In this dissertation I argue that universal attributes are the metaphysical ground of similarity, and that the ultimate reason embracing realism is that an explanation of similarity must posit the existence of universals. Other arguments for the existence of universals are ultimately motivated by the desire to explain phenomena, such as laws of nature, general predication, and general knowledge, that seem to depend on similarity. ;This work is structured on metaphilosophical principles of Lawren…
Read moreIn this dissertation I argue that universal attributes are the metaphysical ground of similarity, and that the ultimate reason embracing realism is that an explanation of similarity must posit the existence of universals. Other arguments for the existence of universals are ultimately motivated by the desire to explain phenomena, such as laws of nature, general predication, and general knowledge, that seem to depend on similarity. ;This work is structured on metaphilosophical principles of Lawrence Lombard and Lawrence Powers. Within this framework I defend principles for assigning weights to theoretical values such as ontological economy, explanatory power, and accordance with commonsense beliefs, and develop a realist semantic position that I use to interpret and defend an argument for the existence of universals from premises that include a simple similarity statement. ;Attributions of similarity to objects are often made with respect to kind-determining attributes or with respect to determinable attributes. I provide a realist se mantic account of such attributions, including an account of attributions of degrees of similarity. While our ways of speaking and thinking about similarities can be explained using a realist approach, nominalist theories cannot explain this central aspect of our experience. Also, nominalist theories require sweeping rejections of commonsense claims. I argue that the ontological economy of the most plausible nominalist theories is not as great as is often supposed. Thus, if we can make sense of the concept of a universal, we should accept realism. ;Because the role of universals as the ground of similarity is the ultimate motivation for positing the existence of universals, the concept of a universal is essentially associated with that role. I support this view of universals by defending a criterion of identity for universals formulated in terms of their role in grounding similarities, and disproving alternative realist criteria of identity for universals. For example, I show that Armstrong is unsuccessful in his attempt to base a criterion of identity for universals on the causal roles of universals. By providing a criterion of identity of universals, in terms of their role in grounding similarities, the essential nature of universals is demonstrated.