Complex Systems: The Role of Interactions*

Robert Jervis

Although we all know that social life and politics constitute systems and that many outcomes are the unintended consequence of complex interactions, the basic ideas of systems do not come readily to mind and so often are ignored. Because I know international politics best and this area is of greatest interest to readers of this book, I will often focus on it. But the arguments are more general and I will take examples from many fields. This is not difficult: systems have been analyzed by almost every academic discipline because they appear throughout our physical, biological, and social world. The fact that congruent patterns can be found across such different domains testifies to the prevalence and power of the dynamics that systems display. Much of this constitutes variations on a few themes, in parallel with Darwin’s summary remark about the structures of living creatures: "Nature is prodigal in variety, but niggard in innovation."1

We are dealing with a system when (a) a set of units or elements are inter-connected so that changes in some elements or their relations produce changes in other parts of the system and (b) the entire system exhibits properties and behaviors that are different from those of the parts.

The result is that systems often display non-linear relationships, outcomes cannot be understood by adding together the units or their relations, and many of the results of actions are unintended. Complexities can appear even in what would seem to be simple and deterministic situations. Thus over 100 years ago the mathematician Henri Poincaré showed that the motion of as few as three bodies (such as the sun, the moon, and the earth), although governed by strict scientific laws, defies exact solution: while eclipses of the moon can be predicted thousands of years in advance, they cannot be predicted millions of years ahead, which is a very short period by astronomical standards. 2

International history is full of inter-connections and complex interactions. Indeed, this one might seem like a parody were it not part of the events leading up to the First World War:

By the end of the summer of 1913 there was a real danger of yet another Balkan conflict: the King of Greece [said] that Turkey was preparing an expedition to recover the island in Greek hands, and from Constantinople the German ambassador reported that the Bulgarian minister to the Porte had informed him of a verbal Turco-Bulgarian agreement under which Bulgaria would attack Thrace in the event of a Turco-Greek war. The danger that a Turco-Greek war could spread beyond the Balkans could not be lightly dismissed. If Turkey and Greece came to blows the Bulgarians could be expected to seek revenge for the defeats of the previous summer; so early a repudiation of the Treaty of Bucharest would offend the Rumanians, whilst the Greeks, if attacked by the Bulgarians, could still invoke their treaty with Serbia. If Serbia became involved no one could guarantee that Austria-Hungary would once again stand aside.3

Ripples move through channels established by actors’ interests and strategies. When these are intricate, the ramifications will be as well, and so the results can surprise the actor who initiated the change. The international history of late 19th and early 20th centuries, centered on maladroit German diplomacy, supplies several examples. Dropping the Reinsurance Treaty with Russia in 1890 simplified German diplomacy, as the Kaiser and his advisors had desired. More important, though, were the indirect and delayed consequences, starting with Russia’s turn to France, which increased Germany’s need for Austrian support, thereby making Germany hostage to her weaker and less stable partner. In 1902, the Germans hoped that the Anglo-Japanese Alliance, motivated by Britain’s attempt to reduce her isolation and vulnerability to German pressure, would worsen British relations with Russia (which was Japan’s rival in the Far East) and France (which sought British colonial concessions).4 There were indeed ramifications, but they were not to Germany’s liking. The British public became less fearful of foreign ties, easing the way for ententes with France and Russia. Furthermore, Japan, assured of Britain’s benevolent neutrality, was able to first challenge and then fight Russia. The Russian defeat, coupled with the strengthening of the Anglo-Japanese treaty, effectively ended the Russian threat to India and so facilitated Anglo-Russian cooperation, much against Germany’s interests and expectations.

In a system, the chains of consequences extend over time and many areas: the effects of action are always multiple. Doctors call the undesired impact of medications "side effects." Although the language is misleading—there is no criteria other than our desires that determines which effects are "main" and which are "side"—the point reminds us that disturbing a system will produce several changes. Garrett Hardin gets to the heart of the matter in pointing out that, contrary to many hopes and expectations, we cannot develop or find "a highly specific agent which will do only one thing.... We can never do merely one thing. Wishing to kill insects, we may put an end to the singing of birds. Wishing to ‘get there’ faster we insult our lungs with smog."5 Seeking to protect the environment by developing non-polluting sources of electric power, we build windmills that kill hawks and eagles that fly into the blades; cleaning the water in our harbors allows the growth of mollusks and crustaceans that destroy wooden piers and bulkheads; adding redundant safety equipment makes some accidents less likely, but increases the chances of others due to the operators’ greater confidence and the interaction effects among the devices; placing a spy in the adversary’s camp not only gains valuable information, but leaves the actor vulnerable to deception if the spy is discovered; eliminating rinderpest in East Africa paved the way for canine distemper in lions because it permitted the accumulation of cattle, which required dogs to herd them, dogs which provided a steady source for the virus that could spread to lions; releasing fewer fine particles and chemicals into the atmosphere decreases pollution but also is likely to accelerate global warming; pesticides often destroy the crops that they are designed to save by killing the pests’ predators; removing older and dead trees from forests leads to insect epidemics and an altered pattern of regrowth; allowing the sale of an anti-baldness medicine without a prescription may be dangerous because people no longer have to see a doctor, who in some cases would have determined that the loss of hair was a symptom of a more serious problem; flying small formations of planes over Hiroshima to practice dropping the atomic bomb accustomed the population to air raid warnings that turned out to be false alarms, thereby reducing the number of people who took cover on August 6.6

In politics, connections are often more idiosyncratic, but their existence guarantees that here too most actions, no matter how well targeted, will have multiple effects. For example, William Bundy was correct to worry that putting troops into Vietnam might not make that country more secure because deployment could not only lead the North to escalate, but also might "(1) cause the Vietnamese government and especially the army to let up [and] (2) create adverse public reactions to our whole presence on ‘white men’ and ‘like the French’ grounds."7 It seems that the American development of nuclear weapons simultaneously restrained Stalin by increasing his fear of war and made him "less cooperative and less willing to compromise, for fear of seeming weak."8 Indeed, it is now widely accepted that mutual second strike capability not only decreased the chance of nuclear war but also made it safer for either side to engage in provocations at lower levels of violence.9 (Similarly, providing security guarantees to the countries of East Europe might lead them to take harsher stances toward minority ethnic groups and make fewer efforts to maintain good relations with their neighbors.) To mention three more surprising cases, in the fall of 1948 General Clay warned that American budget deficits would be seen in Europe as a forerunner of inflation and so would undermine morale in West Berlin; the American pressure on the Europeans to rearm more rapidly in response to the North Korean attack on the South produced squabbles that encouraged the USSR "to believe that contradictions in the enemy camp ultimately would tear apart the enemy coalition....[and so] undermined U.S. bargaining power"; in 1994 the dollar strengthened after President Clinton hired a powerful lawyer to defend him against charges of sexual harassment: as one currency trader put it, "we were starting to lose faith in him and that helped turn things."10

Interactions, Not Additivity

Because of the prevalence of inter-connections, we cannot understand systems by summing the characteristics of the parts or the bilateral relations between pairs of them.11 This is not to say that such operations are never legitimate, but only that when they are we are not dealing with a system. More precisely, actions often interact to produce results that cannot be comprehended by linear models.

Linearity involves two propositions: (1) changes in system output are proportional to changes in input...and (2) system outputs corresponding to the sum of two inputs are equal to the sum of the outputs arising from the individual inputs.12

Intuitively, we often expect linear relationships. If a little foreign aid slightly increases economic growth, then more aid should produce greater growth. But in a system a variable may operate through a non-linear function. That is, it may have a disproportionate impact at one end of its range. Sometimes even a small amount of the variable can do a great deal of work and then the law of diminishing returns sets in, as is often the case for the role of catalysts. In other cases very little impact is felt until a critical mass is assembled. For example, women may thrive in a profession only after there are enough of them so that they do not feel like strangers. Clausewitz noted a related effect:

The scale of a victory does not increase simply at a rate commensurate with the increase in size of the defeated armies, but progressively. The outcome of a major battle has a greater psychological effect on the loser than the winner. This, in turn, gives rise to additional loss of material strength [through abandonment of weapons in a retreat or desertions from the army], which is echoed in loss of morale; the other two become mutually interactive as each enhances and intensifies the other.13

Similarly, the effect of one variable or characteristic can depend on which others are present. Thus even if it is true that democracies do not fight each other in a world where other regimes exist, it would not follow that an entirely democratic world would necessarily be a peaceful one: democracies might now be united by opposition to or the desire to be different from autocracies and once triumphant might turn on each other. (The other side of this coin is that many of the characteristics of democracies that classical Realists saw as undermining their ability to conduct foreign policy—the tendency to compromise, heed public opinion, and assume others are reasonable—may serve them well when most of their interactions are with other democracies.)

To further explore interactions it is useful to start with the basic point that the results cannot be predicted from examining the individual inputs separately. I will then move on to the ways in which the effect of one actor’s strategy depends on that of others, after which I will discuss how the actors and their environments shape each other, sometimes to the point where we should make the interaction itself the unit of analysis.

First Interactions: Results Cannot Be Predicted From the Separate Actions

The effect of one variable frequently depends on the state of another, as we often see in everyday life: each of two chemicals alone may be harmless but exposure to both could be fatal; patients have suffered from taking combinations of medicines that individually are helpful. So research tries to test for interaction effects and much of modern social science is built on the understanding that social and political outcomes are not simple aggregations of the actors’ preferences because very different results are possible depending on how choices are structured and how actors move strategically.

Turning to international politics, Shibley Telhami argues that while pan-Arabism and pro-Palestinian sentiment worked to enhance Egyptian influence when Egypt was strong, they made it more dependent on other Arab states when Egypt was weak.14 From the fact—if it is a fact—that nuclear weapons stabilized Soviet-American relations we cannot infer that they would have a similar impact on other rivalries because variables that interact with nuclear weapons may be different in these cases (and of course may vary from one pair of rivals to another). Within the military domain one finds interaction effects as well: two weapons or tactics can work particularly well together and indeed most analysts stress the value of "combined arms" techniques that coordinate the use of infantry, artillery, armor, and aircraft. Events that occur close together also can have a different impact than they would if their separate influences were merely summed. The Soviet invasion of Afghanistan affected American foreign policy very deeply in part because it came on the heels of the Iranian revolution, which undercut American power, disturbed public opinion, and frightened allies.

In explaining outcomes, we are prone to examine one side’s behavior and overlook the stance of the other with which it is interacting. Although deterrence theory is built on the idea of interdependent decisions, most explanations for why deterrence succeeds in some cases and fails in others focus on differences in what the defender did while ignoring variation in the power and motivation of the challenger, just as much policy analysis in general starts—and often ends—with the strengths and weaknesses of the policies contemplated and adopted. But one hand cannot clap; we need to look at the goals, resources, and policies of those with whom the actor is dealing. Teachers are prone to make the parallel error of not exploring how shortcomings in our students’ performances on tests may be attributable to the questions we ask.

Second Interactions: Strategies Depend on the Strategies of Others

Further complexities are introduced when we look at the interactions that occur between strategies when actors consciously react to others and anticipate what they think others will do. Obvious examples are provided by many diplomatic and military surprises: a state believes that the obstacles to a course of action are so great that the adversary could not undertake it; the state therefore does little to block or prepare for that action; the adversary therefore works especially hard to see if he can make it succeed. As an 18th century general explained, "In war it is precisely the things which are thought impossible which most often succeed, when they are well conducted."15 In the war in Vietnam, the U.S. Air Force missed this dynamic and stopped patrolling sections of the North’s supply lines when reconnaissance revealed that the number of targets had greatly diminished: after the attacks ceased the enemy resumed use of the route.16

Both the success and failures of policies are determined interactively. This means that many cases of intelligence failure are mutual—i.e., they are failures by the side that took the initiative as well as by the state that was taken by surprise. Indeed, an actor’s anticipation of what others will do stems in part from its estimate of what the other thinks the actor will do. In many cases of surprise a state sees that a certain move by the adversary cannot succeed and therefore does not expect the other to take it: the U.S. did not expect the Russians to put missiles into Cuba or Japan to attack Pearl Harbor because American officials knew that the U.S. would thwart these measures if they were taken. These judgments were correct, but because the other countries saw the world and the U.S. less accurately, the American predictions were also inaccurate.17

Third Interactions: Behavior Changes the Environment

Initial behaviors and outcomes often influence later ones, producing powerful dynamics that explain change over time and that cannot be captured by labeling one set of elements "causes" and other "effects." Although learning and thinking play a large role in political and social life, they are not necessary for this kind of temporal interaction. Indeed, it characterizes the operation of evolution in nature. We usually think of individuals and species competing with one another within the environment, thus driving evolution through natural selection. In fact, however, there is coevolution: plants and animals not only adapt to the environment, they change it. As a result, it becomes more hospitable to some life forms and less hospitable to others.

Nature is not likely to "settle down" to a steady state as the development or growth of any life form will consume—and be consumed by—others, closing some ecological niches and opening others, which in turn will set off further changes. To some extent, organisms create their own environments, not only by direct actions (e.g., digging burrows, storing food, excreting waste products), but as their very existence alters the microclimates, nutrients, and feeding opportunities that will affect them and others. Indeed, not only does the amount of rainfall influence the vegetation that grows, but the latter affects the former as well. To take a more readily visible example, elephants thrive on acacia trees. But the latter can only develop in the absence of the former. After a while, the elephants destroy the trees, drastically changing the wildlife that the area can sustain and even affecting the physical shape of the land. In the process, they render the area uncongenial to themselves, and they either die or move on. The land is adapting to the elephants just as they are to it. One Maasai put it well: "Cows grow trees, elephants grow grasslands."18 Most consequentially, the very atmosphere that supports current life was produced by earlier forms, many of which could not survive in the new environment: long before humans, species of bacteria were so successful and generated so much pollution that they poisoned themselves.

Politics, like nature, rarely settles down as each dispute, policy, or action affects others and re-shapes the political landscape, inhibiting some behaviors and enabling others. Campaign financing reforms generated new actors in the form of PACs, new issues in the form of arguments about what PAC activities should be permitted, new debates about the meaning of the first amendment, and new groups that track the flow of money and services. These in turn affect not only how the funds are solicited and given, but also change the allies and adversaries that are available to political actors and the ways in which a variety of other issues are thought of. Political maneuvers create niches for new actors and disputes, often in ways that no one had anticipated. William Miller’s fascinating study of the Southern attempt to control—indeed choke off—the debate about slavery in the 1830s points out that by passing a "gag rule" prohibiting Congressional discussion of petitions asking for the end of the slave trade in the District of Columbia, the South called up "petitions against the gag rule itself" and made a new issue of the right to petition the government.19 Indeed, many protest movements grow as people previously unsympathetic are offended by the way the authorities respond. Each added issue may mobilize the population in a different way than did the original one—and of course the new dispute in turn changes the political environment.

It is clear that, for better and for worse, people change as they are affected by experiences, including those that they have chosen. Personal development does not mean that the person simply turns into what was latent in him. Instead, we need to take account of the situations in which he was placed. To take some examples familiar to academics, when we think about whether one of our bright undergraduates would do well in a Ph.D. program, we are likely to ask whether she enjoys and does well at independent research. But the right question may be whether she will enjoy and do well at it after she has experienced two or three years of graduate school. It is also a mistake to point to the lackluster career of a person who failed to get into a major graduate school or to receive tenure at a top school as justification for these decisions because we do not know how well she would have done in a more stimulating and demanding setting. Similar reasoning explains the limitations of the common argument that international institutions do not matter because states will ignore them "when push comes to shove" and vital interests are at stake. Although the statement is correct, it misses the role institutions can play in shaping interests and seeing that push does not come to shove.

Many of the ways in which deterrence can fail reflect interactions in which the state’s behavior changes its environment. Thus it might seem obvious that for Pakistan to build nuclear weapons could not but decrease the likelihood of an Indian attack. But this would overlook both the danger that India would feel increased pressures to preempt and the likely Indian judgment that world public opinion would be less censorious of an attack against a nuclear-armed Pakistan. Furthermore, the deterrence tactics that bring success at one point can change the other side and make future deterrence more difficult, most obviously by increasing the challenger’s dissatisfaction with the status quo and giving it incentives to "design around" the defender’s threats that previously had been adequate.20 A version of this process may have been at work in Vietnam in the late 1950s, casting a somewhat different light on the standard account of the American advisors erroneously fearing a conventional attack from the North and so training the South Vietnamese army to meet a fictitious danger. But the reason Diem’s enemies turned to guerrilla warfare may have been that he had succeeded in foreclosing the option of fighting a conventional war. The American policy was still in error in failing to anticipate the response it could trigger, but not in having been misconceived from the start.

Because actions change the environment in which they operate, identical but later behavior does not produce identical results: history is about the changes produced by previous thought and action as people and organizations confront each other through time. The final crisis leading to World War II provides an illustration of some of these processes. Hitler had witnessed his adversaries give in to pressure; as he explained, "Our enemies are little worms. I saw them at Munich."21 But the allies had changed because of Hitler’s behavior. So had Poland. As A.J.P. Taylor puts it, "Munich cast a long shadow. Hitler waited for it to happen again; Beck took warning from the fate of Benes."22

Hitler was not the only leader to fail to understand that his behavior would change his environment. Like good linear social scientists, many statesmen see that their actions can produce a desired outcome, all other things being equal, and project into the future the maintenance of the conditions that their behavior will in fact undermine. This in part explains the Argentine calculations preceding the seizure of the Falklands/Malvinas. Their leaders could see that Britain’s ability to protect its position was waning, as evinced by the declining naval presence, and that Argentina’s claim to the islands had received widespread international support. But what they neglected was the likelihood that the invasion would alter these facts, unifying British opinion against accepting humiliation and changing the issue for international audiences from the illegitimacy of colonialism to the illegitimacy of the use of force. A similar neglect of the transformative power of action may explain why Saddam Hussein thought he could conquer Kuwait. Even if America wanted to intervene, it could do so only with the support and cooperation of other Arab countries, which had sympathized with Iraq’s claims and urged American restraint. But the invasion of Kuwait drastically increased the Arabs’ perception of threat and so altered their stance. Furthermore, their willingness to give credence to Iraqi promises was destroyed by the deception that had enabled the invasion to take everyone by surprise. Germany’s miscalculation in 1917 was based on a related error: although unrestricted submarine warfare succeeded in sinking more British shipping than the Germans had estimated would be required to drive Britain from the war, the American entry (which Germany expected) led the British to tolerate shortages that otherwise would have broken their will because they knew that if they held out, the U.S. would rescue them.23

The failure to appreciate the fact that the behavior of the actors is in part responsible for the environment which then impinges on them can lead observers—and actors as well—to underestimate actors’ influence. Thus states caught in a conflict spiral believe that they have little choice but to respond in kind to the adversary’s hostility. This may be true, but it may have been the states’ earlier behavior that generated the situation that now is compelling. Robert McNamara complains about how he was mislead by faulty military reporting but similarly fails to consider whether his style and pressure might have contributed to what he was being told.24

Products of Interaction as the Unit of Analysis

Interaction can be so intense and transformative that we can no longer fruitfully distinguish between actors and their environments, let alone say much about any element in isolation. We are accustomed to referring to roads as safe or dangerous, but if the drivers understand the road conditions this formulation may be misleading: the knowledge that, driving habits held constant, one stretch is safe or dangerous will affect how people drive—they are likely to slow down and be more careful when they think the road is dangerous and speed up and let their attention wander when it is "safe." It is then the road-driver system that is the most meaningful unit of analysis. In the wake of the sinking of a roll-on roll-off ferry, an industry representative said:

With roro’s, the basic problem is that you have a huge open car deck with doors at each end. But people are well aware of this, and it is taken into account in design and operation. You don’t mess around with them. There have not been too many accidents because they are operated with such care.25

Similarly, we often refer to international situations as precarious, unstable, or dangerous. But, again, if statesmen perceive them as such and fear the consequences, they will act to reduce the danger—one reason why the Cuban missile crisis did not lead to war was that both sides felt that this could be the outcome if they were not very careful. Nuclear weapons generally have this effect. Because statesmen dread all-out war, international politics is safer than it would otherwise be, and probably safer than if war were less destructive. Conversely, like drivers on a "safe" stretch of road, decision-makers can behave more recklessly in calmer times because they have more freedom to seek unilateral gains as well as needing to generate risk to put pressure on others. For example, the relaxation of Anglo-German tensions after 1911 may have misled both countries into believing that they could afford dangerous tactics in 1914.

Circular Effects

Systems can produce circular effects as actors respond to the new environments their actions have created, often changing themselves in the process. In international politics, perhaps the most important manifestation of this dynamic is the large-scale operation of the security dilemma—i.e., the tendency for efforts to increase a state’s security to simultaneously decrease the security of others. Because states know that they cannot rely on others in the unpredictable future, they seek to protect themselves against a wide range of menaces. Thus in the 1930s Japan, which was heavily dependent on resources from outside its borders, sought to expand the area it controlled. Immediate economic needs generated by the world-wide depression increased but did not create this impulse. Nor were they brought on by specific conflicts with the Western powers. Rather what was driving was the fear that conflict might be forced upon Japan in the future, which meant that to remain secure Japan needed raw materials and larger markets. The result was the conquest of Manchuria, followed by a larger war with China, and then by the occupation of Indochina. Each move generated resistance that made the next action seem necessary, and the last move triggered the American oil embargo, which in turn pushed Japan into attacking the West before it ran out of oil. Had Japan been secure, her aggression would not have been necessary; it was the fear of an eventual war with the West that required policies that moved Western enmity from a possibility to a reality. (Of course a further irony is that World War II led to the reconstruction of international politics and the Japanese domestic system that brought Japan security, economic dominance of South East Asia, and access to markets around the world.)

Despite the familiarity of the idea that social action forms and takes place within a system, scholars and statesmen as well as the general public are prone to think in non-systemic terms. This is often appropriate, and few miracles will follow from thinking systemically because the interactive, strategic, and contingent nature of systems limits the extent to which complete and deterministic theories are possible. But we need to take more seriously the notion that we are in a system and to look for the dynamics that drive them. A distinguished student of genetics summarized his perspective in the phrase: "Nothing in biology makes sense except in the light of evolution."26 Very little in social and political life makes sense except in the light of systemic processes. Exploring them gives us new possibilities for understanding and effective action; in their absence we are likely to flounder.

End Notes

1. Charles Darwin, The Origin of Species (New York: Modern Library, 1936), p. 143.

2. For a recent discussion, see Robert Pool, "Chaos Theory: How Big an Advance?" Science, vol. 245, July 9, 1989, p. 26.

3. R.J. Crampton, The Hollow Detente: Anglo-German Relations in the Balkans, 1911-1914 (Atlantic Highlands, N.J.: Humanities Press, 1980), p. 131.

The first scholars who applied ideas from ecology to international politics noted that when elements are interconnected, "any substantial change in one sector of the milieu is nearly certain to produce significant, often unsettling, sometimes utterly disruptive consequences in other sectors." (Harold and Margaret Sprout, An Ecological Paradigm for the Study of International Politics (Princeton University, Center for International Studies, Research Memorandum no. 30, March 1968), p. 55.) A more recent study in this vein is James Rosenau, Turbulence in World Politics (Princeton: Princeton University Press, 1990).

4. P.J.V. Rolo, Entente Cordiale (New York: St. Martin’s Press, 1969), p. 121.

5. Garrett Hardin, "The Cybernetics of Competition," Perspectives in Biology and Medicine, vol. 7, Autumn 1963, pp. 79-80, emphasis added.

6. Jonathan Weisman, "Tilting At Windmills," Wildlife Conservation, vol. 97, January/February 1994, pp. 52-57; Lindsey Gruson, "Problem With Clean Harbor: Creatures Devour Waterfront," New York Times, June 27, 1993; Aaron Wildavsky, Searching for Safety (New Brunswick, N.J.: Transaction Books, 1988); Perrow, Normal Accidents; the classic case of "turned" agents was revealed in J.C. Masterman, The Double-Cross System in the War of 1939 to 1945 (New Haven: Yale University Press, 1972); Packer, "Coping with a Lion Killer," pp. 14-17; William Stevens, "Acid Rain Efforts Found to Undercut Themselves," New York Times, January 27, 1994; Richard Kerr, "Study Unveils Climate Cooling Caused by Pollutant Haze," Science, vol. 268, May 12, 1995, p. 802; Kerr, "It’s Official: First Glimmer of Greenhouse Warning Seen," ibid, vol. 270, December 8, 1995, pp. 1565-67; Nancy Langston, Forest Dreams, Forest Nightmares: The Paradox of Old Growth in the InlandWest (Seattle: University of Washington Press, 1995), pp. 148-50; 292-94; "You Want Hair, Get A Prescription," Aspen Daily News, July 28, 1994 (in the end, the FDA decided to permit freer sale of the medication: "Hair-Growth Drug to Be Sold Over the Counter," New York Times, February 13, 1996, p. C 10; Leon Sigal, Fighting to a Finish: The Politics of War Termination in the United States and Japan, 1945 (Ithaca: Cornell University Press, 1988), pp. 215-16. Those who believe that it is healthy to eat food that has not been treated with pesticides will be interested in Jane Brody, "Strong Views on Origins of Cancer," New York Times, July 5, 1994. But as we will discuss in chapters 2 and 7, not all unintended consequences are undesired.

7. Quoted in Larry Berman, "Coming to Grips with Lyndon Johnson’s War," Diplomatic History, vol. 17, Fall 1993, p. 525.

8. David Holloway, Stalin and the Bomb (New Haven: Yale University Press, 1994), p. 272.

9. Glenn Snyder, "The Balance of Power and the Balance of Terror," in Paul Seabury, ed., The Balance of Power (San Francisco: Chandler, 1965), pp. 184-201.

10. Walter Millis, ed., The Forrestal Diaries (New York: Viking, 1951), p. 526; William Stueck, The Korean War (Princeton: Princeton University Press, 1995), p. 6; quoted in Thomas Friedman, "It’s a Mad, Mad, Mad, Mad, World Money Market," New York Times, May 8, 1994, section E, p. 2. As these examples show, people’s expectations—based in part on their beliefs about others’ expectations—are central to the dynamics of systems.

11. Kenneth Waltz, Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979), p. 64; for parallel discussions in social psychology, organization theory, and ecology, see respectively Paul Watzlawick, Janet Beavin, and Don Jackson, Pragmatics of Human Communication: A Study of Interactional Patterns, Pathologies, and Paradoxes (New York: Norton, 1967), pp. 125-26, 135-39; Charles Perrow, Normal Accidents (New York: Basic Books, 1984); Pimm, Balance of Nature?, pp. 249-56.

12. Alan Beyerchen, "Nonlinear Science and the Unfolding of a New Intellectual Vision," in Richard Bjornson and Marilyn Waldman, eds., Papers in Comparative Studies vol. 6 (Columbus, Ohio: Center for Comparative Studies in the Humanities, Ohio State University Press, 1989), p. 30.

13. Carl von Clausewitz, On War, ed. and trans. Michael Howard and Peter Paret (Princeton: Princeton University Press, 1976), p. 253. For a general discussion of war from the perspective of systems effects, see Roger Beaumont, War, Chaos, and History (Westport, Conn.: Praeger, 1994).

14. Shibley Telhami, Power and Leadership in International Bargaining: The Path to the Camp David Accords (New York: Columbia University Press, 1990), pp. 12, 92-106.

15. Quoted in Reed Browning, The War of the Austrian Succession (New York: St. Martin’s Press, 1993), p. 123.

16. Barry Watts, "Unreported History and Unit Effectiveness," Journal of Strategic Studies, vol. 12, March 1989, p. 98.

17. Klaus Knorr, "Failures in National Intelligence Estimates: The Case of the Cuban Missiles," World Politics, vol. 16, April 1964, pp. 455-67; for a related discussion, see James Wirtz, The Tet Offensive (Ithaca: Cornell University Press, 1991).

18. Quoted in David Western, "The Balance of Nature," Wildlife Conservation, vol. 96, March/April 1993, p. 54.

19. William Lee Miller, Arguing About Slavery (New York: Knopf, 1996), p. 278. The classic treatments of these processes are E.E. Schattschneider, Politics, Pressures and the Tariff (New York: Prentice-Hall, 1935) and Schattschneider, The Semisovereign People (New York: Holt, Rinehart, and Winston, 1960).

20. Alexander George and Richard Smoke, Deterrence in American Foreign Policy (New York: Columbia University Press, 1974), pp. 522-30.

21. Quoted in Glenn Snyder and Paul Diesing, Conflict Among Nations (Princeton: Princeton University Press, 1977), p. 187.

22. A.J.P. Taylor, The Origins of the Second World War (Greenwich, Conn.: Fawcett, 1966), p. 242.

23. Fred Charles Iklé, Every War Must End (New York: Columbia University Press, 1971), pp. 42-48.

24. Robert McNamara, In Retrospect: The Tragedy and Lessons of Vietnam (New York: Times Books, 1995); Deborah Shapley, Promise and Power: The Life and Times of Robert McNamara (Boston: Little Brown, 1993), pp. 149-52. "`Ah, les statistiques!’ one of the Vietnamese generals exclaimed to an American friend. `Your Secretary of Defense loves statistics. We Vietnamese can give him all he wants. If you want them to go up, they will go up. If you want them to go down, they will go down’": Roger Hilsman, To Move A Nation (Garden City, N.Y.: Doubleday, 1967), p. 523.

25. Quoted in Stephen Kenzer, "Little Hope for 800 Lost in Sinking of Baltic Sea Ferry," New York Times, September 29, 1994.

26. Theodosius Dobzhansky, "Nothing in Biology Makes Sense Except in the Light of Evolution," American Biology Teacher, vol. 35, March 1972, pp. 125-29.

* Jervis, Robert. How Systems Work: A Perspective on Political and Social Life. Copyright 1997 by Princeton University Press. Reprinted by permission of Princeton University Press.

| Complexity Index | Chapter 4 |