Introduction
Whether Japan could lawfully join in the defense of Taiwan is a question that has drawn renewed attention. In November 2025, Prime Minister Sanae Takaichi told the Diet that a Taiwan contingency involving warships and the use of force could, in her words, “by any reckoning,” amount to a “survival-threatening situation” (sonritsu kiki jitai)—the trigger under the 2015 peace and security legislation for Japan’s limited exercise of the right of collective self-defense.[i] The statutory formula requires “an armed attack against a foreign country that is in a close relationship with Japan” which in turn threatens Japan’s survival. Two different legal questions hide inside that one formula, depending on which foreign country is attacked. If it is the United States—whose forces might intervene in a Taiwan contingency and come under Chinese fire—the statehood question presents no special difficulty; the United States is unquestionably a State, though whether Japan would act remains a separate political judgment. If the victim is Taiwan itself, the analysis changes. The International Court of Justice’s 1986 Nicaragua judgment confirmed that collective self-defense is a right exercised for the benefit of a victim State, one that must itself declare that it has been attacked and request assistance.[ii] Taiwan’s own statehood is precisely what is in doubt. This essay is about that harder question—not whether Japan can defend American forces coming to Taiwan’s aid, but whether Japan can defend Taiwan directly.
Taiwan, as matters stand, cannot readily be evaluated as a State. Effectiveness is not seriously in question within the area Taiwan actually governs; it has a population of some twenty-three million, a defined territory—Taiwan, Penghu, Kinmen, and Matsu—under its unbroken control since 1955, and a government of undoubted capacity. The difficulty, as I have argued elsewhere,[iii] lies instead in the mismatch between that effective control and the Republic of China’s own official claim, which nominally still extends to the Chinese mainland. Taiwan’s constitutional order remains that of the Republic of China. The Additional Articles, in force since 1991, are framed “to meet the requisites of the Nation prior to national unification” and confine themselves to regulating relations between a “free area” and a “mainland area,” without redefining the Republic’s territory as Taiwan alone. Individual leaders have gone further politically than legally—Lee Teng-hui called cross-strait relations a “special state-to-state relationship” in 1999, Chen Shui-bian called Taiwan and China “one country on each side” in 2002, Lai Ching-te insisted in 2025 that Taiwan is “of course” a country. Lee’s and Chen’s statements were each met within days by an official clarification that they changed nothing; Lai’s does not appear to have been withdrawn, but neither has it been entrenched in any constitutional or legislative act. And the closest occasion on which voters were asked to endorse an international-status claim under the name “Taiwan,” a 2008 referendum on U.N. membership, carried 94 percent of the vote cast but failed Taiwan’s own fifty percent turnout requirement and had no legal effect—although abstention does not necessarily imply opposition, and the question itself fell short of a vote on independence. Because recognition is declaratory rather than constitutive, other States’ treatment of Taiwan is relevant evidence but not dispositive. What also matters here is whether Taiwan’s own competent organs have settled on an unequivocal claim reflecting the will of its people and the status quo. On the constitutional record, for now, they have not.
If Taiwan is not a State, a Japanese use of force in its direct defense cannot straightforwardly be justified as collective self-defense in the ordinary Article 51 sense.
Yet the Japanese government has itself suggested an opening in its own doctrine. During the 2015 Diet deliberations, then-Foreign Minister Fumio Kishida was asked whether the existence of diplomatic relations was relevant to identifying a “foreign country in a close relationship,” and answered that since the right of self-defense is exercised by States, the country concerned must be a State—but that, as the government had traditionally explained, “unrecognized states” (mishōnin-koku) and “divided states” (bunretsu kokka) are included.[iv] This was a general proposition, not a claim about Taiwan specifically. Pressed directly on whether Taiwan qualified, Kishida declined to answer, citing Japan’s renunciation of all right, title, and claim to Taiwan under the San Francisco Peace Treaty.[v] If sound, that proposition points to the one route by which Japan might still defend Taiwan directly without conventional statehood: showing both that Taiwan qualifies as a “divided state” and that divided-state status carries some protection under the law on the use of force. Yet neither half of that claim survives testing comfortably.
[i] Statement of Prime Minister Sanae Takaichi, House of Representatives, Budget Committee, Nov. 7, 2025.
[ii] Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States of America), Merits, Judgment, I.C.J. Reports 1986, p. 14, paras. 195, 199.
[iii] Masahiro Kurosaki, “Reformulating Taiwan’s Statehood Claim,” Lawfare, https://www.lawfaremedia.org/article/reformulating-taiwan-s-statehood-claim.
[iv] Statement of Foreign Minister Fumio Kishida, House of Representatives, Special Committee on the Peace and Security Legislation, June 15, 2015.
[v] Kishida, House of Representatives, Special Committee on the Peace and Security Legislation, June 15, 2015.
I. The Diet Formula and Its Cold War Origins
Kishida described the inclusion of unrecognized and divided states as something the government had “traditionally explained.” The “close relationship” concept is a product of that same 2015 legislation, so the exact phrase has no earlier history; the term “divided state” does not share that limitation. What the record does disclose is the doctrine’s substance. During the Vietnam War, the Foreign Ministry’s treaty bureau chief told the Diet that crossing an internationally established line partitioning a divided state would amount to aggression and, in the same breath, to an armed attack, against which the attacked side would naturally hold a right of self-defense—a principle, he added, already settled beyond doubt since the Korean War.[ⅵ] The paradigm is North Korea; never recognized by Japan as an independent State, yet—a United Nations member since 1991—a State capable of committing an armed attack within the meaning of Article 51.
What this history shows is limited: the absence of Japanese recognition or diplomatic relations does not remove an entity from the class of States for the use of force. But that understanding formed in a world where both halves of every divided nation had come to satisfy the criteria of statehood. Whether the formula can do any work for an entity that has not—Taiwan’s situation—depends on what “divided state” actually means, and why it was ever thought to matter for the use of force.
[ⅵ] Statement of Banri Fujisaki, Director-General of the Treaty Bureau, Ministry of Foreign Affairs, House of Representatives, Budget Committee, July 18, 1966.
II. What Is a “Divided State,” and Could Taiwan Be One?
The term is descriptive before it is legal. It entered the vocabulary after the Second World War to capture the parallel predicaments of Germany, Korea, Vietnam, and China: single nations within which rival authorities emerged, each claiming to be the sole legitimate government of the whole. A divided state, by the standard definition, exists where the partitioning line cannot be regarded as a frontier in the legal sense.[ⅶ] Yet the prevailing doctrinal verdict is deflationary: there is no separate juridical category of the “divided state.” Whether the components of a divided nation constitute parts of a single State or two States “depends on general criteria for statehood—not on some postulated category of ‘divided States’ into which a variety of cases are supposed to fit.”[ⅷ] The case histories of Germany, Korea, and Vietnam bear this out. In each case, crucially, both halves could be characterized as independent States by ordinary criteria—before self-determination had crystallized as a cardinal principle, when a people’s own claim to separate statehood was not itself decisive—notwithstanding each side’s own un-renounced claim to the whole nation. This is why Crawford calls the “divided state” label legally idle; it added nothing that statehood criteria did not already supply.
Taiwan’s situation, on the analysis above, is harder to assimilate to the classic cases. Although it satisfies the effectiveness criteria, it has not unequivocally claimed separate statehood within the territory it actually governs—itself a requirement of contemporary self-determination doctrine, not of any recognition-based theory. Whether Taiwan can even be called a divided state is itself disputed, and scholars divide on the question. A further, structural difference reinforces the point: the Taiwan Strait has been stabilized since the mid-1950s by deterrence and geography, not by an agreed demarcation line—its median line was drawn unilaterally by the United States in 1955, and Beijing has never accepted it as binding. Further, there has been no mutual accommodation of the sort that matured the German and Korean divisions into settled fact—no equivalent of the Basic Treaty, no simultaneous U.N. admission. Under the loose, traditional understanding grouping these four cases together, Taiwan can plausibly be counted among them; under any more demanding test, it cannot confidently be. That looser label carries no legal significance here. What matters is the one feature Taiwan lacks—an internationally established boundary—on which the use-of-force theory examined next depends.
[ⅶ] James Crawford, The Creation of States in International Law (2nd ed., Oxford: Oxford University Press, 2006), p. 450.
[ⅷ] Crawford, The Creation of States, p. 451.
III. Why Might Divided-State Status Matter for the Use of Force? The Korean Precedent
The clue lies in Korea in 1950, not in any definition of “divided state.” When North Korean forces crossed the 38th parallel on June 25, 1950, the U.N. Security Council did not pause to resolve whether the Democratic People’s Republic of Korea was a State, or whether the Republic of Korea was, or whether the parallel was an international boundary in the strict sense—all genuinely unsettled at that date. Noting “with grave concern the armed attack on the Republic of Korea by forces from North Korea,” U.N. Security Council Resolution 82 determined that this action “constitutes a breach of the peace”—not quite “aggression” in the Charter’s terms. Two days later, U.N. Security Council Resolution 83 recommended that member States “furnish such assistance to the Republic of Korea as may be necessary to repel the armed attack and to restore international peace and security in the area” (italics added in both quotations). Twenty years later, the General Assembly’s Friendly Relations Declaration articulated a broader principle to similar effect: every State must refrain from the threat or use of force “to violate international lines of demarcation, such as armistice lines, established by or pursuant to an international agreement to which it is a party or which it is otherwise bound to respect,” even where the line falls short of a full international frontier. An inter-State boundary is one limit on the use of force, Crawford notes, “but, as resolution 2625 (XXV) recognizes, there can be other such boundaries.”[ⅸ]
Put together, Korea 1950 and the Friendly Relations Declaration suggest a theory: that crossing an established demarcation line separating the two halves of a divided nation can itself be an unlawful use of force of the gravity the ICJ’s Nicaragua judgment later required for an armed attack—one of “the most grave forms of the use of force”—against which the attacked side, and those coming to its aid, might in principle act in self-defense, without first resolving every question about the attacked side’s formal statehood. On that theory, Japan’s collective self-defense would rest not on Taiwan’s statehood but on the protected status of the line and the armed attack its crossing would constitute. Whether that also supplies the victim-State standing Article 51 requires is the difficulty addressed next.
[ⅸ] Crawford, The Creation of States, p. 450.
IV. Why the Theory Is Too Fragile to Rely On
Attractive as it is, the theory is not itself the holding of any single authority: it is an inference drawn by combining two sources, neither of which, on its own terms, supplies it. Korea 1950 shows that the Security Council can recommend collective assistance to a divided state’s half without resolving its statehood—but the Council did not invoke Article 51 in doing so, only language the United Nations’ own repertory of practice describes as merely having “some bearing on” it.[ⅹ] The Friendly Relations Declaration’s demarcation-line principle, for its part, obliges States not to use force against protected lines but says nothing about who may invoke Article 51 self-defense in response to a violation, and in any event confines the protected lines to those established by an agreement binding the State concerned—a description the Taiwan Strait, as noted above, does not meet. Vietnam shows why that last condition is no formality—and shows it using the very line Japan’s own Foreign Ministry cited in 1966 as the doctrine’s guiding example: the 1954 Geneva Final Declaration provided that the 17th parallel “should not… be interpreted as constituting a political or territorial boundary,” yet the United States invoked collective self-defense under Article 51 to defend South Vietnam against forces crossing it—a controversial claim made independently of Security Council collective-security action. The Taiwan Strait, lacking even Vietnam’s degree of formal agreement, meets the Friendly Relations Declaration’s condition even less.
On the current record, neither route is available: Taiwan has made no independent and unequivocal claim to statehood, and, as just shown, the divided-state theory does not reach the Taiwan Strait either. It is accordingly difficult to identify a solid legal basis on which Japan could defend Taiwan directly, as distinct from defending American forces that come to its aid.
[ⅹ] United Nations, Repertory of Practice of United Nations Organs, vol. 2 (1945–1954), Article 51, p. 430 n. 5, https://legal.un.org/repertory/art51/english/rep_orig_vol2_art51.pdf.
Conclusion
Rightly understood, the divided state is not a shortcut to Japan’s direct defense of Taiwan; tracing its logic to the source shows how much still depends on the requirement of statehood. Japan’s Diet formula—that unrecognized and divided states may be objects of collective self-defense—was offered as a general proposition, and this essay has found no reason to think it wrong. The possibility of a lawful Japanese defense of a divided-state Taiwan is not zero, and the theory could, in principle, be extended to Taiwan’s case. All that said, that theory asks more of its foundations than they supply: the Friendly Relations Declaration’s principle does not extend to the Taiwan Strait, where no relevant line has been established by or pursuant to an international agreement binding the PRC. Taiwan’s possible status as a divided state cannot, on the current state of international law, by itself supply a sufficiently secure legal basis for direct Japanese defense of Taiwan. Short of an independent and unequivocal Taiwanese claim to statehood, then, Japan’s soundest route to involvement in a Taiwan contingency remains the indirect one: defending American forces that come to Taiwan’s assistance, rather than defending Taiwan directly on the divided-state theory examined here. That indirect route, however, necessarily assumes that the United States’ intervention is lawful and that an attack on its forces constitutes an armed attack giving rise to its own right of self-defense.
Views expressed or implied in this article are solely those of the author. They must not be construed as representing the views and positions of RIPS or any other organizations.