In assessing the claim that a realist account of universals is indispensable to our ability to construct propositions, we draw upon three main positions. These are realism, nominalism and conceptualism. The realist school would claim that universals are real, and they are different from the particulars that instantiate them. The types of realism I will reference here are Aristotle’s ‘strong’ realism and Platonic realism. In contrast, the nominalist would assert that only particulars exist and deny that universals exist as real entities. Conceptualists urge that words that might seem to refer to universals actually refer to concepts. They are mind-dependent classificatory schemes.

Realists tend to argue that universals must be distinct entities in order to account for various phenomena. For example, a common realist argument is that universals are required for certain general words to have meaning and for the sentences in which they occur to make sense. Words such as ‘to’, ‘on top of’, ‘honesty’, ‘rationality’ etc. cannot be labeled as particulars because there are no particular ‘things’ that they reference. The most famous of realists was Plato, who argued that for us to know something, that which we know must be unchanging. Since material individuals are subject to change, Plato argued, there must be things that don’t change, suitable as objects of genuine knowledge. Here fit universals. In terms of Platonic realism, he believes that in order to explain the qualitative identity of distinct individuals, we must accept there is another entity besides them- universals, or as Plato would call them: Forms. For example, if two branches are both brown, then it is because there is a Form of Brown, which has manifested itself into both the branches at once. However, there is a problem in that we cannot know what it is in the nature of the Form, which provides for the bit of brown that we see in either branch. Plato believes Forms to be epistemically inaccessible- as they are wholly abstract, outside of space and time. Plato accounts for the problem of Forms needing to be somehow related to the particulars by postulating a theory of participation, and speaks of things ‘participating in’ Forms, and that they get their qualities by virtue of this relation. There are several objections to Platonic realism, the first of which may simply be that Forms are ‘strange’. If the two branches I discussed before are both individuals, then it is implied that they can only be in one place at one time. But if the brownness they share is a universal, then the brownness is a real non-individual. Suppose one of the branches was to be destroyed. There would be one fewer individual- but the brownness would not be diminished. This suggests that the presence of brown in one place doesn’t affect the simultaneous presence of it elsewhere. However it is not made clear how universals can be wholly present in each of the places they exist and at the same time, present in many different places at once. A second problem is that of materiality- it seems to be a mark of materiality that things can be in only one place at a time. If so, then universals cannot be material. Therein lies the problem, in terms of causation. As we usually understand causation, one thing affects another by interacting with it. This is a problem because we can therefore not understand how universals interact with other things that exist. This issue becomes more acute when we then wonder hoe we can know about universals at all. Surely in order for us to know about them, they need to interact with our brain? In this respect, they seem quite mysterious. Aristotle’s realism seems to bypass some of the problems, which cripple Platonic realism. It rejects independently existing Forms, asserting rather that universals are qualities that are in qualitatively identical individuals. We need not have Forms that are abstract, in the sense of being non-temporal and non-spatial, but rather both in space and time, and able to be in two places at once. Multiple exemplification is certainly strange, but not as strange as existence outside of space and time. Ockham’s razor may be used to argue against all kinds of realism, as universals may be deemed unnecessary and this argument states that, all things being equal, the fewer kinds of entity in ones ontology, the better. Those opposed to Realism argue that we can account for qualitative resemblance and identity without universals, and this ought to be preferred. This is demonstrated by searching for alternative accounts such as nominalism or conceptualism.

In the anti-realist perception of universals there are two distinct positions. These are nominalism and conceptualism. Nominalism asserts that there are only particulars and that general terms refer to the resemblances between them. Conceptualists, on the other hand, argue that words that might seem to refer to properties really refer to concepts. They are classificatory schemes that are totally mind dependent. Nominalism is a theory that claims that universals aren’t real. They claim instead that they are sets of objects that resemble each other. Qualitative identity, e.g. ‘is red’ is accounted for by belonging to a given set i.e. set of red things. A set is constructed by adding particulars which resemble each other more closely than other non-members. However, there are certain objections with resemblance nominalism- such as they fail to account for some distinct properties. For example: Has a heart and has a kidney. All who have hearts have kidneys, therefore the set membership is the same and so too is the set. Logically this means that having a heart is the same property as having a kidney. However we can say that the fact we can have one without the other shows that they are two sets. For example, we can remove both kidneys but still have a heart. There is a second problem with resemblance is set construction. We construct sets by grouping particulars that resemble each other. Things can resemble each other in many ways. For example: two identical apples, except one is red and one is green. Both resemble each other therefore they both go in the red set. Therefore the explanation for redness is by reference to a set containing non-red things. In other words it doesn’t explain it at all. A response is that the resemblance is not the ‘right kind’ of resemblance to be in that set. However, what does it mean to be the ‘right kind’ of resemblance? If you mean it has to resemble by being red then you’re explaining redness using redness as a criteria which is circular. The biggest flaw in the resemblance nominalism argument is that resemblance itself is a universal property. Resemblance nominalists make use of the relation of resemblance. i.e. A resembles B which resembles C. A,B and C are particulars, but ‘resembles’ is a universal. In attempting to explain away universals they require universals. Anti-realism conceptualism claims that generality is not a feature of reality, but instead a feature of our minds and the concepts or ideas in minds. Conceptualism is a middle way between realism and nominalism, we form concepts in our minds, which several individuals fall under. Universals do share something but that something is not a metaphysical entity but a commonality in our minds. Concepts can, however, be misapplied. E.g. ‘that is a dog’, when really it is a cat. So the generality of universals must be explained by correct concept application. But then we can ask what makes a concept application correct or incorrect? Either it is to do with the fact that these resemble each other – then we are back to nominalism- or that they share some property, but that puts them back into realism. As neither solution is acceptable, conceptualism simply collapses into either realism or nominalism.

In conclusion, there seems no answer to the question of whether universals are real entities. Either, we must expand our ontology to include some very strange entities, or we get involved in complex linguistic analysis which seems to get tied up in knots.