otherwise always at odds. Tisza was the great defender of Hungary, and Franz Ferdinand made no secret of his dislike of Tisza and Hungarians more generally. 23 With Franz Ferdinand and Tisza urging moderation, Berchtold, a weak personality, would also have pursued a cautious line, and Franz Josef, cross-pressured in 1914, would probably have sided with them instead of Conrad. If so, there would have been no Hoyos mission; Germany would have been consulted with a diplomatic end in mind. The channel for communication with Germany would not have been the hawks in the foreign office but Franz Ferdinand, who had close personal relations with Kaiser Wilhelm and had been used in the past to sound out Berlin’s intentions in the Balkans. Merely changing the victims of the terrorist attack in Sarajevo might have been enough to alter in a fundamental way Austria-Hungary’s response.
Sarajevo provided a necessary incentive and opportunity for Germany. Moltke had pushed for war almost from the moment he became chief of staff because he wanted to fight Russia and France while Germany still had a chance of victory. Although the German general staff had a low regard for the military prowess of their Austrian ally, they were horrified at the prospect of an Austrian decline because it would leave Russia free to concentrate all of its forces against Germany in East Prussia. In Berlin, the assassination was perceived to threaten Austria’s standing as a great power if it exposed Austria’s lack of will to act like one. This additional consideration, when weighed in the balance along with the general concern for the deteriorating military balance, made Chancellor Bethmann-Holweg more receptive to Moltke’s pleas for action at the height of the crisis. The assassination may well have been the catalyst for Bethmann-Holweg’s gestalt shift described in the previous section. 24
Bethmann-Holweg was more prescient than most of his contemporaries in recognizing, as he put it, that a European war was likely to topple more thrones than it would prop up. He accordingly deemed the backing of the Social Democrats, the largest and best organized working class organization in Europe, the sine qua non of military action. Without it, the chancellor would not have taken his “leap into the dark.” Moltke knew this, and in February 1913 had discouraged Conrad from attacking Serbia on the grounds that the German people would not support a war that Austria provoked against a seemingly conciliatory adversary. 25
Sarajevo was a tailor-made provocation for Bethmann-Hollweg. The assassination aroused considerable sympathy for Austria throughout Europe, and not the least among the German working class. Although the Austrian ultimatum was widely regarded as heavy-handed by the politically sophisticated, German opinion perceived their country as a bystander to a Balkan conflict and the innocent target of Russian aggression. The chancellor played up this interpretation, and benefitted from the general fear and dislike of Russia by Social Democrats, who regarded the Czarist regime as barbaric because of its treatment of labor, dissident intellectuals and minorities. The result was the Burgfrieden of 4 August in which the Social Democrats voted for war credits.
Sarajevo created the necessary psychological environment for kaiser and chancellor to overcome their inhibitions about war. Admiral Tirpitz observed that “When the Emperor did not consider the peace to be threatened he liked to give full play to his reminiscences of famous ancestors, [but] in moments which he realized to be critical he proceeded with extraordinary caution.” 26 To his contemporaries, the chancellor came across as a fatalist, as a man who had a deep revulsion of war, but felt powerless to oppose the prevalent view that it was necessary. Kaiser and chancellor were caught on the horns of a dilemma: Germany’s vital interests seemed to require war, or a diplomatic triumph that would break up the Franco-Russe , and the latter could only be achieved at the risk of war, but they were not prepared to accept responsibility for starting such a confrontation. Until July 1914, they procrastinated. By deferring a decision that was too difficult for them to make, kaiser and chancellor preserved their psychological equilibrium. The July crises offered them a way out of their decisional dilemma.
When Hoyos met Wilhelm and Bethmann-Hollweg on 4 July in Potsdam to secure their support, he did not confront them with a choice between peace and war but only with a request to back Austria if Russia threatened intervention in support of Serbia. Kaiser and chancellor both expected Russia to back down as it had in 1909, and this was also the view of their ambassador in St. Petersburg. 27 They doubted that France would come to Russia’s assistance, or that Britain would intervene if the Balkan conflict provoked a continental war. Their support for Austria precipitated its ultimatum and declaration of war against Serbia, Russia’s subsequent mobilization against Austria, German ultimatums to Russia, Belgium and France and German mobilization, which was the equivalent of war. When kaiser and chancellor confronted Russian mobilization, or more accurately, premature and exaggerated reports of Russian mobilization that flooded through the channels of German military intelligence, they convinced themselves that they were only reacting to Russian initiatives, and that St. Petersburg, not they, bore the brunt of responsibility for the war they were about to unleash. 28
To recapitulate, the Sarajevo assassinations changed the political and psychological environment in Vienna and Berlin in six important ways, all of which were probably necessary for the decisions that led to war. First, they constituted a political challenge to which Austrian leaders believed they had to respond forcefully; anything less was expected to encourage further challenges by domestic and foreign enemies. Second, they shocked and offended Franz Josef and Kaiser Wilhelm and made both emperors more receptive to calls for decisive measures. Third, they changed the policymaking context in Vienna by removing the principal spokesman for peace. Fourth, they may have been the catalyst for Bethmann-Hollweg’s gestalt shift. Fifth, they made it possible for Bethmann-Hollweg to win the support of the socialists, without which he never would have risked war. Sixth, they created a psychological environment in which Wilhelm and Bethmann-Holweg could proceed in incremental steps toward war, convincing themselves at the outset that their actions were unlikely to provoke a European war, and at the end, that others were responsible for war.
A striking feature of the July crisis was the tremendous psychological difficulty German leaders had in making a decision for war. Given their unwillingness to accept responsibility for starting a great power war, it is certainly difficult to imagine how kaiser, chancellor, foreign office could have taken this step if they had been compelled to recognize their share of responsibility for it from the outset. If the archduke had not been assassinated, giving rise to this unusual opportunity, Germany might have reached the fateful year of 1917 still at peace with its neighbors. If so, its leaders might have discovered that their fears of a window of vulnerability were greatly exaggerated; their adversaries were constrained from attacking Germany for many of the same reasons that had prevented Germany from exploiting its military window of opportunity in the decade before 1914.
The double assassination was critical in its nature and timing, but it could easily have been avoided. Reports reached Vienna that Sarajevo was seething with discontent and a dangerous venue for a royal visit. Franz Ferdinand explored the possibility of postponing his trip and seems to have been encouraged by the emperor to do so. The High Command nevertheless decided to proceed with its great maneuvers, and the archduke, who was Inspector General of the armed forces, decided that he had no choice but to attend. Duchess Sophie had come to Bosnia with dark misgivings that something dreadful was about to happen to her husband. She was aware that Dr. Josip Sunari_, one of the leaders of the Sabor, the Bosnian parliament, had urged General Potiorek to cancel their impending visit because of the hostility of the local population to the regime. The evening before the assassination, Karl von Rumerskirch, the archduke’s chamberlain, urged him to cancel the next day’s visit to Sarajevo. General Potiorek’s aide-de-camp, Lt.-Col. Eric von Merizzi, interceded and convinced Franz Ferdinand to proceed because cancellation would be a rebuke to his superior. The next morning the archduke and his wife were met at the Sarajevo train station by General Potiorek and the Lord Mayor, and ushered into an open touring car to go to a nearby military camp for a quick inspection before going on to the city hall. The lead car in the procession was supposed to carry six specially trained security officers, but their chief departed with three local policeman instead. On Appel Quay, a long street with houses on one side and an embankment on the other, a young man in a black coat asked a policeman which car carried the archduke and then stepped out into the street to throw a grenade at it. Franz Ferdinand’s Czech driver saw the object coming and accelerated. The bomb fell on the folded roof, rolled off into the pavement and exploded under the rear wheel of the next car in the procession. The would-be assassin jumped over the embankment into the river. 29
Lt.-Col Merizzi and another officer, both in the car behind the archduke, were hit by bomb fragments and were rushed to a military hospital. Franz Ferdinand dismissed the attack as madness and insisted on preceding to the city hall. Following the ceremony there, the archduke asked General Potiorek if there were likely to be any more attacks. Potiorek advised taking a different route and skipping the planned visit to the museum. Other members of the archduke’s party urged him to leave Sarajevo immediately, but he insisted on visiting Col. von Merizzi in the military hospital and then going on to the museum. The cars drove up Appel Quay, this time at high speed, but the lead car turned by mistake into Franz Josef Street and the second car with the police guard followed. Franz Ferdinand’s driver, in the third car, was turning to follow when he was ordered to stop by General Potiorek, back up and continue down Appel Quay. At that moment, Princip, standing at the intersection, took a revolver out of his coat, and a nearby policeman reached out to grab his hand. An accomplice struck the policeman, and Princip fired twice at point blank range into the car containing Franz Ferdinand and Sophie. 30
Numerous minimal rewrites could prevent the assassinations: Princip might have obeyed the order to abort the assassination sent to him by the military conspirators in Belgrade, Austrian authorities in Bosnia might have taken security as seriously as they did the menu and music for the banquets they planned in the archduke’s honor, Franz Ferdinand might have canceled his trip in response to multiple warnings and his wife’s fears, he might have followed the advice of his advisors and left Sarajevo directly after the ceremony at city hall, or his cavalcade could have adhered to the planned route and raced down Appel Quay past Princip. None of these changes strain our understanding of the world because most royal processions do not stray from their intended routes and most security details would have rushed the archduke and his wife to safety at the first sign of violence.
Without the assassinations there would have been no war in the summer of 1914. Could some other country, or combination of countries, have found a reason to start a war? Britain was committed to the status quo and was consumed with its Irish problem. France coveted Alsace-Lorraine, but had been on the defensive in Europe since 1871, and perceived itself increasingly weaker militarily vis a vis Germany. The lynchpin of French security was the Franco-Russe, and France supported Russia in 1914 to preserve this alliance. France had also drawn closer to Britain, and relied on British military assistance in case of a war with Germany, but the French knew this would only be forthcoming if they were attacked by Germany. Italy pursued an aggressive colonial policy that led to war with the Ottoman Empire, and aspired to those parts of Austria-Hungary inhabited by Italian speakers. But before 1914, Italy was constrained by its alliance with Austria and Germany. The Ottoman Empire was everywhere on the defensive and not about to challenge any major power or provide it with a pretext to intervene in support of any of its neighbors. Russia was more or less recovered from its defeat in the Russo-Japanese War and intent on expanding its influence in the Balkans. But Russian leaders did not want war; they mobilized reluctantly in 1914, and in the hope that it would deter Austria from attacking Serbia. Serbia had long-term aspirations to acquire Bosnia-Herzogovina but its energies were fully consumed with overcoming the resistance of its newly acquired subjects in Macedonia. In 1910, German foreign minister Alfred von Kiderlen-Wächter rightly observed that “If we do not conjure up a war into being, no one else certainly will do so.” 31 .
A failed assassination attempt could have had beneficial consequences for Europe. Serbia’s diplomatic humiliation in 1909 encouraged the formation of secret societies aimed at undermining Austrian rule in Bosnia-Herzogovina. Apis was at the center of many of these conspiracies, and supplied arms and other assistance to the archduke’s assassins. Pa_i_ was hostile to the conspirators and knew of their preparations, but felt constrained to provide only veiled warnings to the Austrian ambassador. Ironically, neither Apis nor the other conspirators wanted war with Austria, and Apis did not expect the assassination to provoke one. Apis hoped to strengthen his hand vis a vis civilian authorities, and sought to call off the assassination when he became convinced that it would not buttress his authority. From the perspective of those who mattered in Belgrade, the war was an unintended and largely undesired consequence of an unwanted assassination. An abortive assassination attempt might have allowed Pa_i_ to rein in Apis, which he did in the summer of 1915. Apis was arrested a year later, tried and executed in 1917.
Historians who contend that a European war was inevitable root their argument in the deeper trends they see pushing the powers toward war. Would these trends have become more or less pronounced in the years after 1914? Structural determinists assume they would have intensified, although they offer no reasons why. We can conjure up scenarios of more acute Austro-Russian competition in the Balkans, British-German naval and economic competition and the breakup of the Triple Alliance due to Italy’s defection. With no more imagination, we can identify developments that would have muted underlying tensions and made war less likely.
The principal difference between the Balkan crises of 1908-09 and 1914 was Russian willingness to go to war in support of Serbia in 1914. Some historians maintain that Russia was ripe for revolution in 1914 and that World War I postponed the upheaval for three years. 32 If revolution had broken out in the absence of war, Russia might have been consumed by domestic turmoil for several years afterwards and not have been in any position to pursue an aggressive policy in the Balkans. This could also have happened to Austria. Franz Josef died in 1916 and was succeeded by Prince Karl. If Franz Ferdinand had lived, he would have ascended to the throne. Motivated by hatred of Hungarians and the Ausgleich of 1867, Franz Ferdinand would have sought to reduce the power of Hungary, and had considered several strategies toward this end including “trialism” and a looser form of federalism. The documents he had prepared for the transition indicate that he probably would have introduced universal suffrage in Hungary at the outset of his reign in the hope of increasing the power of minorities at the expense of the Magyars. This would have provoked a strong reaction from Budapest, and further attempts by Franz Ferdinand to undercut the Ausgleich would have raised the prospect of civil war. Vienna would not have been in any position to start a war with Serbia. 33
One of the principal causes of war in 1914 was the German military’s belief that war was inevitable and had to be fought before 1916 or 1917 when improved Russian mobilization and armaments made the Schlieffen Plan unworkable. The most obvious alternative, a direct onslaught on France across the Meuse and Moselle had little chance of success because of the mountainous terrain and French fortifications. Schlieffen contemplated such a campaign in 1894, but quickly gave the idea up as unrealistic. 34 Could Germany have conducted an offensive in the East? Russian railway and fortress construction made an Austro-German offensive in Poland difficult, but not impossible. It could not produce the kind of decisive victory Moltke sought. If the Germans broached the Narew River line, the Russians could withdraw with relative ease into their vast hinterland.
If offensives against Russia or France were unrealistic, the most sensible strategy was a defensive posture on both fronts; German generals knew that France and Russia planned to march against Germany at the outbreak of war. With no German invasion of Belgium, Britain would have remained neutral. The French army would have exhausted itself, as it began to do in 1914, in a series of unsuccessful and costly assaults against the strong German defensive position in Alsace. In the East, the Russian offensive into Prussia was blunted by the meager forces Germany had left in the region and forced to withdraw in great disorder. Reserve forces could have conducted a limited counter-offensive, after which Germany could have called for a restoration of peace on the basis of the status quo ante bellum . It seems improbable that there would have been much support in France or Russia for continuing the war after a series of disheartening defeats. Nor could these powers have resisted British, and perhaps, American, pressures to lay down their arms and accept a reasonable peace. Austria-Hungary would have been preserved, although the Russian empire might have succumbed to revolution. German preeminence on the continent would not only have been maintained but immeasurably strengthened.
Moltke was willing to tinker with details of the Schlieffen plan but reluctant to consider alternatives. Some members of the general staff doubted the likelihood of victory in 1914, but clung to the Schlieffen Plan because of their collective commitment to the offensive. 35 By 1917, according to the calculations of the general staff, Germany would have had no hope of waging successful, sequential offensives against France and Russia. If war had been avoided in 1914, the contradiction between Germany’s strategy and military reality would have become more pronounced. Moltke, or more likely, his successor, would have been compelled to abandon the Schlieffen Plan. As funds for additional troops were out of the question, and the concept and technology of the Blitzkrieg had not yet been developed, there were no viable alternatives to the defensive. The German general staff might have been compelled to adopt such a strategy, or variant of it. If so, they would no longer have had the same incentive to launch a preventive war or preempt in crisis, and to the extent that they became fearful of Russian military capabilities, they might even have become a force for preserving the peace.
I have asked if a European war could have been delayed in the absence of an appropriate catalyst. Could it have come sooner? I am inclined to discount the prospect of an earlier war. Austria considered and rejected going to war with Serbia in December 1912, April-May 1913 and October 1913. Between 1905 and 1914, German leaders spurned Moltke’s several pleas to exploit great power crises as pretexts for war. Austrian and Germans swords remained sheathed because political leaders in Berlin and Vienna saw war as politically and militarily risky and did not feel threatened enough to take these risks.
Contingency and Causation
The origins of World War I are best understood as a confluence of three largely independent chains of causation that came together in 1914. Their interaction has the characteristics of a complex, non-linear system. The value of important variables was not independent, but depended on the presence and value of other key variables. And they in turn depended on the changing understandings actors had of their strategic interactions.
A linear model that specified the presence of A (the set of variables associated with the German security dilemma), B (the set of variables relevant for Austria’s security dilemma) and C (Russian willingness to risk war to support Serbia) would only capture part of the strategic picture. The values of A, B and C were determined by gestalt shifts that took place in 1909 in Russia and in 1914 in Austria-Hungary and Germany. The presence of A + B + C prior to these gestalt shifts would not have produced a war. Nor, I have argued, would their coincidence have been likely to do so after 1916 or 1917 if the political alignment in the Balkans had changed, if Russia’s domestic situation had become more acute or if the Schlieffen Plan had been replaced by a more defense-oriented alternative. War required the coincidence of A + B + C after the gestalt shifts and before important underlying conditions changed to produce further shifts. The catalyst for the C gestalt shift was Russia’s perception of its humiliation in the 1908-09 Bosnian Annexation Crisis, and for the A and B shifts, the twin assassinations at Sarajevo.
World War I is not unique in its non-linearity. World War II, which brought about the next transformation of the international system, was also the product of a highly contingent set of conditions. Aggressive as Hitler was, it is more difficult to imagine Germany starting a war in a less fortuitous context. In the 1930s, France was divided internally, and Britain and France were loath to collaborate with the Soviet Union. Because his imperial policy in Africa had run afoul of Britain, Mussolini abandoned his opposition to German expansion and entered into an alliance with Hitler. In the Far East, Japan attacked China, and posed a serious security to the Soviet Union and the Western powers. Isolationism guaranteed that the United States, whose intervention had determined the outcome of World War I, was no longer a player in the European balance of power. Hitler could attack his enemies piecemeal, while counting on the support of Italy and the neutrality of the Soviet Union and the United States. The end of the Cold War — which brought about the third system transformation of the century — can also be described as the result of complex, path dependent, non-linear interactions. 36
International relations theory needs to consider multiple paths of causation and their possible interaction. Current theories of international relations almost invariably focus on one chain of causation. Power transition theory, for example, attempts to explain the outbreak of wars responsible for system transformations in terms of the changing power balance between hegemons and challengers. 37 Some power transition theorists, notably, Robert Gilpin, identify responses other than war for declining hegemons. 38 Power transition theories that acknowledge choice are indeterminate, and attempts to use them to account for the end of the Cold War elicit the criticism that they are being used to explain ex post facto what they did not and could not possibly have predicted. The indeterminacy problem arises because factors, independent of the power balance, shape actors’ responses to an actual or anticipated decline in their power. 39
In 1914 power was only part of the story; shifts in the political and military balance may have made German and Austrian leaders more willing to consider the use of force but were insufficient cause for them to draw their swords. Power transition theory – or any other monocausal explanation for system transformation – may be a useful analytical starting point, but it is unlikely to offer much analytical purchase in and of itself. Other causal chains inevitably need to be considered as well as their interaction.
This is also true for the end of the Cold War. The Gorbachev foreign policy revolution was undeniably a response, at least in part, to widespread sentiment among the Soviet elite that a stagnant economy severely constrained the country’s domestic and foreign policy choices. But there is a lot more to the story. Cutting loose from Third World commitments, internal democratization, toleration of the non-communist regimes in Eastern Europe and the unification of Germany within NATO were the direct result of Mikhail Gorbachev’s advocacy of glasnost , perestroika at home and “new thinking” in foreign affairs. Other possible Soviet leaders (e.g., Grishin, Romanov, Ligachev) would almost certainly have pursued a different set of domestic and foreign strategies. 40 Gorbachev’s foreign policies and their consequences were also a response to independent developments in Eastern Europe and the successful personal relationship he established with Western leaders, most notably with Ronald Reagan. To understand the end of the Cold War, we need to consider international and domestic structures, domestic politics, leaders and the non-linear ways in which they interacted.
Gorbachev’s consolidation of power, the collapse of communist governments in Eastern Europe and the Soviet Union’s accommodation with the West were also facilitated by a series of catalysts (e.g., Chernobyl, Mathias Rust’s flight to Red Square, the Reagan administration’s endorsement of the “double zero” option in the mistaken expectation that Moscow would reject it outright, the demand of East German tourists in Hungary to travel to West Germany via Austria and Budapest’s initial willingness to accommodate them). It is quite likely that the end game of the Cold War would have unfolded differently in the absence of these events, and it is not too difficult to imagine a series of alternative events that might have pushed it along a very different course. The Berlin Wall might still be standing, and the two superpowers deploying yet another generation of nuclear-tipped missiles. 41
Contingencies in the form of random events and conjunctures of multiple chains of causation are difficult to deal with theoretically. Random events, by definition, lie outside our theories, as do conjunctures, which in turn may be caused by random events. In January 1999, Pope John Paul II made a scheduled stopped in St. Louis during his tour of the United States, and pleaded with Governor Mel Carnahan of Missouri to commute to life imprisonment the sentence of a murderer scheduled for execution the next day. The governor honored the pope’s request even though the condemned man’s crime had been particularly heinous — a triple murder to which he had confessed — and he had no credible claim to leniency on any ground. The next person scheduled for execution, Roy “Hog” Roberts, had attracted national attention because of the strong possibility that he was innocent, and numerous groups and prominent individuals, including the pope, appealed — without success — to Governor Carnahan to commute his sentence. We can surmise that the governor, who had campaigned in support of the death penalty, was not about to commute the sentence of two murderers in a row no matter how deserving the second offender might have appeared. Had the pope arrived in town a few weeks later and made his initial plea for the life of Roberts it is likely that the governor might have commuted his sentence. Studies of the death penalty and its application would never have predicted the actual outcome, because they would not have considered the pope’s itinerary as a relevant variable. 42
As in the case of the pope’s visit, coincidences that have unpredictable consequences may be critical to the explanation of particular decisions or courses of action. For this reason alone prediction in individual cases is an unrealistic goal. Conditional forecasting may be a more appropriate strategy for attempting to cope with the manifold uncertainties associated with as complex as the events responsible for system transformations. Conditional forecasts use existing theories and behavioral regularities as a starting point to develop alternate scenarios of likely future developments, or of a system transformation. They consider multiple chains of causation and look at some of the possible interactions that might take place among them, as well as the paths that might lead from one scenario to another. They also stipulate the kind of information or events that will used to determine the extent to which events track according to the expectations of any of the scenarios. As events unfold, researchers repeatedly revise their scenarios and expectations in light of the new information. Such a process is messy and time consuming, but it is the only reasonable way of taking into account coincidence and random events. At the very least, it can provide early warning of major changes in a system or of the faulty expectations of those who are tracking its performance. 43
Catalysts: In many physical processes catalysts are unproblematic. Chain reactions are triggered by the decay of atomic nuclei. Some of the neutrons they emit strike other nuclei prompting them to fission and emit more neutrons, which strike still more nuclei. Physicists can calculate how many kilograms of Uranium 235 or Plutonium at given pressures are necessary to produce a chain reaction. They can take it for granted that if a “critical mass” is achieved, a chain reaction will follow. This is because trillions of atoms are present, and at any given moment enough of them will decay to provide the neutrons needed to start the reaction. Wars and accommodations, and the system transformations they may bring about, involve relatively few actors. And unlike the weak force responsible for nuclear decay, political catalysts are not inherent properties of the interacting units. For both these reasons, we can never know if or when an appropriate political catalyst will occur.
Political catalysts differ from their physical counterparts in another important respect. They are often causes in their own right, as was Sarajevo. The twin assassinations caused the Austrian leadership to reframe the problem of Serbia. Risks that had been unacceptable in the past now became tolerable, even necessary. The independent role of catalysts creates another problem for theories and attempts to evaluate them. All the relevant underlying causes for an outcome may be present but absent a catalyst it will not occur. The uncertain and evolving relationship between underlying and immediate causes not only renders point prediction impossible, it renders problematic more general statements about the causes of war and system transformations — and many other international phenomena — because we have no way of knowing which of these events would have occurred in the presence of appropriate catalysts, and we cannot assume that their presence or absence can be treated as random. It is thus impossible to define the universe of such events or to construct a representative sample of them.
The independent role of catalysts in some classes of events also renders statistical tests meaningless because of the impossibility of coding outcomes. If war is the dependent variable, researchers will distinguish between cases or interactions that ended in war from those that did not, and look for association between these outcomes and their independent variables. Their results will be misleading if war would have occurred in the presence of their variables if a suitable catalyst had been present. Meaningful statistical studies would require two stage data sets that accounted for this variation.
Theorizing about catalysts is difficult because they are so often situation specific. It is nevertheless useful to distinguish between situations in which actors are actively looking for an excuse for war, and those in which the catalyst reshapes the way the think about the situation, making them more willing than they were previously to consider high-risk options because of the greater perceived costs of inaction. Classic examples of the former include Hitler was in 1939 and Lyndon Johnson in 1964; both leaders invented pretexts to go to war when their adversaries failed to provide them. Sarajevo may be the paradigmatic case of the situation in which catalysts play important, independent roles.
Once again, the best way to address the problem of catalysts in specific contexts is likely to be through some form of scenario generation. Analysts can reason forward and ask themselves what kind of event(s) would be required to prompt behavior likely to bring about war, accommodation and system change. Or, they can reason backwards by identifying the kind of events most likely to occur and ask themselves if any of these would serve as effective triggers.
Actors: Most models of political behavior assume instrumental rationality, and this may be the norm in everyday decisionmaking. The evidence from both world wars and from the end of the Cold War suggest that this is not true for the series of decisions responsible for system transformations. In all these cases, the behavior in question led to results diametrically opposed to those intended by key actors, and there was ample information available at the time to suggest that this would occur.
The pathology of German decisionmaking in the two World Wars has been extensively documented. The Austrian case in 1914 is less well-known, and provides another striking example of deviance from instrumental rationality. In a seeming fit of emotion, Austrian leaders went to war to uphold the honor of the royal family, crush Serbia and remove the domestic and foreign threats to the security of their empire. They had no appropriate mobilization plan for coping simultaneously with Serbia and Russia, and could not fight the short war they deemed essential. In a longer war, they recognized they would become increasingly dependent upon Germany, and end up losing the very independence they were fighting to maintain. More enigmatic still, was their crisis policy that maximized the likelihood that Russia would intervene and confront them with the two front war they knew they were unprepared to wage.
Gorbachev aides and Western scholars agree that Gorbachev would never have made many key policy decisions if he had realized their probable consequences. This is particularly true of the cluster of decisions and statements that encouraged political change in Eastern Europe and led to the demise of communist regimes, the dissolution of the Warsaw Pact, unification of Germany within NATO and the breakup of the Soviet Union. The last of these consequences might still have been averted in Gorbachev had responded earlier and more appropriately to the threat of a coup against him by right-wing forces. None of these failures can be attributed to lack of information; they were the result of faulty strategic conceptions and overconfidence. 44
Determinists might counter with the argument that some, if not all, of these miscalculations were the result of the structural dilemmas these leaders faced. The claim is unpersuasive because in each instance there were other members of the policymaking elite who advocated different responses, and almost certainly would have pursued them if they had been in power. This counterfactual has been explored most systematically in the case of the Soviet Union from 1985 to 1990. A forthcoming study analyzes in detail how different leaders in Moscow and Washington would have interacted, and makes the case for a variety of possible outcomes, including a continuation, even intensification, of the Cold War. 45 If so, the Soviet Union would not have collapsed — at least in the short-term — and the international system would not have been transformed.
Unintended consequences have been described as emergent properties of systems. 46 As presently used in the international relations literature, the concept of emergent properties elides two significantly different consequences. The first, and most common, are the unanticipated outcomes we would expect in a stable system. Examples include arms races, runs on banks or the 1990s Asian financial crisis. These phenomena are the aggregated, unintended and counterproductive result of individual behavior based on instrumentally rational calculations of self-interest. The second kind of consequence, and the one relevant to my argument, are those that are doubly unintended. Here, behavior has unintended consequences for the actors and their strategic interaction, and also for the system as a whole. It sets in motion a chain events that lead to the system’s transformation. None of our theories, as presently constituted, can account for this outcome.
Niklas Luhmann suggests that systems are repertories of codes, and that outside influences must be translated by the logic of the system to have an effect upon its operation. 47 Outside influences in economics, for example, are translated into prices. Simple structural theories like that of Kenneth Waltz completely ignore outside effects. Waltz acknowledges that an international system can change, but says nothing about how such change might come about. More evolutionary approaches to system – Robert Jervis makes the case for them in his recent book – acknowledge that the structure and operating principles of systems undergo fundamental shifts in response to outside stimuli. World War I and the end of the Cold War indicate that systems can also be transformed through their internal operations. These cases point to the existence of a self-referential loop by which actors change their understanding of themselves, the system and how it operates.
The possibility of system change through actor learning has important implications for the study of international relations. It suggests that structural change may be the product, not the cause, of behavior — the opposite of what most realist theories contend. It also directs our attention to the understandings actors have of each other and of their environments, and how these understandings evolve and become shared, an underlying premises of constructivism. Finally, it constitutes a possible conceptual bridge between scholars in the neo-positivist tradition who privilege structures and those in the constructivist tradition who privilege ideas.
Endnotes
Note 1: F. H. Hinsley, “The Origins of the First World War,” in Keith Wilson, ed., Decisions for War, 1914 (New York: St. Martin’s Press, 1995), p. 4. Back.
Note 2: Randall L. Schweller, Deadly Imbalances: Tripolarity and Hitler’s Strategy of Conquest (New York: Columbia University Press, 1998), p. 2; A.F.K. Organski, World Politics , 2nd ed. (New York: Knopf, 1967), 202-03; A.F.K. Organski and Jacek Kugler, The War Ledger (Chicago, IL: University of Chicago Press, 1980). Robert Gilpin, War and Change in World Politics (New York: Cambridge University Press, 1981), 200-01. Back.
Note 3: Gabriel Almond and Stephen J. Genco, “”Clouds, Clocks, and the Study of Politics,” World Politics 29 (1977), 496-522. Back.
Note 4: On the conduct of counterfactual experiments in history and international relations, see Philip E. Tetlock and Aaron Belkin, “Counterfactual Thought Experiments in World Politics: Logical, Methodological, and Psychological Perspectives,” in Philip E. Tetlock and Aaron Belkin, eds., Counterfactual Thought Experiments in World Politics: Logical, Methodological, and Psychological Perspectives (Princeton: Princeton University Press, 1996), 15-16, and Richard Ned Lebow, “What’s So Different About a Counterfactual,” World Politics 52 (July 2000), forthcoming. Back.
Note 5: Max Weber, “Objective Possibility and Adequate Causation in Historical Explanation,” in The Methodology of the Social Sciences (Glencoe, Ill.: Free Press of Glencoe, 1949 [1905]), 164-88. Back.
Note 6: Patricia L. Kendall and Katherine M. Wolf, “The Analysis of Deviant Cases in Communications Research,” in Paul Lazarsfeld and Frank Stanton, eds., Communications Research: 1948-49 (New York: Harper, 1949), 152-57; Arend Lijphart, “Comparative Politics and the Comparative Method,” American Political Science Review 65 (September 1971), 682-93; Harry Eckstein, “Case Study Theory in Political Science,” in Fred I. Greenstein and Nelson W. Polsby, eds., Strategies of Inquiry (Reading, Mass.: Addison-Wesley, 1975). Back.
Note 7: Fritz Stern, “Bethmann Hollweg and the War: The Limits of Responsibility,” in Fritz Stern and Leonard Krieger, eds., The Responsibility of Power: Historical Essays in Honor of Hajo Holborn (Garden City, N.Y.: Doubleday, 1967), 252-85; Konrad J. Jarausch, “The Illusion of Limited War: Chancellor Bethmann Hollweg’s Calculated Risk, July 1914,” Central European History , 2 (1969), 48-76; The Enigmatic Chancellor: Bethmann Hollweg and the Hubris of Imperial Germany (New Haven: Yale University Presses, 1973); Karl Dietrich Erdmann, “War Guilt 1914 Reconsidered: A Balance of New Research,” in H. W. Koch, ed., The Origins of the First World War 2 nd ed. (Oxford: Oxford University Press, 1984), 343-70; Wolfgang J. Mommsen, Das Zeitalter des Imperialismus (Frankfurt, Fischer Bucherei, 1969), “Domestic Factors in German Foreign Policy Before 1914,” Central European History 6 (1972), 3-43; Andreas Hilgruber, “Riezlers Theorie des kalkulieren Risikos und Bethmann Hollwegs politische Konzeption in der Julikrise 1914,” Historische Zeitschrift 202 (1966), 333-51; Egmont Zechlin, Krieg und Kriegrisiko: Zur deutschen Politik im Ersten Weltkrieg (Düsseldorf: Droste, 1979); David Kaiser, “Germany and the Origins of the First World War,” Journal of Modern History 55 (1983), 442-74; Hartmut Pogge von Strandmann, “Germany and the Coming of War,” in Evans and Pogge von Strandmann, The Coming of the First World War , 87-124; Holger H. Herwig, The First World War: Germany and Austria-Hungary, 1914-1918 (London: Arnold, 1998), Ch 1; John C. G. Röhl, “Germany,” in Keith Wilson, ed., Decisions for War, 1914 (New York: St. Martin’s Press, 1995), 27-54. On Austria, see Samuel R. Williamson, Jr., “Influence, Power, and the Policy Process: The Case of Franz Ferdinand,” The Historical Journal 17 (1974), 417-34, “The Origins of World War I,” Journal of Interdisciplinary History 18 (Spring 1988), 795-818, and Austria-Hungary and the Coming of the First World War (London: Macmillan, 1990); R. J. Evans, “The Habsburg Monarchy and the Coming of War,” in R. J. Evans and Hartmut Pogge von Strandmann, eds., The Coming of the First World War (Oxford: Oxford University Press, 1988), 33-56. Back.
Note 8: Kurt Riezler, Tagebücher, Aufsatze, Dokumente. Eingeleidet und hrs. Von Karl Dietrich Erdmann (Göttingen: Vandenhoeck & Ruprecht, 1972); Koch, ed., The Origins of the First World War , Introduction. Back.
Note 9: J. L. Mackie, “Causes and Conditions,” in Myles Brand, ed., The Nature of Causation (Urbana: University of Illinois Press, 1976), 308-44; Roy Bhaskar, The Possibility of Naturalism. A Philosophical Critique of the Contemporary Human Science (Brighton: Harvester Press, 1979); Rom Harré and Peter Secord, The Explanation of Social Behaviour (Oxford: Basil Blackwell, 1973); Heikki Patomäki, “How to Tell Better Stories about World Politics,” European Journal of International Relations 2 (March 1996), 105-34. Back.
Note 10: Robert Jervis, System Effects: Complexity in Political and Social Life (Princeton: Princeton University Press, 1997). Back.
Note 11: Barry Nalebuff, "Brinkmanship and Nuclear Deterrence: The Neutrality of Escalation," Conflict Management and Peace Science 9 (1986), 19-30; Robert Powell, "Crisis Bargaining, Escalation, and MAD," American Political Science Review 81 (September 1987), 717-27, and "Nuclear Brinkmanship with Two-Sided Incomplete Information," American Political Science Review 82 (March 1988), 155-78; Steven J. Brams, Superpower Games (New Haven: Yale University press, 1985), 48-85 Back.
Note 12: Richard Ned Lebow, Between Peace and War: The Nature of International Crisis (Baltimore: Johns Hopkins University Press, 1981), ch. 2. Back.
Note 13: Senator Gravel Edition, The Pentagon Papers: The Defense Department History of United States Decisionmaking in Vietnam 4 vols. (Boston: Beacon Press, 1971), vol. 3, 687-91; Townsend Hoopes, The Limits of Intervention: An Inside Account of How the Johnson Policy of Escalation Was Reversed (New York: David McKay, 1969), p. 30. Back.
Note 14: Joachim Remak, “1914 — The Third Balkan War: Origins Reconsidered,” Journal of Modern History 43 (September 1971), 353-66. Back.
Note 15: F. R. Bridge, From Sadowa to Sarajevo: The Foreign Policy of Austria-Hungary, 1866-1914 (London: Routledge & Kegan Paul, 1972), 334-35. Back.
Note 16: Remak, “1914 — The Third Balkan War”; Bridge, 335-37; Herwig, The First World War , 8-18; Bridge, From Sadowa to Sarajevo , 335-36. Back.
Note 17: Samuel R. Williamson, Jr. to the author, 22 January 1999. Back.
Note 18: Luigi Albertini, The Origins of the War of 1914 (new York: Oxford University Press, 1965), II, 129-30; Williamson, Austria-Hungary and the Origins of the First World War , (London: Macmillan, 1990), p. 85. Back.
Note 19: Williamson, “Influence, Power, and the Policy Process,” and Austria-Hungary and the Origins of the First World War , p. 51. Back.
Note 20: Alexander Spitzmüller-Harmersbach, . . . und hat auch Ursach, es zu Leben (Vienna: William Frick, 1955), p. 103. Georg Franz, Erzherzog Franz Ferdinand und die Pläne zur Reform der Habsburger Monarchie (Vienna: G.D.W. Callwey, 1943), p. 107, quotes a similar statement Franz Ferdinand made to his wife in 1913. Back.
Note 21: Williamson, “Influence, Power, and the Policy Process,” and “The Origins of World War I.” Back.
Note 22: Williamson, Austria-Hungary and the Coming of the First World War , p. 434; David Stevenson, The First World War and International Politics (Oxford: Oxford University Press, 1988), p. 14, believes war much less likely in general if Franz Ferdinand had remained alive and continued to exercise restraint in Vienna. Back.
Note 23: József Galántai, Hungary in the First World War (Budapest: Akadémiai Kiadó, 1989(, 100-18; Gabor Vermes, István Tisza: The Liberal Vision and Conservative Statecraft of a Magyar Nationalist (New York: Columbia University Press, 1985), 230-31; Williamson, “Influence, Power, and the Policy Process.” Back.
Note 24: Konrad Jarausch, email to the author, 4 January 1999. Back.
Note 25: Moltke to Conrad, Der grosse Politik der europäischen Kabinette , ed., A, Mendelsohn-Bartholdy, I. Lepsius and F. Thimme, 40 vols. (Berlin, 1922-26), 10 February 1913, 38, I, No. 12,824; Franz Conrad von Hötzendorff, Aus meiner Dienstzeit , 5 vols. (Vienna: Rikola, 1921-25), 3, 144ff. Back.
Note 26: Alfred von Tirpitz, Politisiche Dokumente , 2 vols. (Berlin: Cotta, 1924-26), I, p. 242. Back.
Note 27: Lebow, Between Peace and War , 122-29, on the hold of the Bosnian precedent on German leaders in 1914. Back.
Note 28: Irving L. Janis and Leon Mann, Decision-Making: A Psychological Analysis of Conflict. Choice and Commitment (New York: Free Press, 1977), 59-60, 205, and Lebow, Between Peace and War , 135-45. Back.
Note 29: Vladimir Dedijer, The Road to Sarajevo (New York: Simon & Schuster, 1966), 10-12, 408-09. Back.
Note 31: Herman Kantorowicz, The Spirit of British Policy and the Myth of the Encirclement of Germany (London, 1932), p. 360, and Imanuel Geiss, German Foreign Policy, 1871-1914 (London: Routledge & Kegan Paul, 1976), p. 126. Back.
Note 32: Hans Rogger, “Rusian in 1914” in Walter Laqueur and George L. Mosse eds., 1914: The Coming of the First World War (New York: Harper & Row 1966), 229-53; Arno J. Mayer, “Domestic Causes of the First World War,” in Fritz Stern and Leonard Krieger, eds., The Responsibility of Power: Historical Essays in Honor of Hajo Holborn (Garden City, N.Y.: Doubleday, 1967), 286-93. Back.
Note 33: Cited in Z. A. B. Zeman, “The Balkans and the Coming of War, in R. J. Evans and Hartmut Pogge von Strandmann, eds., The Coming of the First World War , p. 24; Dedijer, The Road to Sarajevo , Ch. 7. Back.
Note 34: Gerhard Ritter, The Sword and the Scepter: The Problem of Militarism in Germany , trans. Heinz Norden (Coral Gables, Fl.: University of Miami Press, 1970), II, p. 196. Back.
Note 35: Stig Förster, “Der deutsche Generalstab und die Illusion des kurzen Krieges, 1871-1914: Metakritik eines Mythos,” Militärgeschichtliche Mitteilungen 54, no. 1 (1995), 61-95. Back.
Note 36: William C. Wohlforth, “Realism and the End of the Cold War,” International Security 19 (Winter 1994-95), 91-129; Richard Ned Lebow, “The Long Peace, the End of the Cold War, and the Failure of Realism,” and Kenneth A. Oye, “Explaining the End of the Cold War: Morphological and Behavioral Adaptations to the Nuclear Peace?,” in Richard Ned Lebow and Thomas Risse-Kappen, International Relations Theory and the End of the Cold War (New York: Columbia University Press, 1995), 23-56, 57-84. Back.
Note 37: Organski and Kugler, The War Ledger , Gilpin, War and Change in World Politics ; Charles F. Doran and Wes Parsons, "War and the Cycle of Relative Power," The American Political Science Review 74 (December 1960), 947-65. Back.
Note 38: Gilpin, War and Change in World Politics . 191-92, 197. Back.
Note 39: Wohlforth, “Realism and the End of the Cold War”; Lebow, “The Long Peace, the End of the Cold War, and the Failure of Realism.” Back.
Note 40: Archie Brown, The Gorbachev Factor (New York: Oxford University Press, 1996), passim. Back.
Note 41: George Breslauer and Richard Ned Lebow, “Leadership and the End of the Cold War: A Counterfactual Experiment,” in Richard K. Herrmann and Richard Ned Lebow, eds., Learning from the Cold War , forthcoming. Back.
Note 42: Jim Salter, Associated Press, 9 March 1999, “Missouri Set to Execute”; Court TV website, “Missouri Death Row Inmate Executed After Pope’s Futile Plea for Governor’s Mercy,” 10 March 1999; Evelyn Nieves, “Being in the Wrong Place, New York Times , 9 May 1999, New of the Week in Review. Back.
Note 43: Steven Weber, “Prediction and the Middle East Peace Process,” Security Studies 6 (1997), 167-79; Janice Gross Stein et al. “Five Scenarios of the Israeli-Palestinian Relationship in 20002: Works in Progress,” Security Studies 7 (1998), 195-212; Steven Bernstein, Richard Ned Lebow, Janice Gross Stein and Steven Weber, “God Gave Physics the Easy Problems,” European Journal of International Relations , 6 (March 200), 43-76. Back.
Note 44: See the chapters by Archie Brown, Jacques Lévesque, Raymond L. Garthoff, and James Davis and William Wohlforth in Herrmann and Lebow, Learning from the Cold War . Back.
Note 45: Breslauer and Lebow, “Leadership and the End of the Cold War.” Back.
Note 46: Kenneth N. Waltz, Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979); Jervis, System Effects ; Lars-Erik Cederman, Emergent Properties in World Politics: How States and Notions Develop and Dissolve (Princeton: Princeton University Press, 1997). Back.
Note 47: Niklas Luhmann, Die Gesellschaft der Gesellschaft , 2 vols. (Frankfurt am Main: Suhrkamp, 1998) I, ch. 2. Back.
