Universals: An Opinionated Introduction

The problem of universals is one of the most longstanding philosophical problems. The problem is about what makes a class of tokens mark off a type. For example, the red apple and the red fire engine being both red are of the same type. But does this sameness involve a strict identity? Is there some one thing involved in the apple and the fire engine that makes them both red? Realists about universals answer that the very same thing is present in the cases of tokens of the same type; for example, the apple and the fire engine have an identical property of redness in common. Nominalists reject such a postulation as uneconomical and problematic; if the very same redness is present in the apple and the fire engine, then the same thing is at two possibly very distant places at once! In place of such universals, Nominalists introduce such things as primitive natural classes, primitive resemblances, and tropes in order to answer the problem. David Armstrong is one of the most influential philosophers in the debate about universals, and in this book he introduces the problem of universals and covers at least six proposed solutions. Each solution is presented and evaluated carefully. As the subtitle suggests the introduction is opinionated, and Armstrong ultimately favours a moderate form of Realism. However, he also candidly presents the advantages of alternative solutions and the disadvantages of his own. The related topic of the nature of particulars is also covered, and Armstrong introduces the proposal that particular things are mere bundles of properties and the rival proposal that they are substances instantiating properties. Crucial notions such as those of thin and thick particulars and states of affairs are also explained. The book is well organized, and generally written clearly. It is not long, but is nevertheless comprehensive without sacrificing depth. Helpful references and recommended reading for each proposal are included. Students unfamiliar with analytic philosophy might find it quite difficult because there is some terminology that is not explained, and the discussion is quick at some points. The book is intended primarily for advanced undergraduate and graduate students in philosophy, but may also interest professional philosophers. The book is recommended especially for advanced students interested in the problem of universals in particular. For a more general introduction to metaphysics, I recommend Jonathan Lowe's "A Survey of Metaphysics".

This is a very manageable and, as has been noted already in reviews here, very readable, relatively slim introduction to universals. Armstrong presents his own and others' views. He considers strengths and weaknesses of the various views. A good place to start reading about universals or learn one person's take on the subject. Bc it is an introduction, interested readers will want to continue with other, more in-depth books, including those suggested by other reviewers. Readers may also want to read some of the philosophers Armstrong mentions in the book.

If, like many people, your first introduction to philosophy came through the writings of Ayn Rand, you might have a significant amount of "unlearning" to do. Although Rand had some interesting things to say (generally limited to culture and politics), her writings in purely technical areas of philosophy are highly problematic and at times grossly ill informed. Her discussion of the "problem of universals" is paradigmatic.The problem of universals is, contrary to Rand, a problem of metaphysics, not epistemology. Why's that? Because it concerns the nature of reality and whether or not there exists something beyond "particulars," something which "instantiates" these particulars. For example, there are many objects (particulars)that are red. But where (if anywhere) is "redness" (a universal)? That's a metaphysical, not an epistemological question. And, contrary to Rand, there are only two theories of universals: nominalism and realism. (Alternative theories, such as conceptualism, are versions of nominalism, generally speaking.) Rand's theory, it turns out, is best seen as a version of moderate nominalism.I don't know of an introductory work on this subject that is suited for someone without much knowledge of philosophy. However, D.M. Armstrong's UNIVERSALS: AN OPINIONATED INTRODUCTION is a good place to start if you have some background in philosophy (or are willing to work a little harder than usual).I've also purchased J.P. Moreland's UNIVERSALS which is even more advanced and takes issue with Armstrong on certain points.

Armstrong's book is what it promises: an opinionated introduction to the problem of universals, which in the most general terms is the problem of how to understand philosophically the phenomenon of similarity or resemblance between different things. Resemblance between things is the basis of classification and generalization, and it is presumably by noticing similarities among numerically distinct things that we form many abstract concepts and general terms. A realist about universals holds that similar things get to be similar by having some entity in common--repeatable features that are present identically in all of them, either as a constituent of the particular or as "instantiated" or "exemplified" by it in some mysterious way, without being in it. These repeatable features he calls "universals"--general things, as it were, to correspond to general terms, as names of particulars correspond to them. A nominalist denies that there are universals, and seeks to account for similarity without positing them. Armstrong spends most of the book comparing the relative merits of five different metaphysical accounts of similarity: natural class nominalism, resemblance nominalism, a realist "bundle theory" of particulars, a realist substance-attribute theory (his own favorite), and finally theories of tropes (that is, particular or non-repeatable properties and relations). Natural class nominalists take the notion of a "natural class", that is, a set of noticeably similar things, as primitive or undefined, and seek to account for similarity in terms of membership in a natural class. Resemblance nominalists take similarity among particulars as primitive instead, and seek to account for similarity among a class of things by way of resemblance to paradigm cases of such things. Both approaches suffer from serious defects, as Armstrong points out, not the least of which is that these are very awkward primitives to have. Next comes the extreme realist, who thinks universals are all there are, and that particulars are nothing but bundles of universals. This theory also has its share of difficulties, not the least of which is that it seems to be unable to explain what the bundling relation is. Next there is the substance-attribute realist, in which camp Armstrong belongs. Such a realist faces the problem of analyzing similarity in terms of identity of universals. Among particulars this is straightforward; the more universals they have in common, the more similar they are. But what about the universals themselves? Isn't the color red, say, more like orange than it is like green? Armstrong says he is working on this problem, but gives little indication how to solve it. Finally, there is the doctrine of tropes, that is, particular properties and relations which are in themselves simple, the doctrine I favor. The trope nominalist can readily substitute for universals equivalence classes of exactly resembling tropes. (No problems analyzing resemblan