‘counterparts are even more probably corporeal; if 0, you are essentially corporeal ‘An attribute that something shares with all ts ‘counterparts isan essential attribute ofthat thing, part oF ts essence. The whole ofits essence is the intersection of its essential atribute, the attribute it shares with all and only its counterparts, (The attribute, because there is no need to distingyish aruibutes that are coextensive not only in the actual world but aso in every possible work) ‘There may or may not be an open sentence that expresses the aribute that isthe essence of rome- thing; to assert thatthe ateibute expressed by ¢ is the essence ofthe thing denoted by {is to assert Fa( feo & VO(ISO >. YS = & BY Y(TyB >.Cya = $y) (The acwal world contains unique uch that, 9x; and for anything 7 in any world fi is a counterpart of ifand only i€ @)) ) ‘This sentence is not equivalent to the translation ‘of any modal sentence, Essence and counterpart are interdefinable. We Ihave just defined the essence of something as the auribute it shares with all and only its counter- parts; a counterpart of something is anything hay ing the atribute whichis its essence (This is not two say that that attribute is the counterparts essence, or even an essential attribute ofthe coun terpart) Perhaps there are certain aibutes that can only be essential attributes of things, never accidents Pechaps every human must be essentially human; ‘more likely, peshaps everything corporeal must be essentially corporeal. The attribute expressed by is ofthis sor, incapable of being an accident, just in ease it is closed under the counterpart relation; ‘that is, justin case VANE Uap ke IP, ke Cyn te Gfa.3 fy) (or any counterpart yn any word of any- thing ain any world, if @2a then $y) ‘Thisia simplified equivalent of the translation of Eva(62 3 048) ‘We might wonder whether these attributes incap~ able of being accidents are what we call “natural kinds.” Bur notice frst that we must disregard the necessarily universal attribute, expressed, for Counterpart Theory and Quantified Modal Logic instance, by the open sentence =a, since itis an essential attribute of everything. And notice second that arbitrary unions of atibutes incapable ‘of being accidents sre themselves aceibutes incap- able of being accidents; so to exclude gerryman- ders ve must confine ourselves 19 minal axtributes incapable of being accidents. All of these may indeed be natural kinds; but these can not be the only natural kinds, since some unions and all intersections of natural Kinds are them= selves natural nds IV. Modal Principles ‘Translation into counterpart theory can sete dis puted questions in quantified modal logic. We can testa suggested modal principle by secing whether its translation is «theorem of counterpart theory; or, if not, whether the extra postulates that would ‘make ita theorem are phusible. We shall consider ight principles and find only one that should be aecepted 6 < C09 (Beckers principte) ‘The translation is not 2 theorem unless ¢ is a closed sentence, but would have been a theorem in general under the rejected postulate that the ‘counterpare relation was transitive, <10d (Brouwer's principle) ‘The translation is not a theorem unless @ is a closed sentence, but would have been a theorem jn general under the rejected postulae that the counterpart relation was symmetric. my 0 <0 =a (and anot the same variable} “The translation isnot a theorem but woul have been under the rejected postulate tht nothing i any world had more thn one counterpart in any other world adm < Om Am (oy and agnot the sxe variable) “The translation is not theorem, but would have Deen under the rejected postulate that m0 two ‘things in any world had a commen counterpart i any other world aD