Introduction

Ancient Greek historian Thucydides is well known for his writing on the origin of the (Second) Peloponnesian War (431 to 404 BC). He is remembered and frequently cited by contemporary international relations scholars, especially realists, for his memorable reporting of this devastating conflagration engulfing ancient Greece. His dramatization of the negotiation between Athens and Melos (the Melian Dialogue, 416 BC) captures the essence of political realism as reflected in the words of Athenian negotiators that “the strong do what they can and the weak suffer what they must.” Because they refused to surrender to Athens and pay tribute to it, the Melians were subjected to a siege which ended in their defeat (in 415 BC). Athenians killed all of Melos’ adult men and sold their women and children to slavery.

Thucydides’ popularity has enjoyed another resurgence recently with the publication of Graham Allison’s book [1], which recalls this ancient historian’s belief “that the truest cause [of the Peloponnesian War], but the least spoken of, was the growth of Athenian power, which presented an object of fear to the Spartans and forced them to go to war” [quoted in 20, pp. 2–3]. Allison was motivated to write his book by the recent rise of China that in his view could pose a challenge to the global preeminence that the U.S. has enjoyed heretofore. He has used the ancient conflict between Athens and Sparta as an analogy to warn about the possible consequences of the ongoing process of power shift in Sino-American relations. Although he explicitly disagrees with Thucydides that the dynamics of power transition (referring to the overtaking of an incumbent hegemon by a late rising power) [31] makes a war between these leading states “inevitable,” he clearly thinks that the danger of such a conflict rises significantly during this process. He reports that out of his sample of sixteen cases of past historical episodes of power transition, twelve had ended in war. Allison and Thucydides are “structuralists” in the sense that they both emphasize the impersonal forces of international system to be the primary determinant of war and peace. Remarkably, even though Thucydides speaks of the “fear” that Athens’ rise had caused for the Spartans, neither Allison nor other contemporary analysts attend to emotional factors – such as anxiety, envy, arrogance, overconfidence in addition to fear – that connect systemic processes of power transition to the policy choices made by individual leaders and their cohorts. There is therefore a critical missing link in their narratives, and their accounts of historical events tend to overlook human agency – how leaders’ actions can make a difference in keeping peace or provoking war. After all, people make decisions on war and peace and structural conditions only provide conducive circumstances that may incline individuals to certain choices. It takes both ‘willingness” (human motivation and purpose even though they may be misguided) and “opportunity” (facilitative or inhibiting conditions) to explain foreign policy [30] (while at the same time making due allowance for chance and accident).

This view does not necessarily suggest that structural approaches to international relations analyses are worthless. It does argue, however, that we need to observe certain canons of social science inquiry for our analysis results – regardless of their epistemological or methodological basis – to be accorded credibility. Moreover, although historical analogies can be valuable for decision making, they can also mislead and cause policy mistakes if they are misapplied [22]. In the current context, in order for us to accept the basic claim of Thucydides’ Trap, we will have to suspend our disbelief that new developments since the days of premodern Greece some 2500 years ago – such as the advent of modern state system, nationalism, and nuclear weapons – do not compromise the basic validity of its proposition claiming that the danger of war rises (or even becomes “inevitable”) when a late rising power catches up to or overtakes an existing dominant power.

Indeed, the historical parallel introduced by Thucydides’ Trap invites us to accept ancient Sparta as an analog of today’s U.S. and ancient Athens as an analog of today’s China. The irony of this juxtaposition is of course that Sparta was an authoritarian oligarchy and Athens was a democracy according to at least the standards prevailing at its time. Are those who want to apply Thucydides’ claim to the contemporary world serious in implying that the nature of political systems is not germane to the conduct of statecraft, or that a democratic America would become so fearful of a rising China so that Washington will be inclined to consider launching a preventive war against Beijing? A further irony in applying – or misapplying – the analogy from Thucydides is the fact that Sparta was an agrarian society and its power came from its infantry – the hoplites in phalanx formations. It was a classic land power. Conversely, Athens was known for its naval power and its overseas trade. Thus, the facile transfer of premodern Athenian-Spartan relations to contemporary Sino-American relations again breaks down over these concerns. It overlooks the considerable amount of evidence from recent social science research showing that the nature of a country’s political regime and of its power base (whether it is a continental state commanding a strong army or whether it has a powerful navy and is a trading state) are important variables that influence its foreign policy and other states’ perceptions of the threat it poses to them. There is a large body of literature reporting on the phenomenon of “democratic peace,” with Bruce Russett and John Oneal’s book providing a leading example [37]. Moreover, evidence from Europe between the Thirty Years’ War and World War II seems to suggest that a continental power with a large standing infantry tends to be perceived more threatening than a maritime power with a strong navy and an orientation emphasizing overseas commerce. Other states tend to pursue balance-of-power policies against the former in regional politics, but they do not typically engage in similar policies against the latter at the systemic or global level [26, 27].

The following discussion attends to two topics. First, I discuss some problematic aspects of the claim advanced by Thucydides’ Trap such as its scope conditions, sample selection, criteria for admissible evidence, and treatment of alternative historical interpretations. Many of these issues have been taken up by critical reviews about the power-transition theory which shares its core premise [6, 7, 9, 13, 24, 25, 34, 41]. Second, I argue that wars in general and the Peloponnesian War in particular (as reported by Thucydides himself) can stem from multiple and not necessarily mutually exclusive factors. In other words, this ancient historian’s account offers reasons to believe that there is more to the causes of war than just the structural condition of power transition.

Analytic Problems

Thucydides’ Trap is basically a bivariate proposition claiming that when a late rising power catches up to or overtakes an existing dominant power, the danger of war between them rises. Implicit in this formulation is the contention that the influence of other variables on such a clash will be overwhelmed by the dynamics of power transition, which will have the predominant, even decisive, influence over war occurrence between this pair of countries. We are essentially presented with a monocausal explanation of the war phenomenon without, however, any attempt to show that the basic causal attribution will still stand after we account for rival hypotheses on why wars happen. For example, could physical proximity, cultural affinity, armament races, economic interdependence, regime characteristics, alliance commitments, and leaders’ personalities play a role in encouraging or inhibiting war? Naturally, it may be difficult to consider all such variables in a single analysis to assess their influence in exacerbating or mitigating the impact of power transition on war occurrence. But multivariate analysis has been the norm of social science research for some time now, and analysts nowadays routinely confront the charge of spurious explanation by taking into account rival hypotheses in their studies.

The central concern of Thucydides’ Trap pertains to changes in relative national power. If you will, this is its independent variable. But how does one define and measure national power? Allison does not tell us, whereas Organski and Kugler point to gross national product as the most valid and succinct measure. How close will Country B’s power have to approach Country A’s power for us to conclude that a power transition is occurring or has occurred? That is, what is the zone of danger for war to break out? Allison is again silent on this question, whereas Organski and Kugler have stipulated that the former must have achieved at least 80% of the latter’s capabilities [31 pp. 44, 48]. Others such as Tammen et al. [31 p. 41] have stated that a power ratio of 4:5 or 5:6 between the latecomer and the leading state presents the most dangerous moment for a war to occur between them. This is hardly a trivial question because it is germane to a proposition’s (or theory’s) scope conditions. What is the empirical domain covered by its claim(s)? If we do not have an answer to this question, we cannot begin to confirm or disconfirm the validity of Thucydides’ Trap.

The World Factbook published by the U.S. Central Intelligence Agency [42] estimates that in 2017 China’s gross domestic product was $11.94 trillion at the official exchange rate, compared to $19.36 trillion for the U.S. In other words, China’s economy was about 60% of the U.S. economy. The same CIA source reports that in 2016 China spent about 1.9% of its gross domestic product on the military, compared to 3.29% for the U.S. Indeed, U.S. military expenditures had in recent years exceeded the same expenditures by China, Russia, Britain, France, Germany, Japan, India, and Saudi Arabia combined! Indeed, Washington enjoys an unassailable command of the world’s commons: its oceans, air, and even the outer space [32]. Still other scholars such as Michael Beckley [3] and Steve Chan [8] have shown that the U.S. has a commanding advantage in the leading technologies which provide the basis for enduring economic and military prowess [29, 34]. It seems a bit odd that at about the same time when some are concerned about a Sino-American power transition, others have observed that the world is out of balance because of undisputed U.S. supremacy [5]. Given this latter view, one may be rightly puzzled by the discourse on Thucydides’ Trap.

The U.S. is in an incomparably stronger position than China given its lead in technology and armament, its numerous and powerful allies, and its extensive overseas forward deployment of military forces. Whereas the U.S. is flanked by two oceans and has borders with only two relatively weak neighbors, China finds itself located in a crowded neighborhood within close physical proximity of several other great powers such as Japan, Russia, India and of course, the U.S. with its Asian outposts. It is difficult to imagine any U.S. leader would want to exchange his/her country’s strategic position with China’s.

Does the proposition that power transition portends war apply to only the major powers, or does it apply to minor powers as well? And if only the former, which countries would qualify? Organski and Kugler’s theory is specifically concerned about the two most powerful countries at the pinnacle of international system contending for global hegemony. Yet they exclude the U.S. from their analysis of the outbreak of World Wars I and II even though it had already overtaken Britain prior to both conflicts. By this exclusion, they can characterize these conflicts as the result of Germany’s challenge to Britain’s global dominance – even though Germany’s gross domestic product and its per capita income never surpassed Britain’s [25, pp. 22, 28; 34], p. 325]. One may also question whether it would be more appropriate to consider the British Dominion rather than just Britain alone as the proper unit of analysis in this case.

It is also important to note that Germany’s economy never overtook that of the U.S. Had Organski and Kugler included the Anglo-American transition in their analysis – the only case in modern history when the world’s leading economic power was actually overtaken by a latecomer – it would have disconfirmed their theory because war between these countries did not occur. Moreover, although their theory is supposed to be about a contest for global primacy and a revisionist attempt by the latecomer to alter the prevailing world order, Organski and Kugler’s analysis has included the Franco-Prussian War and the Russo-Japanese War – which do not appear to meet those standards just mentioned. At the same time, the Spanish-American war was omitted from their analysis. Consequently, their selection of both countries and wars that are supposed to fall within their theory’s scope seems quite problematic, even arbitrary. It is a bit odd that a theory of power transition can exclude the U.S. from its analysis of contenders for global supremacy on the eve of both world wars (especially after the U.S. had arguably played the decisive role in the outcome of World War I) but assigns this status to China today even though its influence is still limited mostly to the East Asia region.

Depending on one’s stipulation of those countries (typically “great powers”) that fall under the purview of Thucydides’ Trap or power-transition theory, the number of pertinent cases for testing can vary. This variation, however, does not nullify that fact that wars tend to be rare, and wars among great powers tend to be even rarer. The resulting small N problem means that it is difficult to reach firm conclusions, especially when the number of putative variables exceeds the number of cases (the so-called degrees-of-freedom problem). The inclusion or exclusion of just one or two countries or wars, such as the U.S. and the Spanish-American War, in the sample can significantly affect an analysis’s results. To compound this difficulty, power-transition analyses sometimes treat clashes between each pair of great powers as independent observations when in fact these states’ decisions to wage war were very much interrelated. This is problematic because Britain’s decision to fight Germany in 1914 had something to do with Germany’s decision to invade France through Belgium, which in turn had something to do with France’s support for Russia and Russia’s confrontation with Austria-Hungary (a German ally) over the latter’s clash with Serbia (a Russian protégé).

It also seems that some wars happened in the absence of power parity, not to mention transitional overtaking. Although one can certainly debate about whether Spain was a great power in the late 1800s, its exclusion from the ranks of these states conveniently bypasses the Spanish-American War and the troubling question whether contrary to the expectation of the power-transition theory, such conflict can be initiated by a dominant power (for the U.S. had certainly established its regional hegemony in the Western Hemisphere by 1898 and had become the world’s leading economy). One may also ask if the Spanish-American War should be omitted from analysis, what is the rationale for including the Russo-Japanese War at about the same time? Had Japan become a great power by 1904? For that matter, had Japan reached power parity with the U.S. by 1941, or did its leaders acknowledge that the U.S. was eight or nine times stronger when they decided to attack Pearl Harbor [2, 18, 19]? This case suggests that war can be initiated by the weaker side in a lopsided relationship and in the absence of approaching parity or overtaking.

For some mysterious reason, the Korean War was also omitted in Organski and Kugler’s analysis even though they consider both China and the U.S. to be members of the international relations’ “central system” by 1950. In 1951 when China intervened in this conflict to fight the U.S. [10, 45], no one could reasonably argue that it had come close to matching U.S. power. Thus, power parity or overtaking again appears to be neither a necessary nor sufficient condition for war to break out among major powers. Proponents of power-transition theory and Thucydides’ Trap overlook such cases of the occurrence of the unexpected – that wars sometimes happen in the absence of power parity or transition.

There is an additional problem with Organski and Kugler’s original analysis. They ask whether a specific war was preceded by power shifts among its then future belligerents (only four wars were included in their analysis: the two world wars, the Franco-Prussian War, and the Russo-Japanese). This feature of their research design means that they are unable to address false alarms, that is, power shifts that were not followed by war in their wake. This is not a problem with Allison’s analysis which includes cases of power shift with both peace and war as their outcomes. But as with Organski and Kugler, the criteria used by Allison to include or exclude cases are unclear.

Allison’s analysis clearly admits more than just the two most powerful states at the pinnacle of the international system. For instance, his sample includes Germany’s peaceful overtaking of Britain and France in the 1990s. It is not clear, however, why Britain and France are considered “ruling powers” during this period (long after the U.S. had become the world’s predominant power) whereas Germany is described as a “rising power.” It is also not clear why he includes these cases in his inventory of power transitions, while at the same time omitting China’s peaceful overtaking (economically) of these countries as well as Russia and Japan. Had he included these latter cases in his sample, the evidence against Thucydides’ Trap (that power transition augurs war) would be much weaker. There seems to be considerable arbitrariness regarding which cases are admissible for inclusion in the evidence for testing the claim of Thucydides’ Trap and which cases are out of bounds.

The causal arrow suggested by both the power-transition theory and Thucydides’ Trap points in only one direction from power shifts to war. It is nevertheless possible, even plausible, that war can produce power shifts. As pointed out by Ned Lebow and Benjamin Valentino [24], some past power transitions among major European states can be attributed to the defeat or collapse of the Spanish Empire, the Ottoman Empire and the Austro-Hungarian Empire rather than just catch-up by the upstart states. Surely, the rise of the U.S. was facilitated by the Europeans’ economic and military exhaustion after the two world wars. Similarly, the rise of Italy and Germany cannot be separated from conflicts related to their campaigns of national unification. The victory of Chinese communism can also be in part traced to the Japanese invasion, and the fall of the Soviet Union (in part due to its exhaustion from competing with the U.S. during the Cold War) had certainly assisted in bringing about the U.S. unipolar moment. Wars and power shifts can thus have a reciprocal influence on each other rather than just a one-way causality going from the latter to the former.

The basic unit of analysis for both Thucydides’ Trap and the power-transition theory is the dyad, or states that pair up in contentious relationships. These analyses tend to privilege some dyads over others and sometimes dismiss entirely still some dyads. As already mentioned, the Anglo-American dyad was excluded from Organski and Kugler’s analysis before 1950, This analytic move in turn enabled them to exclude the case of U.S. overtaking Britain and to omit the peaceful transition between these two countries that would otherwise have refuted their proposition. And as remarked earlier, Allison’s historical inventory also omits China’s overtaking of Japan, Russia, Britain, and France, occasions when power transition again did not result in war.

Other troubling questions come to mind. Should we accept the historical interpretations offered by proponents of the power-transition theory and Thucydides’ Trap? These narratives do not entertain alternative explanations, competing hypotheses, or just nagging puzzles. For example, if Germany had overtaken Britain, why did it not win both world wars? And if it was on an inevitable trajectory to overtake Britain, why did it not wait? And if by all measures, Japan still lagged badly behind the U.S. in national capabilities in 1941, why did it launch the Pearl Harbor attack? Moreover, Britain had surely faced multiple rivals in the late 1800s and early 1900s. Besides Germany, others such as France, Russia, Japan and indeed, the U.S. had challenged its influence in different ways and in different places. Surely, Germany was not the only expansionist power. Why then did Britain end up fighting Germany and not the other rising powers? This question in turn directs our attention to the relevance of statecraft. States are not locked into a path leading to war because of changes in their power trajectories. They live in a multilateral world where allies can become adversaries and vice versa.

As Karen Rasler and William Thompson [35, p., 310] have remarked, “declining incumbents select, to some extent, which challengers they will fight and with whom they will ally to meet the intensive challenge.” Prior to 1914, London concentrated its attention and resources on the German threat. At the same time, it decided to concede the Western Hemisphere to Washington and conciliate with Paris and St. Petersburg. It even accepted Tokyo as a junior ally in Asia [43]. That these countries became Britain’s allies and fought on its side had something to do with its victory over Germany in the two world wars (except of course for Japan which switched sides to fight on the Axis side in World War II). The choices made by these British allies were not just a result of relative power shifts but were at least in part a response to how Britain had behaved toward them. Similarly, that Germany and Britain had found themselves at war was not unrelated their respective policies toward each other, such as the distrust and tension caused by the Anglo-German naval race. There was a certain element of self-fulfilling prophecy in their reciprocal hostility. The preceding remarks suggest that monocausal explanations are inadequate. Adjustments and refinements are required to gain cogency and validity.

These remarks also argue that it is too simplistic to view international relations in strictly bilateral or dyadic terms. Germany lost World Wars I and II because its coalition was weaker than the opposing coalition. This observation in turn underscores the fact that although China’s economy and its military have expanded rapidly in the recent past, it is still in a far inferior strategic position relative to the U.S. which can count on many powerful allies. Today as before the two world wars, several large countries have been gaining relative power. Besides China, there are emergent or re-emergent powers such as Japan, Germany, Russia, and India. Thus, it behooves us to consider how the growth or decline trajectories of multiple major powers interact, and not to treat these changes in isolation or in strictly dyadic ways [15].

Being primarily structural formulations, power-transition theory and the proposition of Thucydides’ Trap tend to be quite sparse in historical details. They do not delve into diplomatic history and tend to be colored by hindsight bias [16]. They see those wars that happened as somehow inevitable or at least highly probable, and they often fail to consider alternative explanations of these conflicts. For example, some proponents of the power-transition theory try to explain away the peaceful Anglo-American transition by claiming that Britain and the U.S. had a cultural affinity and that they shared democratic institutions (even though both countries had very limited adult suffrage before World War I). Such explanation has a decided ad hoc and revisionist tendency because it cannot explain why cultural affinity did not prevent Austria and Prussia from going to war (and conflicts between other “cultural” neighbors such as the Chinese and Japanese, Israelis and Arabs, and Indians and Pakistanis, not to mention recurrent tension on the Korean peninsula and across the Taiwan Strait), and because it overlooks acrimonious Anglo-American relations during much of the nineteenth century.

Historian Kenneth Bourne [4, p., 408] points out that “The United States remained an enemy of Britain’s calculations… until 1895-96. Until after the Venezuelan affair any increase in the territory and strength of the United States was regarded as a direct threat to the British possessions and British power and influence in the western hemisphere.” As late as 1896, British Prime Minister Lord Salisbury said to his finance minister, “a war with America, not this year but in the not distant future — has become something more than a possibility” (quoted in 4, p. 337]. Expressing his regret that Britain had failed to intervene on the Confederacy’s side in the American Civil War, Salisbury was quoted saying, “It is very sad, but I am afraid America is bound to forge ahead and nothing can restore the equality between us. If we had interfered in the Confederate Wars, it was then possible for us to reduce the power of the United States to manageable proportions. But two such chances are not given to a nation in the course of its career” [quoted in 28, p. 38].

There is not only a tendency to whitewash the history of some dyads, but also a tendency to make the relations of other dyads more antagonistic and the prospect of war seem more certain than might have been the case. Although history shows that Britain and Germany went to war twice in the twentieth century, this does not mean that their conflict was inevitable or that relative power changes were responsible for their conflict. Dale Copeland [13] offers a compelling alternative explanation for both world wars, arguing that in 1914 and again in 1941 Germany initiated a preventive war aimed at Russia/the USSR, which Berlin saw as a rising future threat. Adolf Hitler had blamed London for failing to see that “an Anglo-German combination [made] the most natural of alliances” [quoted in 38, p. 114], and for not realizing that the USSR was “now the greatest power factor in the whole of Europe” [quoted in 36, p. 54). “Everything I undertake is directed against the Russians; if the West is too stupid and blind to grasp this, then I shall be compelled to come to an agreement with the Russians, beat the West, and then after their defeat turn against the Soviet Union with all my forces” [quoted in 13, p. 135]. According to this interpretation, war between Germany and Britain happened not because Berlin had wanted to fight London but rather because it was unable to persuade London to remain on the sideline. By extension, if a war between China and the U.S. should happen over Taiwan, it is not because Beijing wants to challenge Washington’s global hegemony but rather because it has failed to persuade the U.S. to stay out of its unfinished civil war. As with even the two world wars (which were fought primarily over contested mastery over especially Europe but also Asia in the 1930s and 1940s rather than a struggle for worldwide hegemony), there is a tendency for power-transition theorists to confuse regional competition with challenges to a hegemon’s global primacy [14]. Even though China has made great strides in gaining relative power, it is still a regional power and as such is not capable of challenging America’s global dominance.

Other Pathways to War

Although Thucydides concludes that power-transition dynamics made a war between Athens and Sparta “inevitable,” his own account offers many other persuasive reasons for the occurrence of this conflict. Even a casual reader of his writing would realize that the Peloponnesian War had more than one cause and that its occurrence was not inevitable. Moreover, it is far from a settled matter that Athens’ power was growing in the years prior to the war’s outbreak. Some have argued that Athens’ power did not grow between 445 and 435 BC. [20 p. 345]. Moreover, there is considerable evidence that Athens did not have an insatiable imperialist appetite and that the Spartans were not especially alarmed by or afraid of Athens’ power [20, pp. 345–346]. On the last point, in their effort to lobby Sparta to support their cause, envoys from Corinth rather alleged that the Spartans were dull, complacent and lethargic in responding to the threat coming from Athens. This characterization is far from the depiction of fear, alarm and even panic attributed to the Spartans by the oft-quoted remark by Thucydides.

From both Thucydides’ own account and the available historical record, war was far from inevitable between Sparta and Athens. The decision to support Sparta’s ally, Corinth, was decided by a divided vote in the Spartan assembly. Sparta’s citizens and elites were far from united in their determination to go to war against Athens. Even after their decision to support Corinth, the Spartans had sent at least three envoys to Athens. Although one may question whether these envoys were sent in good faith to negotiate a peaceful settlement or whether they were instead intended to buy time or gain favorable publicity, there was a good chance that had the Athenians rescinded the Megarian Decree as demanded by the Spartans, war could have been avoided. This decree had imposed an economic blockade against a Spartan ally, and its withdrawal would have imposed only rather limited symbolic cost on the Athenians. Pericles’ refusal to withdraw it, however, made Athenian policy appear unnecessarily aggressive and threatening, and made the momentum to war ever more difficult to reverse. Yet even then, actual hostility between the belligerents did not start until one year after Sparta’s declaration of hostility and it began only after Thebes, one of Sparta’s allies, invaded Plataea to advance its own agenda. In view of these considerations, war between Sparta and Athens was far from “inevitable” and several “exit ramps” were available for their leaders to deescalate tension.

Thucydides’ own account provides many reasons to doubt the inevitability of the Peloponnesian War, and it invites us to consider alternative explanations of this war’s outbreak or at least additional explanations besides the dynamics of power transition. These other explanations are not necessarily mutually exclusive, and they can be assigned to general categories of “push,” “pull,” and “misjudgment” as David Welch has suggested [44].

“Push” refers to domestic factors that motivate leaders to undertake foreign expansion and adopt belligerent policies. Leaders can try to bolster their political position and popularity, attack and undermine their domestic opponents, and seek to manipulate public attention by scapegoating foreigners and demonstrating their own nationalist credentials. Any passing acquaintance with Athenian and Spartan politics tells us that neither polity could be treated as unitary actors. Each had political factions that could be loosely described as “war” parties and “peace” parties. The political intrigues and contests involving leading figures such as Cimon, Pericles, and Thucydides are well known (Thucydides was banished to exile as he was a domestic rival of Pericles, Athens’ paramount leader).

The outcome of Sparta’s debate on whether to support Corinth was far from a foregone conclusion. King Archidamus spoke at the assembly in favor of patience, prudence, and moderation. He warned about the hardships and uncertainties associated with a long war and sought to slow down the rush to confront Athens. His effort, however, was not sufficient to win over the majority who were persuaded by Sthenelaidas’ exhortation to seek revenge against perceived insults coming from Athens and to act against rising Athenian threat. Sthenelaidas also appealed to his fellow countrymen’s sense of moral indignation and self-righteousness. As mentioned above, however, even after a majority of Sparta’s assembly had opted to support Corinth in a split decision, the domestic struggle between the peace and war parties continued. Archidamus’ influence and his policy position rose and fell due in part to Athenian action or inaction. The peace envoys mentioned above were probably sent at his instigation, and Pericles’ refusal to rescind the Megarian Decree in turn undermined Archidamus’ political stature and discredited his counsel for moderation. Thucydides’ own narrative tells us that there is more to the origin of the Peloponnesian War than just the changing power relations between Athens and Sparta. As emphasized by recent scholarship on international relations, the interaction of domestic and foreign dynamics also matters [33]. Domestic coalitional politics and attempts to appease popular sentiments often played a large role in foreign expansion and intervention, such as in the case of Japan and Germany’s imperialist agendas before the two world wars, Britain’s involvement in the Crimean War, and the U.S. decision to start the Spanish-American War [40]. Popular jingoism and sensationalism also played a role in abetting belligerence.

“Pull” factors refer to a country’s external ties that tend to draw it to foreign conflicts, sometimes even contrary to its best interests or original intentions. Contrary to the depiction that the Peloponnesian War stemmed directly from a bilateral rivalry between Athens and Sparta, its origin can in fact be traced to a conflict among secondary states that in turn engulfed these respective leaders of the Delian and Peloponnesian Leagues. As in other large conflicts such as World Wars I and II, the initial spark that set off a chain reaction came from a distant, lesser power (e.g., Serbia, Poland). The competition between Corinth and Corcyra to control Epidamus was the catalyst for the eventual war. When Corinth refused arbitration and Athens agreed to offer a defensive alliance in response to Corcyra’s request for assistance, the conditions were laid for the further bipolarization of alliance politics. Significantly, the Athenians did not reach their decision to offer a defensive alliance to Corcyra easily. It took two meetings of their assembly, with the second one reversing their initial decision that had favored the Corinthians. Thus, the Athenians’ decision was not preordained.

Faced with Athens’ overwhelming power, Corinth appealed to its ally Sparta for help, thereby setting the stage for the eventual showdown between Sparta and Athens. Significantly, Sparta had tried to exercise a moderating influence in the initial dispute between Corcyra and Corinth over Epidamus, and it had tried to restrain its ally Corinth in this dispute and had moreover sought to discourage others from getting militarily involved in this dispute. The motivation to defend an ally and to protect one’s honor and reputation in standing by one’s commitment to this ally was part of the rationale presented by Sparta’s war party for taking up the supposed challenge coming from Athens. In the parlance of contemporary studies of alliance politics, entrapment by one’s alliance commitments and the contagion effect of chain-ganging (allies being drawn into a conflict that did not initially engage their direct interests) were at least part of the story for the outbreak of the Peloponnesian War [12, 39]. One historian of this conflict describes the situation as “the Corinthian tail wagging the Spartan dog” [20 p. 26]. Had it not been for Corinth’s lobbying efforts and its threat to leave the Peloponnesian League led by Sparta, the latter would probably not have decided to accept a conflict with Athens.

Misjudgments – including those related to hubris – are also part of the story. Arguably, Pericles misjudged the effects of the Megarian Decree and his ultimatum to Potidaea (a colony founded by Corinthians) on Spartans’ perception of Athenian ambition and aggression. Conversely, Sparta’s war party arguably over-estimated the threat coming from Athens without realizing that Pericles’ policies were partly motivated by defensive reasons. For example, they failed to reckon that Athens had offered only a defensive alliance to Corcyra and had thus sought to communicate that it did not have any offensive intention to expand its empire. In his own memorable words, Pericles had warned his fellow Athenians about the danger of self-inflicted mistakes, urging them “not to try to extend [their] empire at the same time as [they] are fighting the war and not to add self-imposed dangers, for I am more afraid of our own mistakes than the strategy of our opponents” [quoted 20, p. 192]. Although Pericles is often characterized as an advocate of Athenian expansion, he seemed in this instance to be cautioning against the danger of strategic overreach and imperial overstretch [21]. Finally, misjudgment can have its source in self-righteousness when all contestants in a dispute believed that the gods would favor their just cause. The evidence for this tendency is abundant in Thucydides’ narrative.

Conclusion

Space limitation does not permit a thorough discussion of how Thucydides’ Trap can affect current and future Sino-American relations [8, 9, 23, 25, 34, 41]. It should be clear, however, that a power transition is not required for these two countries to fight. In 1950, China intervened in the Korean War and fought the U.S. in the next three years. China was much weaker at that time than now and it certainly was not in the process of matching U.S. power or overtaking it. Since the Korean War, China and the U.S. had been involved in several military crises in the Taiwan Strait. Chinese forces also saw action against India, the USSR and Vietnam. These episodes happened before Beijing’s economic reforms and its rapid growth dating from the late 1970s.

Chinese leaders are quite capable of accepting a military confrontation even when faced with a highly unfavorable balance of power. As Thomas Christensen [11] has argued, China can “pose a problem” for the U.S. without catching up. On past occasions when Beijing has resorted to arms, its leaders were evidently motivated by a sense that their foreign or domestic position would become even worse if they were to fail to take any action. In other words, they were disposed to act to prevent a further deterioration of their position. Similarly, Taylor Fravel [17] shows that Beijing has undertaken military action against other foreign opponents in the past because of its sense of domestic or foreign vulnerability rather than out of a feeling of confidence stemming from relative power gains in its international position. The cumulative impression from these and other experts on China points to a conclusion that very much diverges from the suggestion of Thucydides’ Trap. This remark does not mean that the claim of Thucydides’ Trap may not be relevant in the future. It does mean, however, that power transition is not the only reason or condition for China to enter a conflict. Historically, when China fought it has usually acted out of a defensive motivation due to an acute sense of vulnerability and a desire to forestall a deteriorating situation – even when it faces a very lopsided balance of power to its disadvantage. In the future, nationalist public opinion and competition among leadership factions may present additional pressure for Beijing to adopt a belligerent stance over issues such as Taiwan and South China Sea.

Like other conflicts, the Peloponnesian War cannot be explained by a single dynamic. Monocausal explanations are inherently problematic. Wars tend to instead result from a confluence of factors, and both structural conditions and human agency play a role in their outbreak. If a structural condition – the overtaking of an existing hegemon by a late rising power or just the approach to parity between the two – is truly an inevitable or even just a highly probable cause of systemic war, what then are its policy implications for preventing such conflict? Does this imply that for the sake of avoiding such conflict, all states other than the dominant power must refrain from growing economically and improving their military capabilities lest the hegemon feels that its dominant position is threatened? Popular rendition of the so-called Thucydides’ Trap focuses excessively on one possible explanation to the exclusion of other and even more persuasive explanations. Like its cousin the power-transition theory, its selection of cases and its interpretation of history are problematic. Its logic and evidence are shaky. Moreover, it fosters the danger of self-fulfilling prophecy to the extent that leaders in Beijing and Washington are inclined to believe in its prophecy. Of course, we do not have access to the minds of Chinese and U.S. officials to determine whether and if so, the extent to which they believe in Thucydides’ Trap. Popular and scholarly attention given to this proposition and the related topic of power transition can, however, serve the purpose of agitating and mobilizing public and elite opinion. Hardliners in both countries may be empowered by these claims to advance their respective agenda. If a power transition portends a future war, a natural implication will be to contain China’s rise before it is too late. If Chinese officials believe that their U.S. counterparts hold this view, their natural inclination will be to perceive America with suspicion, distrust and hostility. The result could be easily an escalating series of mutual recriminations. Lebow and Valentino [24, p. 408) have remarked,

Should war come between the United States and China in the future it will not be a result of a power transition. The greater risk is that conflict will result from the misperception that such a transition is imminent, and the miscalculation by decision-makers in the United States (or China) that China will soon be in a position to do what no state has done before – unilaterally dictate the rules of the international system. Power transition theory would be made self-fulfilling – generating its own corroboration where history has failed to oblige.