Jens Tomas Anfindsen
Department of Philosophy, University of Uppsala
Published electronically by HonestThinking.org
When we know anything, what kind of
thing is it that we know,
some particular thing or something universal?
Particulars in the philosophical tradition are items that are numerically one. ‘A particular’ is ‘one thing’, like a dog, a jet plane, a stone or an angel, say. Usually particulars are thought of as material and perceptible items; that is the most obvious connotation the term has, anyway.
It is difficult to give a strict definition of what a universal is which at the same time satisfies all philosophical parties. For whether or not there are universals, and what they are if they are, has been the subject of great dispute in the history of philosophy. This subject is really and in the full sense of it something of an enigma! But as a merely nominal determination it seems fair to submit that by a universal we shall mean ‘something which is apt to be truly predicated of more that one thing’. Quite typically, by ‘a universal’ people will mean ‘a nature or an essence which more than one thing can (or could) have’. Let’s consider some candidates for the title.
‘Dog’. ‘Dog’ seems to be something all dogs are, it is a predicate that can be truly predicated of all dogs. So what is it that we predicate when we say ‘dog’ of something? (I exclude pejorative usages from consideration). It would seem that we are saying of some particular item x that it has the general property P; where x is a particular dog and P is the predicate ‘dog’. “Fido is a dog”, would be an example. But how can a particular thing have a general property? Say Fido has a buddy, Beefy, and that Beefy is a dog, too. Now “Beefy is a dog” would also be true. Are Fido and Beefy having the same property? How can two different things have the same property? And if two things can have the same property, what kind of thing is that, the kind of thing that may be had by several things?
If you are inclined to say that the kind of thing that two or more things can have in common is something independent of the human mind, then you are inclined to be a realist. You are inclined to believe that universals are extra-mentally real.
If you are inclined to say that the kind of thing that two or more things can have in common is a mental item, a mental representation or a concept, then you are inclined to be a conceptualist. You are inclined to believe that universals are concepts.
If you are inclined to say that the kind of thing that two or more things can have in common is just a name, a linguistic item, then you are inclined to be a nominalist. You are inclined to believe that universals are nominal.
There are subtle sub-differentiations of each of these positions, but I think the three models just outlined correspond to the three main philosophical alternatives with regard to the ontology of universals.
In what follows, I will attempt to give an historical overview of the most important philosophical models for explaining the nature of universals, covering the period of time reaching from Plato to William of Ockham. I emphasize that this is an overview and that I will not do justice to the full intricacy with which this subject has been discussed. Also, it is primarily just doxographic. That is, I will not so much engage in critical discussions of these various positions as I will simply report them. My presentation has particularly benefited from the historical exposition given Richard Bosley and Martin Tweedale in (Bosley 1997), and Gyula Klima in (Klima 2004).
Plato is without a shadow of a doubt philosophy’s archetypical realist. Plato is known for the view that universal predicates are names of Forms, and to have held that these Forms themselves are eternal, non-temporal and non-spatial, and merely instantiated in perceptible particulars. Plato says that the sensible particulars “imitate” or “participate” in the Forms of which they are instantiations. Plato takes the Forms to be archetypes in the sense that they are patterns for the production of the sensible world, and that they are, in some sense not to be dwelled upon here, more perfect than sensible particulars which imitate them. The Forms are also active causes of their own instantiation in particulars; it is in virtue of the activity of the Forms that sensible particulars may have universal traits, features or essences which the Forms are.
Plato’s theory of Forms is highly disputed and there are various interpretations of what Plato’s Forms really are. It is impossible for me to do justice to this controversy here. I will remark, though, that the controversy is of fairly modern date. An interpretation of Plato’s theory of Forms along the lines that I have outlined above was, I think, something of a standard and received view in antiquity and the middle ages. When one speaks of Plato as a realist or when one speaks of “Platonic metaphysics”, one normally means something like this: The view that for every universal and non-privative predicate there is a Form, and that these are the eternal and perfect patterns after which the world is produced, and that these, not the particulars which instantiate them, are the subjects of truth and knowledge.
Aristotle believed, just as much as Plato did, that universals existed, but he denied that universals were singular beings, something numerically one, as Plato had thought. Aristotle also denied that any universal whose definition would include a material element (such as ‘nose’ would include ‘flesh’) could exist “separately”, i.e. at least he denied that those kinds of universals could be immaterial substances existing in separation from material things. Whether or not Aristotle thought that there were some other universals (notably what the schoolmen called “transcendentals”, such as unity, truth, goodness and being), which would be separate and either be or belong to some single being, is an open question which I personally consider to have been insufficiently explored.
The standard Aristotelian view and the most accessible part of his doctrine, anyway, is that universals are kinds of things, not things. ‘Dog’, e.g., would be a kind, so Fido and Beefy would be individuals of the same kind. Their natures would have the exact same definition, but the dogginess of Fido would be numerically distinct from the dogginess of Beefy. But there is a dubbelness here which lead to Aristotle talking about forms in a dual manner: Primary form is the essence which is expressed in a definition, it is this essence which is the subject of scientific knowledge; secondary form is an essence which is actualized in a particular thing, a form+matter compound. The former is “a such”, the latter is “a this”, as Aristotle expresses it; the former is a universal the latter is a particular. Both exist according to Aristotle, but they exist in different ways. The problem is just that it is so hard to understand how “suches”, the universals, exist.
Are universals just classes or sets of things, and nothing whatsoever apart from its members? If so, then universals cannot be substances, as Aristotle also clearly says. But it is precisely forms which are primary substances, according to Aristotle, and we have just seen that it is the universal essences of things that are primary forms. Also, it is the primary forms which are the object of science, and Aristotelian science, to be sure, is concerned with what is. How can this be? My answer: I do not know! But it is the case that Aristotle repeatedly says that nothing universal can be a substance.
Plotinus’ student, Porphyry, wrote the Isagoge (The Introduction) as a systematic exposition of Aristotelian logic. Even more than Aristotle’s own works, it was the Isagoge which formed and defined the conception of Aristotelian logic throughout the middle ages. This was not least the case due to Boethius’s translation and commentary of this work into Latin. Important though the Isagoge is, it evades that very question which is so crucial to us, what universals are:
I shall beg off saying anything about (a) whether genera and species are real or are situated in bare thoughts alone, (b) whether as real they are bodies or incorporals, and (c) whether they are separated or in sensibles and have their reality in connection with them. Such business is profound and requires another, greater investigation (Isagoge VI.3.1.; transl. in: (Bosley 1997, p. 358)).
I wanted to remark the Isagoge nevertheless, because it demonstrates how the Aristotelian terminology of genus, species, differentiae and so forth, would be employed by Platonists and Aristotelians alike, and would be employed whether one took a realist or a nominalist stance on the ontology of universals. It is Porphyry’s formulation of the problem of universals that lay the ground for and sparked of that debate among the schoolmen.
Boethius wrote a commentary to Porphyry’s Isagoge, which at the time had become something of a standard introduction to Aristotelian logic. In discussing the ontological status of universals, building on the philosophical commentaries of Alexander of Aphrodisias, he gives an admirably clear formulation of an Aristotelian brand of moderate realism.
The ontological status of universals is analogous to that of a line, Boethius explains: A line only subsists in some body and no line can exist separated from body. Yet when the mind cognizes the image of a complex perception, and since the mind possesses a capacity to divide and isolate the elements of complex perceptions (as well as synthesizing them), it can perceive within itself the image of- and form concepts of such elements individually. Among the elements perceived in complex perceptions are lines, and hence the mind may perceive within itself the image of- and form the concept of a line. In that case, the mind cognizes an incorporeal item of such a nature as to never be able to exist without a body, yet the mind would, according to this argument, cognize that item with respect to its true properties – in fact, the argument claims that the only way to cognize the true properties of lines (or any universal) is to cognize them in abstraction from any other thing. So though the mind cognizes in isolation something which can never exist in isolation, it nevertheless cognizes that something with respect to its true properties. This is how the mind cognizes genera, species, differentiae, and all other universals, according to this argument.
So how then, on this model, does universals exist, in things or in minds? Well, in a sense in both, though in different ways. The mind possesses universal representations of particular things, isolating in thought what cannot exist in isolation, yet such as to cognize true properties of particular things. Insofar as some such property is regarded universally, it only exist in as a representation in the mind, as an abstraction the active intellect. But insofar as some such property is true of individual things, it exists in those particular instances.
The above outlines Boethius formulation of what he considered to be the Aristotelian solution to the question of universals. (Boethius’s own stance on the matter was probably Platonic, positing the ultimate ground of universality in the objects of the divine mind.) The outlined position is known in the history of philosophy as “moderate realism”. I would call it a conceptualist variant of moderate realism.
St. Thomas would later defend Boethius’s version of moderate realism, but with an interesting extension. Though claiming that we can only get hold of universal concepts through experience of particulars, i.e. inductively, he also acknowledged that reality has its basis in the ideas of the divine mind. Thus sensible particulars are, to some extent at least, modeled by the divine ideas, and the universal concepts we form of them will resemble those divine ideas according to the degree with which they accurately represent the essential nature of things. So Aquinas’s moderate realism is coupled with the theory that any actual essence resembles a divine exemplar in the divine mind, so called “exemplarism”. It seems fair and well to say that Aquinas position, here, essentially resembles a traditional conception of Platonism, cf. (Roberts 2005).
In the subtile and difficult work Logica Ingredientibus Abelard argues that the only kind of thing which corresponds to Aristotle’s definition of universals in De Interpretatione – “that which is apt to be predicated of many things” – are words. So only words can be universals. Abelard thus represented a nominalist stance on universals, and his rejection of the idea that logic basically treats of “real” entities instead of words and sentences, was influential in steering the scholastic movement away from realism and paved the way for Ockham and, later, the British empiricists.
On an Aristotelian view, words have meaning in virtue of signifying mental concepts. But now, if one rejects conceptualism in opting for pure nominalism, one needs some sort of account of why it is that some words have a universal representative function. The question Abelard has to answer is this: In virtue of what do we give a common name to a potential infinity of individuals belonging to the same kind?
This question requires an answer in terms of a common cause. But this common cause cannot be a common thing in the manner of a Platonic form or a universal concept. According to Abelard, the common cause for the universal representative function of universal-words is something he calls “the status” of the things they represent. It is extremely difficult to understand what Abelard really means by the “status” of things. Relying on the exposition of Gyula Klima, it seems that a status corresponds to that which cause a proposition to be true. For example for the proposition “this is a man”, that which makes the proposition true is there being a man. So “being a man” is a status of things, the status of things that will determine the truth value of propositions employing the universal-word “man” in its predicate. But “being a man” is neither a word, nor a concept of our minds, nor a thing, nor something universal (since every man is individual), nor a bundle of common properties. Still “being a man” is the status which causes the determination of the truth-values of propositions employing the universal-word “man” in the predicate. So what is the “being-x” function pertaining to a status?
According to Abelard, a status is an object of the divine mind, and so the “being-x” function of any status must simply consist of its being perceived as such in God.
This leaves us, I think, with the following Abelardian picture. There are individual things. Things have a certain status which can be referred to in essential predications, and that status is an object in the divine mind. The universal-words that correspond to such status are, precisely, just words. Individual word-users have mental concepts corresponding to those words, but there is no telling whether various users of the same universal-word have the same concept (presumably, no two mental concepts will be completely alike) and there is no telling whether or to what degree individual concepts correspond to the divine status. However, Abelard contends that names can nevertheless signify the status of things because their inventor meant or intended to impose them in accordance with certain natures or characteristics of things.
Abelard’s nominalism is quite peculiar and it seems fair to say that his doctrine of status retained a trace of realism with regard to universals. Nominalism, in the sense with which it has been standardly perceived of in our philosophical tradition, was formulated by Ockham.
Ockham is the founder of nominalism in the so-called “modern school” – via moderna. Science has to do, he maintains, only with propositions, not with things as such, since the object of science is not what is but what is known. Science deals with general concepts which, as such, exist only in the human mind. Ockham’s epistemology thus antedates logical positivism, to the extent of claiming that science is only supposed to and can no more than describe the content of perceptions in a formal way.
Ockham rejects the idea of moderate realism that universals are something in particulars which is distinguished from them not realiter but only formaliter. He considers the universal as an intention of the mind, a symbol representing several objects, but being numerically one and an individual mental representation.
External objects call forth sense-impressions in us, which are transmuted by the active intellect into mental images. These images are a product of the active intellect, not species which flow from the object into the potential intellect. The reality of these images is thus, in the modern use of the terms, not objective but subjective. The result of this line of reasoning is the absolute subjectivity of all concepts and universals, and the limitation of knowledge to the mind and its concepts – although these are real entities because of their subjective existence in the mind.
Bosley, R. N. T., Martin, Ed. (1997). Basic Issues in Medieval Philosophy. Ontario, Broadview Press.
Klima, G. (2004). The Medieval Problem of Universals. 2005.
Philosophy, T. I. E. o. (2005). Nominalism, The Internett Encyclopedia of Philosophy. 2005.
Roberts, M. (2005). Dietrich von Hildebrand and St. Thomas Aquinas on Ideal Essences, Aletheia. 2005.
Ross, W. D. (1951). Plato's Theory of Ideas. Oxford.
 There exists an enormous amount of material on Plato’s theory of Forms. An instructive and systematic introduction which also handles the delicate issue of how the theory of Forms was developed, is given by Ross, in (Ross 1951).
 For this section I have benefited from The Internett Encyclopedia of Philosophy’s article on “Nominalism”, .