By James Lythgoe

Look to any lake-adjacent patch of grass during the daytime and you will almost certainly see ducks. The question of Metaphysical Realism (I shall from here on use ‘Realism’ to refer to the metaphysical position) and Nominalism can be brought to this: why are the various quacking, feathered, beaked objects ducks?

Just what is it that makes these objects, as opposed to any other object, ducks; in what does ‘being a duck’ consist? And in general what makes any given object a member of a group or set? I intend to briefly cover the two principal responses to these questions; Realism and Nominalism (I will leave aside the third traditional response: conceptualism) and aim to demonstrate Nominalism to be the superior account.

The crux of the argument lies in whether we want to include universals in our ontology or not. Universals are things which are common to more than one object (particular). For example, if objects x and y are both red then they posses, or instantiate, the selfsame property of redness.

The Realist position is to accept universals into their ontology as mind independent objects i.e. to posit that they exist independently of any perceiving consciousness . So, by turning to my example of what makes ducks ducks, they must instantiate the property of duckness. This alone is not very illuminating: we must ask what is universal duckness? Here it is prudent to distinguish between two forms of Realism: Ante Rem Realism (AR) and In Rebus (moderate) Realism(IR).

AR will be painfully familiar to anyone who has taken, or is taking an ancient philosophy module and studying Plato’s Republic. AR states that universals (Plato’s Forms) are immortal, unchanging and exist outside of time and space in a so called Platonic heaven. Particular objects or particulars manifest or instantiate these universals or forms which never change; even if they have no instantiations they persist. So ducks are ducks by virtue of their standing in the relation of instantiation to the universal of duckness or, to keep with Plato, the universal form of ‘the duck’.

Even this bare sketch has thrown up several problems. Firstly, ascribing duckness with a place in some non-spatiotemporal realm and immutability does not seem to shed light on matters: in fact quite the opposite. Secondly, even if we can comprehend of non-spatiotemporal existence, then it raises the question of how non-spatiotemporal objects can stand as paradigms, or even be the root of, spatiotemporal properties, relations etc. Conversely, how can spatiotemporal objects instantiate the non-spatiotemporal? And just what is this strange relation of instantiation anyway?

IR can be seen as attempting to flee from the difficulties of AR whilst maintaining the mind independence of universals. IR states that universals exist in their instances; so, for example, the property of duckness is spatiotemporal but diffuse through all the instances of ducks existing at a particular time. Should there cease to be ducks, there would cease to be duckness, for an IR property is simply its instantiations.

IR evades all the objections from AR by moving universals to the spatiotemporal realm. It changes the instantiation from a relation of a spatiotemporal object to a non-spatiotemporal object, and into the presence of common spatiotemporal features in spatiotemporal objects.

However it must be asked just how it is possible for one and the same thing (a universal) to have a spatiotemporally diffuse existence? Furthermore, this diffuse existence gives rise to odd situations. For example, if some unpardonable scoundrel were to round up and destroy all objects instantiating duckness, then we may have to say that he has destroyed duckness. But won’t people still remember ducks and have an idea of what made them ducks? Will this idea not be a further instantiation of duckness? If the Realist responds that an idea does not instantiate duckness then how can it, on the realist world view, be an idea of ducks? If the realist admits that the idea is an instantiation of duckness, then the universal has become mind dependent — the very thing the Realist wished to avoid. To make this clearer, imagine that only one person still has an idea of ducks; if he were to lose his memory of duckness then the universal would cease to exist.

I shall now discuss a Nominalist account and attempt to show that it is not only preferable to a realist approach, but that it captures our use of language with regards to objects. Put simply, Nominalists hold that universals are not to be included in our ontology, that they do not exist. All that exists are particular objects. We group them, not by virtue of their pertaining to one universal or another, but as a convention of language based on our perceptions of them. For example, we find that looking at a clear sky, a clean body of water and a lapis gives us a similar colour sensation; so we group them under the name blue. Similarly, we observe that certain water fowl share certain biological features and call them ducks.

Imagine that we were to come across a society of people, perhaps in some hitherto undiscovered under croft of central hall, that only had words for juxtaposed pairs of colours and instead of individual colours. Imagine also that they group lake-life not by physiognomy but by habitat. What would a realist say about such a society? It seems they will have to say that the new society has plainly missed the point: after all, on the realist account, the new society has merely failed to recognise the universals of blueness and duckness. Does the difference in our ways of grouping rest on some superior knowledge of reality? I think not. The Realist might try to object that there is in fact a form of the new society colour, bleen (a juxtaposition of blue and green), or the new society genus Derwentium (those animals and birds which nest around Derwent College).

However once this step has been made there may be no going back; are we then to suppose that there are universals which correspond to every conceivable mode of grouping objects? Positing a universal for every group of objects defeats the purpose of universals as a means of distinguishing objects: we are left in the same position as if there were no inherent relations between objects and with a cumbersome and pointless ontology. If the determined realist still wants to object, they may account for the difference in our use of language and deny that all conceivable universals exist, by saying that universals only exist when matching actual (rather than merely possible) language forms. At this stage, the realist must have some account of language’s ability to create universals that are mind-independent, spatiotemporally dispersed objects, and also the ontological status of language itself as being mind independent. Without doing so their position collapses into nominalism.

In our ordinary speech we talk about an object being blue or being a duck; we do not talk of it being an instantiation of such and such universals. Imagine someone telling you that they have adopted one of the campus ducks; you naturally ask which one, and they reply ‘the instantiation of duckness instantiating beakedness, featheredness and goldness’. This shows how realism is at odds with our everyday language use for two reasons. Firstly it would obviously have made more sense had he said ‘the golden duck’. And secondly, a full realist unpacking of the event would have meant I had to cash out ‘imagining’, ’someone’, ‘telling’,’you’, ‘adoption’,’one’ and ‘campus’ in terms of their relevant universals. However we can understand the event without talking or even thinking about any universals; by contrast nominalism states that our grouping of objects and defining them thereby is just a convention of our language, there is no special relation of object to properties to look for behind the use of the phrase ‘the golden duck’.