1.4 Three characteristic difficulties in discussions of representation
I have hinted that accounting for the nature of representation – whether it be the meaning of utterances or the content of our mental states – is not easy. There are several reasons for this, and it is as well to take note of some of them from the outset.
One is that there seem to be several different senses of ‘meaning’, ‘represents’ and related terms like ‘stands for’, ‘being about’, ‘expresses’ – differences that have been glossed over here but will need to be distinguished. Moreover, there are different kinds of thing that can ordinarily be said to represent or mean or stand for or express. Sometimes we talk of people meaning something, at other times we talk of their words meaning something, and often we talk of their utterances (i.e. the actions people produce using words) as meaning something. Different again is the meaning of a mental state, i.e. what is represented in thought rather than through spoken or written language.
Another source of difficulty is that meaning is not quantifiable in the way that, for example, temperature or humidity are. Generalising from this point, ordinary scientific methods do not seem suited in any straightforward way to assist us in our efforts to understand either what representation is or how it is possible for human beings to represent. This is unfortunate because scientific methods have, without doubt, been extremely effective in other spheres of enquiry.
A further reason for thinking that science and the topic of representation are not well suited to one another has to do with the close connection between representation and the notion of correctness. Consider representation as it figures in belief. One way of distinguishing a belief that cats have kittens from a belief that dogs have puppies is that these two beliefs have different correctness conditions from one another. The first belief is correct if, and only if, cats have kittens. The second belief is correct if, and only if, dogs have puppies. But correctness and incorrectness – being right and being wrong – are normative or evaluative properties. As such, science seems ill-suited to describing them. Physicists and geologists, for example, do not include evaluative notions in their theories: a microscopic particle or a rock formation would never be described as ‘correct’. It is true that geologists might evaluate a rock as a ‘nice sample’, but this is only an indication of its usefulness to the interests of the scientists, perhaps for illustrative purposes. Being ‘nice’, like being ‘correct’, is not part of any actual geological theory.
Finally, represents differs in strange ways from ordinary relations such as bumps into or is late for. To illustrate, consider our ability to think and talk about – to represent – things that do not exist. It is not possible to bump into things that do not exist. Similarly one can only swim in, read or kick things that actually exist. How is it possible that we can be related to a non-thing? Yet ‘x represents y’ seems to be just such a relation. Were it not so, we would not be able to think about or talk about Santa Claus or God unless they both existed, and it seems plain that we can think about and talk about both without making any such assumption. For example, one can think about and discuss whether they exist. A phenomenon related to our capacity to think and talk about things that do not exist is our capactity to think and talk about one thing without thereby thinking or talking about a second thing, even though that ‘second’ thing is in fact identical with the first thing. Examples bring out the contrast with ordinary relations. With ordinary relations like bumps into, if x bears that relation to y, and y = z, then x also bears that relation to z. If by accident I bump into Mr Jones on the bus, and Mr Jones is identical with Plastic Freddy the performance artist, then it is also true that I have bumped into Plastic Freddy. Representing, or thinking/talking about, is different in this respect. When Oedipus thought longingly about Jocasta without realising that Jocasta was his mother, there is a sense in which he was not thinking longingly about his mother. Representing someone as Jocasta is not the same as representing them as one's mother even when Jocasta is one's mother. (The example comes from the ancient Greek tragedy, Oedipus Rex. Oedipus unwittingly kills his father so as to marry Jocasta, taking out his own eyes on discovering his mistake.)