著者

横路 佳幸

出版者

日本哲学会

雑誌

哲学 (ISSN:03873358)

巻号頁・発行日

vol.2018, no.69, pp.259-273, 2018

---

<p>The Principle of the Identity of Indiscernibles (hereafter the PII) states that ifany individuals exactly resemble each other, then they are necessarily identical. Intuitively,the PII seems valid, but Max Black attempted to refute it by introducingthe possibility of a symmetry universe in which two iron spheres <i>c</i> and <i>p</i> can resembleeach other exactly. This counterexample (hereafter BU) seems easy to ruleout using a weak discernibility strategy (hereafter WD) according to which <i>c</i>, beingspatially separate from <i>p</i> and not from <i>c</i> itself, is not indiscernible from <i>p</i>. WD, however,leads to 'the presupposition problem', because obtaining <i>c</i> as spatially separatefrom <i>p</i> presupposes the distinctness of <i>c</i> and <i>p</i>. In this discussion, I will give an outlineof a defense of the validity of the PII that evades the presupposition problemthrough the elucidation of some aspects of 'identity'.</p><p>In my view, 'identity' has two aspects: one is simply self-identity as a universalmonadic property (hereafter identity-1), and the other is identity as an equivalencerelation entailing indiscernibility (hereafter identity-2). The basis or ground for identity-1obtaining with regard to an individual <i>x</i> can be called the individuator for <i>x</i>,but it is no wonder that the individuation and articulation of <i>c</i> and <i>p</i> are prior to orground for obtaining <i>c</i> as spatially separate from <i>p</i>. So far as the PII is concernedwith identity-1, it may not be valid. However, we can characterize identity-2, followingDavid Wiggins's lead, in terms of what is called the sortal dependency of identity-2and the extended Locke's Principle (hereafter ELP), according to which, for anysortal concept <i>F</i>, <i>x</i> falling under <i>F</i> is identical with <i>y</i> falling under <i>F</i> if and only if <i>x</i>is the same <i>F</i> as <i>y</i>, and <i>x</i> is the same <i>F</i> as <i>y</i> if and only if a) <i>x</i> and <i>y</i> share <i>F</i> and b) <i>x</i>is not spatially separate from <i>y</i>. If ELP is valid, we can regard BU as merely a generalcase to which WD is applied. And if the Wigginsian idea of the sortal dependencyof identity-2 is also right, there is no longer a presupposition problem. I hence conclude that the PII is valid to the extent that it is concerned with identity-2.</p>