【作者】 汪伟民；

【导师】 石源华；

【作者基本信息】 复旦大学 ， 国际政治， 2005， 博士

【摘要】 自人类发生战争与冲突以来,“联盟”(又称“同盟”,alliance or alignment)便成为国家或其它政治行为体对付威胁,求取安全,增强权力,维护或扩展其利益的最常用、乃至最有效的手段或方法之一,联盟政治、联盟战略也由此成为国际政治和国际安全研究的核心问题之一。然而,由于联盟行为同时兼具国际政治的冲突与合作的双重特性,因此联盟作为一种抽象理论的研究一直缺乏明晰的研究本体和独特的研究范畴,致使联盟理论的研究始终难以走到国际关系理论研究的前沿。 目前美欧学者主要在联盟起源的研究方面取得了较大的突破和较多的研究成果,而联盟理论流派的分野也大致以不同的起源论来进行界定的。传统现实主义的联盟理论将权力作为主要的分析变量,其理论基点是“权力平衡论”,认为“联盟是权力的增强器”,是权力平衡的工具;新现实主义则在对传统现实主义的联盟理论进行优化的基础上,分别以“威胁”和“利益”作为主变量,先后提出了“威胁平衡论”和“利益平衡论”,并使其联盟理论更具系统性和实证性。近十多年来,联盟理论的研究出现了一个不大不小的高潮,并在上个世纪90年代引起了激烈的学术论争。与此相伴,新的研究成果和范式也不断涌现,一方面现实主义的理论阵营不断推陈出新,如提出了“联盟困境理论”、“制衡与追随理论”、“囚徒捆绑与推卸责任理论”等诸多新的专门理论,另一方面制度主义者和建构主义者在指出现实主义联盟理论范式解释力有重大缺失的同时,力图从“制度”和“认同”的角度对联盟的起源与护持作出新的解读。可惜的是,后起的“制度规范论”者和“观念认同论”者都未能建构起完整的联盟理论体系,它们所采取的单一变量的分析方法似乎只能为居于主导地位的现实主义的联盟理论范式提供多样的分析视角而已。 本文试图立足新现实主义和新古典现实主义的基本分析框架,着重借鉴兰德尔·施韦勒“利益平衡论”的分析体系,尝试提出一种全新的联盟理论,即“威权式联盟”(authoritative alliance)理论。这一理论事实上并非一种一般意义上的联盟理论,而是针对某种类型或性质的联盟而展开的理论分析。这一理论分析既揉合了传统联盟理论分析通常所使用的某些变量,如权力、威胁、利益,也引入了制度、认同、恐惧等新的分析变量,并将这一联盟运行的逻辑或理论基点假定为“利益平衡”和“安全平衡”。具体而言,威权式联盟理论主要由互为一体的三个主要的理论假设所构成。假设一:在冷战后的国际体系中真实地存在着一个史无前例的一元化的“威权式联盟”体系,它使其它旧式的或零散的联盟难以取得足够的生存空间,事实上,目前几乎所有国家都在寻求与更多还原

【Abstract】 Since war and conflict came into being, alliances (or alignments) have been perceived as one of the most frequently-used or the most efficient instruments to fight against threats, obtain security, aggregate power, safeguard or enlarge national interests. Thus alliance politics and alliance strategy becomes one of the core themes of international politics and security studies. Yet, as a kind of abstract theory, the studies of alliances always lack for obvious identity and peculiar research realm, which makes it rather difficult to come to the forefront of the theoretical studies of international relations.So far the U.S. and European scholars’ academic achievements or breakthroughs are mainly concentrated on the exploration of alliance origins which roughly determined different schools of alliance theories. The classical realist alliance theory chooses power as the main analytical variable and assumes balance-of-power as its basic theoretical proposition and supposes that alliances are tools to aggregate power in order to achieve the balance of power. Through refining the classical realist alliance theory, the neo-realist alliance theories pick up threat or interests as their respective analytical variable, thereby constituted the balance-of-threat and balance-of-interests theories, both of which make the alliance theories more systemic and verifiable. In the past decade, a modest invigoration occurred in the studies of alliance theories and 90’s of last century witnessed a round of intense academic argumentation by which new paradigms and academic achievements burst into existence. On one hand, the realist camp weeded through the old to bring forth the new, some partial theories such as "alliance dilemma" theory, "balancing and bandwagoning" theory, "chain-ganging and buck-passing" theory were put forward one by one. On the other hand, the institutionalist and constitutionist attempted to provide new interpretations to the origins and preservations of alliances from the perspectives of institution and identity, while they pointed out sharply the serious shortcomings and disadvantages in the realist paradigms because of their lack of interpretation force. But it is really a pity that as later comers, the institutionalist and constitutionist failed to constitute any systemic alliance theories. It seems that their single-variable analytical patterns perhaps can only supplement the dominant realist alliance paradigm with dynamic perspectives.By employing the neo-realist and neo-classical realist paradigm, and mainly drawing on Randall Schweller’s analytical framework, this dissertation attempts to put forth a brand-new alliance theory, i.e. authoritative alliance theory. This theory is actually not an alliance theory in general meaning, it is merely a theoretical analysis concerning a particular type of alliance or an alliance of peculiar nature. This kind of theoretical analysis absorbs those often used variables, such as power, threat and interests while pays more attentions to some new variables, for example,更多还原