In this paper I propose a solution to the qualitative version of David Miller's verisimilitude reversal argument. Miller (1974) shows that verisimilitude rankings are relative to language choice and hence, are not objective. My solution stems from a reply to an earlier solution proposed by Eric Barnes (1991). Barnes argues that the verisimilitude reversal problem can be solved by revealing an epistemic dimension. I show that Miller's problem cannot be solved by side-stepping foundational metaphysical claims as his epistemic solution suggests. Rather, a substantive metaphysical basis grounds identity relations among properties. The problem of verisimilitude cannot be solved without embracing the fundamental metaphysical distinctions between basic and composite properties that ground the relationship of partial identity among properties.