Are There Universals?
The proponents of universals claim that on top of the particulars (physical objects in observable world) there is an additional type of thing in the world, the universal, from which any particular derives its properties (what it is, its colour, shape, abilities, etc.), and which any objects which share properties have in common. Namely, that if a is F, and b is F, both a and b exhibit F-ness, and this is how we can see that there are similar. More specifically, rather than simply believing that such relations are a convenient manner of speaking, believers in universals hold that such things actually exist in a concrete form, though presumably not visibly or tangibly. I hope to show that it is, indeed, an intriguing suggestion, and that the discussion of such objects is not ridiculous, but that it is ultimately futile, and attempting to explain a problem that simply is not there.
The question of whether there are what are now known as Universals is a very old one - dating back to Plato and his theory of forms. That it has lasted so long without being agreed upon either way must suggest that there is some value in such a metaphysical theory, yet that it is not so obvious as to be accepted as correct by all. Plato suggested that, at a higher level (in an unclear sense) than the world we perceive is a more real world, populated by forms, of which the objects we perceive are merely lesser reflections.
Such ideas have, more recently, been dogged by a large number of objections, many of which are fair, but various which seem pretty facile, and which focus on misreadings of Plato's works. Various philosophers, for example, seemed to suggest that Plato was claiming that 'heaven' is basically a big warehouse, filled with tables, leaves, love, justice etc. - one of each. It seems likely that we should give Plato a little more credit than this, and he certainly does not suggest that the world of forms be just like the world we perceive, but with only one of each thing (contained, presumably in a warehouse with a single wall), and where everything sparkles because it's so clean. Even if we are kinder to Plato, however, analogous criticisms might hold water - questions such as where these forms he speaks of would be located, and why we can't perceive them; is it possible that we might, exploring Iceland hit upon the so-called 'World of Forms' ? But even if we did find them we might not be able to see or even feel them - it all seems a little far-fetched, yet it is fair of critics to demand answers to just these sorts of questions.
Russell is probably the philosopher that most re-ignited the debate recently, under the new heading of 'Universals'. Approaching the question from a slightly different direction, he suggested that the way we look at language is very narrow minded, and that we spend far too much time considering nouns and not enough considering the relationships between them. The example he uses is that of the sentence 'Charles I's head was cut off' - he states that whilst we consider Charles I, his head, and the event of his head being cut off, we tend not to concentrate on the word 'cut' (and by extension, other words like it). There is a tendency, Russell suggests, to view verbs and adjectives simply as inferior things that predicate nouns, and not view them as things in themselves. His example, however, seems to employ a remarkably circular logic. He says that people don't spend much time thinking about universals and aren't interested in them, not viewing them as complete things in themselves, but should be. Why? Because, he says, such things exist independently and are therefore interesting, and deserve to be thought about. He suggests that the fact that people view predicates as 'demanding' a context fails to see that they have an innate meaning in themselves, and suggests that people view them as subordinate to nouns, which people are happy to consider on their own.
Such an objection to the 'ordinary' way of looking at language seems very unfair. Firstly, people don't seem to view nouns as things to be used on their own (and don't imagine that 'Russell' would be a meaningful sentence on its own, any more than 'cut' would be) and secondly even such a suggestion wouldn't mean that people presume them to be more important than predicates. Finally, and probably most importantly, however, I can't see how, even if a subject-predicate way of looking at language might place a greater emphasis on nouns, we should view 'fairness' as a basis for a linguistic theory, and why we should spend as much time considering verbs as nouns. I'd say we actually try to spend more time considering sentences than any part of speech - though Russell might complain that the consideration of 'Charles I's head was cut off' is simply a consideration of the event.
Worse still for Russell, few philosophers (though there are a few) would go so far as to deny that there are such things as relations - but simply would not go so far as to say that they exist as free-standing objects on their own. It would be hard to claim that 'Charles I's head was cut off' is simply the same as saying 'Charles I', 'Charles I's head' and 'An event in history' - on their own this list does not convey anything, and most would agree that these things are somehow linked together (or rather, the first two cease to be!). The thing that Russell has not yet shown is that such a linking is any more than just a concept - a way of talking - and holds as a thing in itself. Russell will have to prove in other terms why universals deserve as much consideration, and why they should be thought of as existing independently of mind; luckily he does attempt to.
A much stronger reason for believing in universals is what is known as Plato's 'One to Many' problem - namely the difficulty we will have in applying the same property to a number of different particulars if we cannot say that the property itself is a single, independent thing. How could, as in one of Russell's example, we recognise two entirely different sorts of objects both to be white if there is not a single unifying thing which we might call 'whiteness'?
The best objection to this line of argument is probably that we certainly do have a concept of whiteness, through which these things are linked, but that it is simply a concept in our minds, and not indicative of any deep truth about the world. Russell, however, seeing the possibility of this objection, proclaims that 'In the strict sense, it is not whiteness that is in our mind, but the act of thinking of whiteness'. This seems highly unsatisfactory to me. Those, for example, who do not believe there is a physical world as distinct from what we have in our mind would say things like 'That orange is merely in my mind', and only within a framework when we believe a distinction must be made between the object itself and our perception of it would we deny such a statement - but this is exactly the sort of thing Russell is trying to prove. He is trying to show that whiteness and our conception of whiteness are distinct simply by stating a claim that relies on our already accepting that they are.
A better response from Russell comes in his questioning how, if whiteness is merely a thought, two men could conceive of it in the same way. This suggestion, that thoughts are incommunicable and no two people can have the same thoughts, seems at once both too strong and too weak. Do we, for example, believe that when two people think through the addition of 1+1=2 that they have a different experience of doing so? It seems that such a process is so simple that any two instances of this thought process would be pretty much the same - and certainly not distinguishably different - and in this case Russell's suggestion is too strong. If we are unable to prove that they are different, we will be unable to claim that two instances of thinking of whiteness might be different. His suggestion is also too weak, because it suggests that any experience of a universal must be communicable and observably the same, and we would hope to have more rigid requirements than the observation simply referring to a universal rather than a particular. In his own example, that of the colour white, there is no logical necessity (and, in fact, I think it is provably false) why two people observing a white thing should see it as exactly the same colour, regardless of their perspectives, standards of eyesight, etc., and if this is not the case, then how can Russell claim that a universal helps us to identify a common attribute at all? Such a claim would also seem to preclude any form of language, unless he is suggesting that the very words we use are common examples of universals which are communicable.
Russell even concedes, however, that we identify whiteness by abstracting from the white things that we see around us, and claims we are extrapolating from the particular to the universal. Clearly this is a major problem for his theory, since it implies that we never have direct contact with the universal, and as such it is not reasonable to claim that we move (as Plato also suggested, with his theory of forms) from an acquaintance with particulars to an acquaintance of the universal rather than simply deriving a general idea in our minds of what being white is - which would be purely a thought, and a way of talking.
Russell's final attempt to establish universals as distinct from thoughts about these sorts of things is to argue that universals exist, and statements about them are true regardless of minds. Again, this seems to lead to huge problems. If whiteness, for example, is to be totally independent of minds, we must know what it would be for something to be white regardless of minds - it appears the answer to 'If a tree falls in an empty wood, does it make a sound?' would be rendered entirely trivial. Yet since whiteness appears just to be the brain's way of interpreting light with a particular mixture of wavelengths it seems rather dubious to start talking about the colour of light without an interpreting medium.
An advocate of universals might respond that we've answered our own objection - that wavelengths of light exist without an observer to see it, but such a claim would seem to miss the point, since the experience of colour and the wavelengths that move about the universe seem to be very different sorts of things, even if one provokes the other, and the suggestion that if there were no minds there would not just be wavelengths but also the concept of colour seems a difficult one to accept without much stronger proof (and definitions of terms).
Another (better, in my mind) response would be that there is still the universal, but no particulars derived from it, almost like there in a universal but no instantiation of it. With this defence, however, we seem to be back where we started, since without minds, suddenly everything that was coloured would cease to be, and, whilst the fanatical universal-lover might like this idea, we would have lost the connection between the normal linguistic usage of words like 'white' and the concept of the universal, 'whiteness', and it would appear to have a very flimsy connection to the particulars that hold it as a property.
Perhaps the most talked about single example of something that is difficult to view as anything other than a universal is that of numbers. We definitely seem to have a communicable idea of 'two' without having to resort to pointing out sets of two things, and whilst we could argue that these are derived from observations, it seems unlikely that more complicated mathematical examples could be - how, for example, can we extrapolate to an idea of p, and it doesn't seem likely that we could invent a number that is so pervasive throughout nature, especially since how we write it is dependent on the base we use - and yet it seems to be useful regardless of base. Russell's attempt, however, to show that numbers are not simply thought-concepts, whilst slightly more plausible than his more general strategy, is again flawed. He gives the example that all products of integers we never think of are over 100 (since we have thought of all products of integers below 100), and suggests that, since any product we think of will be ruled out by the question, it cannot be to do with thought processes. Whilst this is a clever route to take, I am not convinced that this does in any way suggest that numbers are not to do with thought. Even invented concepts can be played with; we might give as an analogous example the suggestion that every custom humans never use would have to have been invented by a human - yet he would not (I hope) suggest that there were 'universals' of customs, which we have to observe to pick up a particular custom, rather than them being constructs invented by society.
It is Quine, however, who gives the most impressive rebuttal of the universal-advocates suggestion that they exist in the world, rather than simply (for example) in thought. He suggests that there is simply a mistake being made, where people confuse naming something with something having a meaning. Just because we are able to give a particular thing a name, does not mean that the name thing has real content. His primary example is that of the Pegasus as opposed to the Parthenon. Whilst both the Parthenon and the Parthenon-idea have content, he claims that only the Pegasus idea does. As to Russell's suggestion that there is whiteness as distinct to the whiteness-idea, Quine would presumably claim that the idea has content whilst whiteness itself does not, and can only be used as a shorthand for the idea.
The way that Quine escapes the trap set by the follower of universal is to avoid concepts like whiteness altogether. Rather than saying that a has the property of F-ness he claims we need only say that a is F - that Russell's attempt to complicate the matter is fallacious. If we can view properties in terms of simply things that particulars happen to be we need not worry that the properties might be things in themselves. a can be similar to b simply by being of the same ilk, rather than sharing an abstract thing, from which each of them gets their ilkness. All names, Quine suggests, can be converted easily to descriptions, and this renders the example of colours very trivial. Rather than white things have the property whiteness, they simply are white.
Jackson, however, disagrees with Quine's formulation, saying that he has examples of sentences where the universals cannot be converted into descriptions. 'Red is a colour' is one of his examples, where it just does not seem possible to rephrase this only using descriptions - 'Anything red is coloured' just doesn't seem to work, since it might, coincidentally, be the case that anything red is circular, for example. The problem, he suggests, is that without universals we will not be able to hang together two separate qualities. We will not be able to explain what function the 'is' is performing in sentences like 'Red is a colour' without universals to be joined together. Rather than 'Red is a colour' Jackson suggests a reformulation to 'Red has the property of being a colour'.
Devitt agrees with Quine's assessment of the situation, arguing that there is in fact no real problem present, suggesting that a is F simply because of physical facts about the world. He questions, further, why 'a is F' is any worse than 'a has the property F-ness', since both seem to have as little an ability to link together properties shared between particulars as one another. What is the meaning of this 'has the property', for example, and why should we not extend it one stage further, saying that 'a and F-ness exhibit a having' or even further to 'a and F-ness and having are had'. Clearly we are never going to get to a position where the meaning of the connection are understood, at least not by this root. The most important thing that Devitt claims, however, is that no explanatory gain will be made in such a reformulation, and this seems to be clear now that we are aware that the connections are no more clear - surely we should go for fewer vaguely understood connections, rather than more which introduce new variables?
Overall, therefore, I would have to side with the universals-sceptics, and claim that the question the creation of such things tries to answer is simply not a problem. If Russell were correct, and there were some substantial difference in saying that 'a is F' and 'a exhibits the property F-ness' then the position might seem slightly more plausible, but as it is, with no advance in understanding, and no further explanatory power, it does not seem sensible to presume that there are such concrete invisible intangible entities, rather than simply concepts and names created for classes.