Join War on the Rocks and gain access to content trusted by policymakers, military leaders, and strategic thinkers worldwide.
The standard explanations for why Taiwan is so important to the Chinese Communist Party are familiar: territorial integrity, credibility of deterrence commitments, great power competition with the United States, the strategic value of the first island chain, and the semiconductor industry concentrated in TSMC’s fabrication facilities. These explanations work, but they don’t tell the whole story. Rational strategic calculation does not produce a posture in which a nuclear-armed state with the world’s second-largest economy treats a democratic island of 23 million people as a permanent, non-negotiable crisis. But what is rational depends largely on how history has defined you and the ways in which you have defined yourself.
The Chinese Communist Party’s posture toward Taiwan is driven not primarily by strategic calculation, at least not in the way it is understood from outside. Its posture is driven by a political trauma the party has consistently chosen to adopt and emphasize: three millennia of Chinese political history, interpreted through what I call the fragmentation pattern. This describes the situation in which a displaced regime establishes a rival power node that continues to claim sovereignty over the whole. It then either reconquers the center or is destroyed. The fact that this binary is false is irrelevant: It is the story that has been embraced, and cynically cultivated.
Beijing is finding that echoed civilizational trauma invoked in the name of regime stability is not easily discarded. The resulting ways in which the party is obliged to act are shaped by this narrative. For the United States, the consequence is significant: Deterrence built around imposing costs may fail if Beijing perceives inaction itself as a threat to regime survival. Understanding the historical template through which the Chinese Communist Party interprets Taiwan is therefore necessary to understanding what will deter Beijing and where U.S. strategy is most likely to miscalculate.
Victoria Tin-bor Hui’s comparative analysis of Chinese and European state formation argues that neither Chinese unification nor European fragmentation was structurally inevitable: Both were the products of strategic choices, particularly the self-strengthening reforms that allowed dominant states to absorb rivals. She argues specifically against geographic determinism in favor of political agency: The Chinese heartland was just as prone to political fragmentation as any other region. There was no foreordained pattern. China was never destined to be a centrally controlled single state. The fact that it happened repeatedly has no intrinsic value, but tremendous propaganda worth to anyone wishing to point to a tradition of political unity. The patterns of history may not themselves be deterministic, but they are politically useful, and the Chinese motif of unification has proved extraordinarily useful to the Chinese Communist Party.
Fragmentation and reunification have been the defining rhythm of Chinese political life, and a generation of Chinese citizens — including today’s cadres — have been schooled in it the way American officers are schooled in the lessons of Pearl Harbor and the Cuban Missile Crisis. On both counts the events themselves were real and documented, but American education focuses on the events as a product of their time and geopolitical circumstance, not a historical motif. Mark Edward Lewis’s histories of the early Chinese empires document the cycle from its origins: the Warring States period that preceded Qin unification in 221 B.C. and the collapse of the Han into the Three Kingdoms. The pattern repeated through the Sui, the Tang, the Song partition, and the Yuan Mongol conquest. The Ming expelled the Mongol Yuan, reunified China, then fell to the Manchu Qing. Southern Ming loyalist courts persisted for nearly two decades claiming continued legitimate authority over all of their former empire. The Qing collapse in 1912 produced warlordism, civil war, and left China vulnerable to Japanese invasion.
The lesson that the modern Chinese Communist Party has chosen to learn, and rely on, is this: When a Chinese dynasty falls, the displaced regime or its successors do not accept the loss and establish an independent state. They relocate, claim continued sovereignty over the whole, and persist as a standing challenge to the legitimacy of whoever holds the center. The Eastern Jin fled south after the fall of the Western Jin and ruled from Nanjing for over a century while non-Chinese dynasties held the north. The Southern Ming established rival courts after the Manchu conquest and fought on for nearly 20 years. The conclusion that can be drawn if only considering these cases is that the rival claimant represented an existential challenge to the legitimacy of the ruling dynasty, and the dynasty that tolerated the rival claimant’s existence was eventually judged to have failed the fundamental test of governance.
That test has a name that is so familiar it has become a cliché, which is precisely why its operational significance is underestimated: the Mandate of Heaven. The concept, which dates to the early Zhou dynasty, is not divine-right doctrine in the European sense. It is much better understood as a performance standard. Dingxin Zhao traces how the Western Han fused Confucian moral legitimation and Legalist administrative control, with the mandate as the test of that system: The ruler held it as long as he governed competently, maintained order, and preserved the unity of the realm. The mandate’s withdrawal is demonstrated by natural disasters, popular unrest, and above all by the fracture of political unity — when the ruler morally fails.
The existence of a rival claimant is, then, evidence that the mandate is in doubt and represents a direct refutation of the regime’s legitimacy. It is part of the evaluative framework through which political legitimacy has been assessed far beyond living memory, and it is the framework that Beijing has chosen to evaluate itself by. Zheng Wang’s research on Chinese political memory shows how the party maintains this framework through its schools. The patriotic education campaign launched after 1989 made the choice of which history to remember, putting the Century of Humiliation and the equation of national unity with national survival at the center of Chinese identity. Wang, borrowing the psychiatrist Vamik Volkan’s term, calls this a chosen trauma. The template is not merely inherited but actively reinforced by the state. This reinforcement only further constrains Beijing on the Taiwan question. To back down becomes a blow to legitimacy out of all proportion to its actual significance.
Taiwan fits the pattern Beijing has chosen to embrace with a precision that should alarm anyone responsible for deterrence planning in the Western Pacific. In 1949, the Chinese Nationalists under Chiang Kai-shek retreated to Taiwan after losing the civil war to Chairman Mao Zedong’s Communist forces. They did not establish an independent state. They established the Republic of China on Taiwan and maintained, for decades, that they were the legitimate government of all of China. The Nationalists occupied China’s seat at the United Nations until 1971. They maintained a posture of eventual reconquest well into the Cold War. The rival claim has softened over time as Taiwan democratized and developed its own identity, but the structural fact remains: A defeated regime retreated to a nearby position, claimed sovereign authority, and has persisted for 77 years under foreign protection.
All of this tells a very neat story, and it’s the story that the Chinese Communist Party has been telling for decades. But the Century of Humiliation narrative, stretching from the Opium Wars through the Japanese invasion to the Communist victory, is less historical analysis than political engineering. Rana Mitter shows how the narrative of Chinese experiences in World War II was deliberately constructed to prime the threat architecture of the Chinese population for a specific emotional payload: the memory of foreign exploitation and civilizational degradation, with the Chinese Communist Party positioned as the sole agent of restoration. General Secretary Xi Jinping’s “great rejuvenation of the Chinese nation” is both a slogan and a commitment — framed in the language of the mandate — that the Chinese Communist Party will complete the civilizational task that every previous dynasty was measured by. Taiwan’s continued separation is, within this framework, proof that the task is unfinished.
The Chinese Communist Party’s narrative seems to have survived its own historical counterexamples. The Northern Song Dynasty coexisted with its Liao neighbors after decades of warfare. Under the 1005 Treaty of Chanyuan, it accepted the Liao ruler as an equal emperor. The Eastern Jin, which the party’s reading treats as a rival claimant, opened a multi-state era: From A.D. 317 to 589, through the Northern and Southern Dynasties, China functioned as a system of rival states for over 250 years. The Yuan court, as the Northern Yuan, resisted every Ming attempt to extinguish it in Mongolian exile. These historical episodes seem to call into question the truism that a rival claimant continuing to maintain total sovereignty is fundamentally intolerable. Yet the Chinese Communist Party, having fully committed itself to the pattern, cannot accept the record of history it did not choose.
Treating rival claimants for sovereignty, breakaway territories, and every other flavor of territorial dispute as an existential problem is not a trait limited to the Chinese Communist Party. North and South Korea technically remain in a state of war and both governments treat that fact as a critical reality, yet peace has largely been maintained on the peninsula for seventy years. Time has decoupled reunification from regime survival. Russia has shown repeatedly its willingness to use military force against its neighbors for various reasons, all framed as existential. They are existential: just to Vladimir Putin, not to the Russian state. What sets apart the Chinese pattern is the deep pedigree of the falsehood. China can rightly claim to have a linguistic and civilizational record that stretches back millennia, and this is exactly what the Chinese Communist Party has tried to tap into. In doing so it has solidified its control over a staggering amount of territory and people, but it has also tied itself to an outdated template it found useful to cultivate and did so in such a way that it cannot be easily abandoned by the regime.
A choice hardens into a constraint once it is written into institutions, and this one has been written everywhere. The 2005 Anti-Secession Law put the claim into statute. The 2023 Patriotic Education Law makes reunification a mandatory subject of state education and calls completing it a “sacred duty” of all Chinese, Taiwanese included. A 2022 white paper declared reunification indispensable to national rejuvenation — Xi’s signature slogan. In 2013 Xi told a Taiwanese envoy the problem could not be passed down from one generation to the next, which constitutes a sharp break from Deng Xiaoping’s willingness to wait a thousand years. Each layer raises the price of walking away, however contingent the origins. After three decades of reinforcement, reunification is more tightly bound to regime survival in Beijing than when the campaign began.
Those bonds do, however, have some slack. Survey experiments by Jessica Chen Weiss and Allan Dafoe found that Chinese citizens punish leaders for empty threats, but approve of propaganda that frames present restraint as a path to future success. The party can sell patience at home far more easily than anything that looks like retreat under foreign pressure. North Korea shows the instrument can be set down and what the cost can look like. When Supreme Leader Kim Jong Un abandoned reunification as a goal in January 2024, he had to declare South Korea a “primary foe,” abolish the agencies built around reunification, and order the destruction of a monument his father built to it. Pyongyang could afford that because its legitimacy now rests on a dynasty, not on finishing a national task. Beijing has staked its legitimacy on rejuvenation, and its own white paper calls reunification indispensable. Setting the instrument down would mean admitting the slogan cannot be delivered. Any move by Beijing in Pyongyang’s direction would therefore show the party loosening its grip on the fear it cultivated. A leadership change would give us unparalleled insight into whether the embrace of the template is institutional or personal. A new general secretary of the Chinese Communist Party could radically alter China’s priorities, and such a change would allow a face-saving way to quietly abandon the template. If that — or anything like it — happens, then we will be privy to exactly what kind of template was in place. The real test will play out over years. Does the Chinese Communist Party’s rhetoric continue to emphasize reunification with Taiwan, or does it simply gradually disappear from party communications, never having been formally abandoned?
This is where wargaming reveals a gap in American planning. Most wargames or analyses modeling a Taiwan contingency assign Chinese decision-making to a rational actor framework: Beijing calculates costs, weighs probabilities, and acts when the expected value of military action exceeds the expected value of restraint. Adjudicators know the imbalance between modeling and actual decision-making well. The models are precise about capabilities and logistics — where data exist — but can be crude about how the adversary actually decides, where it matters most. Fidelity typically depends on the Red Cell leader’s intimate knowledge of Chinese political life. A game can report sortie rates to the decimal and still rest on a guess about what Beijing fears.
Stephen Downes-Martin has shown that standard adjudication methods break down on exactly these novel strategic problems, where data are thin. The most detailed public example, the Center for Strategic and International Studies’ wargame series on a Chinese invasion of Taiwan, is built on operations research rules and “most likely” base values. By design and definition, it was a rational actor model, and a rational actor model alone risks unsophistication in its treatment of the adversary’s threat perception. The center did not build its game to address this problem specifically, but any game constructed to gauge reaction and retaliation must consider it. Mirror imaging, in which planners assume their adversaries assess costs the same way they would, remains a recognized problem in how wargames are built.
As a wargame designer and adjudicator at the U.S. Naval War College, I treated it as a prudent working rule that omitting the fragmentation template carries risk: To be complete, games must incorporate it, even if only as an understanding among the adjudication cell. Beijing’s decision calculus must include not just the military cost of action but the civilizational cost of inaction. The relevant comparison is not “can we win,” but “can the party survive being the dynasty that lost the mandate.”
A game built this way should behave differently. When the adversarial cell is played by experts who understand modern China as a polity, Red would be harder to deter at lower rungs of escalation, more willing to absorb costs that rational actor models flag as prohibitive, and more sensitive to symbolic provocations that capability-focused models dismiss as strategically irrelevant. This is the behavior that open analysis of China’s Taiwan calculus already predicts. Some recent analyses read reunification as bound to regime legitimacy, and describe not calculated wars of choice, but wars of necessity: undertaken because the cost of inaction is regime survival itself. Beijing’s template does not make conflict inevitable. It makes the threshold lower and the triggers different than most American planners assume.
So what does any of this mean for actual strategy? If the Chinese Communist Party’s posture toward Taiwan is driven by rational cost-benefit calculation, then deterrence is a matter of raising costs until the expected value of military action falls below the expected value of the status quo. This is the logic that underlies most American deterrence planning, and it has a solid intellectual pedigree in the work of Thomas Schelling and the broader rational deterrence tradition. But if the Chinese Communist Party’s posture is driven by the activation of an outdated civilizational threat template that it chose for itself in the name of national unity, then the cost-benefit framework is incomplete. Rational actor theory still holds, but as the Chinese Communist Party applies it to itself, it must be understood that Beijing’s chosen threat template warps fragmentation into regime death.
Taken to its apparent logical extreme, treating China and the Chinese Communist Party with kid gloves because they have tied themselves to an idea and have a credible nuclear arsenal seems like a recipe for appeasement. The reality is that effective deterrence requires an accurate model of the adversary’s threat perception, and the model most American planners are using is incomplete. Deterrence that accounts for the real architecture would take seriously that the Chinese Communist Party’s leadership may accept costs that the current rational actor model classifies as irrational, because their chosen threat template classifies the costs of inaction as possible regime death. It would understand that public demands, conspicuous shows of force, and language that positions the United States as the latest in a line of foreign powers dictating terms to China do not deter. They escalate.
A harder version of the objection is that this rewards Beijing: If invoking trauma buys American restraint, Beijing has every reason to invoke it more. But what it buys is restraint on symbolism only. Arms deliveries, deployments, training, and the commitment itself stay where they are, so the invocation purchases nothing Beijing can fight with. Every invocation tightens the handcuffs. The louder Beijing ties its legitimacy to reunification, the costlier any visible retreat becomes and the less room it has to choose. Withholding pretexts while delivering capability leaves Beijing paying for its rhetoric and collecting nothing.
Washington has lately done the opposite. Asked in May whether he would approve a pending $14 billion arms package for Taiwan, President Donald Trump said he was holding it “in abeyance” and that it “depends on China,” adding: “It’s a very good negotiating chip for us, frankly.” Putting capability on the table while the summits continued is the one trade this argument rules out.
Depriving Beijing of rhetorical leverage means rethinking several elements of the current approach. It does not mean prostration: It means exploiting the handcuffs the Chinese Communist Party has placed itself in. First, in the signaling environment, the distinction that matters is between two kinds of statements. Statements about what the United States will do if force is used speak to Beijing’s cost calculation, and that is where deterrence lives. Statements about what Taiwan is, its status, and its place in a contest between democracy and autocracy speak to the template. Washington should make the first kind clearly and often, and the second as seldom as it can.
Congressional delegations to Taipei are designed to signal resolve to Beijing. In practice they hand Beijing a pretext. Then-Speaker Nancy Pelosi’s 2022 visit is the clearest case. Its lasting effect was to give Beijing the occasion to erase the median line of the Taiwan Strait: 302 People’s Liberation Army aircraft crossed it in August 2022 alone, and by 2025 crossings ran at about 1,500 a year and no longer made the news. Beijing may have wanted a pretext anyway, but the visit supplied it. Actions like this publicly replicate the pattern of foreign powers intervening in what the Chinese Communist Party must process as a threat to itself. A carrier transit of the Taiwan Strait or a quiet weapons sale signals something completely different, and it is a signal that can be understood by Beijing without requiring a high-profile response.
Time can be used to the advantage of the United States. The longer Taiwan remains an independent polity, the less existential reunification becomes, provided that the pattern is not constantly re-invoked. The United States cannot abandon Taiwan or cease military preparation, but it must distinguish between signals that deter by raising costs and signals that escalate by feeding the template and creating a public spectacle that the Chinese Communist Party must respond to or lose legitimacy. It must be disciplined about which forces are deployed and when. Quiet capability building deters by changing the military calculus without activating the civilizational narrative, even when the capability being built is devastating. There is a simple litmus test: Does the action change the military balance, or does it only make a statement to which the Chinese Communist Party must respond in escalatory terms? Beijing’s reactions can legitimately be considered as they bear on things that the United States intends to say. Beijing gets no veto over who goes to Taipei. Washington should send people when a visit gains something and hold them back when it would only hand Beijing a pretext. What the United States does is entirely its own business.
The main danger of this approach is under-signaling. If the United States only publicly announces things which meaningfully shift the military balance in the Pacific, there is a possibility that the Chinese Communist Party feels emboldened to test established boundaries. Fortunately, this is a problem the current situation is suited to solve. Beijing overreacts reliably to U.S. signaling: Indeed, that is the crux of this entire analysis. That means that any under-signaling can be easily fixed. Persistent under-signaling is rectified with one clear and targeted declaration, but the reverse is not true. Under-signaling is cheap to fix in terms of time, money, and personnel. Over-signaling is not.
None of this is Washington’s or Beijing’s alone to settle, and Taipei would not welcome some of it. Taiwan has changed: 62 percent of residents now identify as Taiwanese only, and under 3 percent as Chinese only. The rival claimant frame fits a Taipei that claimed all of China. It fits today’s Taipei poorly, and Beijing’s template has not adjusted. Taipei has also agreed to a status quo, its own: President Lai Ching-te pledged at his inauguration to “neither yield nor provoke” and to maintain it. What Taipei rejects is Beijing’s definition. The approach here asks Washington to defend Taipei’s version with capability and to stop arguing about either version in public.
The approach would cost something in Taipei, where American visits read as reassurance. A majority of Taiwanese welcomed Pelosi’s visit even after the exercises that followed — though most Kuomintang voters did not. Reassurance matters more now that only about a third of Taiwanese see the United States as a credible country. Washington can offset it by delivering capability. It has also asked restraint of Taipei before. In 2003 a Republican president publicly opposed President Chen Shui-bian’s referendum because Washington opposes unilateral changes to the status quo by either side. Taipei’s current government holds that Taiwan is already sovereign and needs no declaration, and it has largely kept to that line.
Second, the diplomatic vocabulary: Language that frames Taiwan as a pawn in great power competition — a strategic asset or a democratic beacon to be defended against authoritarian expansion — is accurate from Washington’s perspective, but unwise from the standpoint of template management. It confirms exactly what the Century of Humiliation narrative predicts: that foreign powers view Chinese territory as a playing piece in their own strategic games. Decoupling Taiwan’s security from the broader competition frame would reduce the template activation without reducing the commitment. Taipei, which leans on that framing to build support abroad, will not like this, but Washington can defend Taiwan’s democracy without making it a front line in its own contest with Beijing.
Third, the internal coherence of the American position: The current approach simultaneously tells Beijing that the United States does not support Taiwan independence, sells Taiwan advanced weapons systems, and sends senior officials to Taipei. From the current rational actor perspective these are calibrated signals of ambiguity. From the perspective of the Chinese Communist Party’s chosen template, they can be seen as incoherent provocations that confirm the party’s deepest fears about foreign interference. Strategic ambiguity works only if the adversary processes ambiguity the way you intend. If the adversary must publicly process it through a template that converts ambiguity into threat, the policy is escalatory, not ambiguous. A more coherent approach would align words and actions around a single message, a statement about an American redline rather than about Taiwan’s status: The United States, regardless of administration, takes no position on how the two sides settle their differences, only that they do so peacefully, and it will never accept a unilateral military alteration of the status quo. That was true when Chiang proposed — and fully planned — an operation to retake the mainland in the 1960s. It remains true now. The Eisenhower administration was not practicing appeasement when it wrote the constraint into a 1954 exchange of notes attached to the treaty, the Johnson administration was not practicing appeasement when it blocked the invasion’s launch in 1965, and no modern administration would be doing so now by enforcing the exact same principle.
The most stable outcome may require not the resolution of the Taiwan question, but the abandonment of the template the Chinese Communist Party adopted, cultivated, and bound itself to: maintaining a status quo that avoids activating the fragmentation trigger while preserving Taiwan’s functional autonomy. In the wargaming context, building the fragmentation template into adversary decision models is intended to produce higher quality games, not fatalism: games where Blue team strategies are stress-tested against an adversary that behaves the way rational actor theory predicts, when that theory is calibrated with as much fidelity as possible. The strategies that survive that stress test are the ones worth building policy around. The adversary’s threat map is legible to anyone willing to both read three millennia of evidence and understand how it has been selectively applied. The question is whether American strategy is built on what the evidence actually says.
Charles Edward Gehrke is the former deputy division director for wargame design and adjudication at the U.S. Naval War College. He holds a B.S. in Quantitative Economics from the U.S. Naval Academy and an M.A. in National Defense and Strategic Studies from the U.S. Naval War College. He is now a Sloan Fellow at the Massachusetts Institute of Technology.
The views expressed in this article are the author’s own and do not represent those of the U.S. Naval War College, the Department of the Navy, the Department of Defense, any part of the U.S. government, or the Massachusetts Institute of Technology.
Image: Japanese Ministry of Defense via Wikimedia Commons
