Nominalism

2. The medieval period

The orientation we call nominalism is typically traced back to medieval debates over universals. A major source of these debates was Boethius’ commentary on Porphyry’s Isagoge, where we find a detailed discussion of the ontological status of universals (see Boethius, A.M.S.; Porphyry §1). Boethius’ commentary became a pivotal text for medieval philosophers, and by the twelfth century the debate over universals had become a dominant topic of philosophical concern. Two opposing views emerged. One, championed by William of Champeaux, was an extreme form of realism. On this view, a genus or a species is literally the same in all its members; the individuals falling under the kind are rendered distinct by the addition of forms to the common essence, and those forms are predicates of the essence. The opposing view, defended by Roscelin of Compiègne, represented an extreme version of nominalism. Insisting that all entities are particulars, Roscelin argued that talk of universals is merely talk about linguistic expressions that can be applied to a number of different particulars, and he held to an austere interpretation of linguistic expressions where universals are mere flatus vocis or vocalizations.

Abelard attacked both views (see Abelard, P. §5). He argued that since the forms which allegedly diversify any common essence are contraries, William’s realism commits us to the view that a single entity can simultaneously exhibit incompatible properties; and he appealed to Aristotle’s definition of the universal as that which can be predicated of many to call into question the claim that nonlinguistic entities can be universals. That definition, he contended, applies exclusively to things that can function as predicates in subject-predicate sentences, to what we nowadays call general terms. But while agreeing with Roscelin that we can ‘ascribe universality to words alone’ Abelard rejected Rocelin’s claim that universals are mere flatus vocis, insisting that they are expressions that are significative or meaningful; and he argued that any adequate account of universals must show how, in the absence of a common essence, general terms can be meaningful. In doing so, it must answer two questions: (1) what is the cause of the imposition of a common name; and (2) what is it that we grasp when we grasp the signification of a common name?

Abelard’s answer to the first question is disarming. The things to which the term ‘man’ applies, for example, all agree in being men. Their being men is the ground of the imposition of the common name ‘man’. Abelard denies, however, that this agreement involves any common entity. He takes it to be an irreducibly primitive fact that all the things called men agree in that they are all men. In response to the second question, Abelard argues that what we grasp when we understand the common name ‘man’ is not any of the particular men named by that term; nor is it the collection consisting of all those particulars. To explain the kind of cognition associated with general terms, Abelard appeals to the distinction between perception and intellection. In perception, we grasp the particular named by a proper name; the cognition associated with general terms is, however, intellective. Here, the mind is directed toward an object of its own making, a res ficta. The res ficta is a kind of image, one that is communis et confusa (common and indifferent). It is common to all the items named by the associated general term and proper to none. Accordingly, it represents them all indifferently. Since it is something distinct from any of the particulars that fall under the Aristotelian categories, it is neither a substance nor an accident. It is the product of the intellect’s activity of abstraction, and it is what is signified by the associated general term - it is what we grasp when we understand such a term.

The century after Abelard brought a number of different developments. The appearance of the complete Aristotelian corpus gave a clearer picture of Aristotle’s views on universals, and the rich framework of semantical concepts associated with the developing terminist logic made possible the articulation of a more powerful form of nominalism than those defended by Roscelin and Abelard. That articulation came from William of Ockham (see William of Ockham §6). Following terminist logicians, Ockham distinguished between categorematic and syncategorematic terms. Categorematic terms are expressions whose significance derives from their having a ‘definite and determinate signification’. Syncategorematic terms, by contrast, do not serve as signs of objects; their significance derives from the roles they play when used in conjunction with categorematic terms. Categorematic terms are further divided into discrete and common terms, where this is the contrast between expressions that signify just one object and expressions that signify many and so are predicable of many.

Like Abelard, Ockham identifies universals with common terms. Insisting that every existing thing is a particular, he construes the distinction between universals and particulars as a distinction between categorematic terms that signify just one thing and those that signify many. But where Abelard takes conceptual representations to be nonlinguistic items that function as the significata of the various common terms, Ockham wants to claim that a common term like ‘man’ signifies the various particulars of which it is truly predicable, and that the conceptual representation corresponding to the term ‘man’ is itself a linguistic entity. His idea is that thinking is inner dialogue, best understood by way of the familiar concepts appropriate to spoken and written language. Thus, concepts are mental terms, judgments are mental propositions, and inferences are mental syllogisms. Conceptual linguistic items differ, however, from spoken or written words in that while the latter are only conventionally significant, the former are natural signs. The phoneme ‘man’ has the meaning it does only by virtue of a complicated system of conventions, but the concept man is something whose intrinsic nature it is to play the linguistic role it does; and although we describe conceptual representations by the use of concepts derived from our characterization of spoken and written language, mental language is prior to both. Just as written language is an outgrowth of spoken language, spoken language is an extension of mental language, a sort of ‘thinking out loud’.

The distinction between discrete and common terms, therefore, applies to conceptual representations, so that there are conceptual universals - conceptual representations that are predicable of many - and these are the genuine universals. Since they are naturally significant, the common terms of mental language are in their intrinsic nature items predicated of many; and their universality is the root of the merely conventional universality of the common terms of spoken and written language. Corresponding to the spoken/written term ‘universal’ there is a mental common term universal. To bring out the contrast between his own form of nominalism and its realist alternatives, Ockham tells us that this mental term is a term of second rather than first intention; it is a term that signifies not extramental entities but, rather, intentions of the soul, those that are in their intrinsic nature signs of many.

In characterizing the conceptual items that are in their intrinsic nature universals, Ockham mentions three possible views. One harkens back to Abelard and construes the mental term as a res ficta not found in any Aristotelian category; a second construes mental terms as qualities of the soul that serve as objects of its acts of understanding; the third identifies the mental term with the act of understanding itself. Over the course of his career, Ockham wavers between these views, but he ultimately comes to endorse the third view on grounds of theoretical simplicity.

Ockham’s nominalism extends beyond a concern with universals and the distinction between common and discrete terms. He is also interested in the distinction between concrete and abstract terms, between terms like ‘man’ and ‘humanity’, ‘courageous’ and ‘courage’, and he is concerned to undermine what initially appears to be a plausible account of this distinction. A natural response to the distinction is to say that whereas concrete terms signify familiar concrete particulars (the particulars which are men and courageous), their abstract counterparts signify the abstract entities (humanity and courage) those particulars exhibit. Since the distinction between concrete and abstract terms is found in all ten Aristotelian categories, the view unfolds into the claim that for each category there is a categorically different kind of abstract entity such that, in virtue of exhibiting an entity of that kind, a particular comes to be characterized by the appropriate concrete term.

In combating this view, Ockham argues that the categories do not represent a classification of nonlinguistic objects; they are, rather, a classification of linguistic expressions according to their mode of signification. As he sees it, there are nonlinguistic objects corresponding only to the categories of substance and quality, and the entities in question are all particulars. So there are particular substances (like this man) and particular qualities (like the whiteness of this piece of paper). Abstract terms from the category of substance do not signify anything distinct from the particular substances signified by their concrete counterparts. Abstract terms from the category of quality do tend to signify entities distinct from the familiar substances that we say are white and courageous; but a term like ‘courage’ does not signify some one quality that all courageous individuals share. ‘Courage’ is better construed as a general term signifying individual qualities, the various courages in virtue of which individual human beings are called courageous. In none of the other Aristotelian categories do abstract terms signify any entities distinct from those signified by their concrete counterparts. Indeed, Ockham wanted to claim that abstract terms from categories other than that of quality are eliminable from discourse, that sentences incorporating terms like ‘paternity’ and ‘burglary’ can be replaced, without loss of content, by sentences in which those terms do not appear, but their concrete counterparts (‘father’ and ‘burglar’) do; and a significant portion of his ontological/logical works is dedicated to showing how these translations are to go (see Aristotelianism, medieval).