To: Gordon F.
From: Geoffrey Klempner
Subject: Quine's ontological relativity and the nature of concepts
Date: 12 June 2006 08:47
Thank you for your email of 31 May, with your third essay for the Philosophy of Language program, inspired by the question, 'What is the significance of Quine's doctrine of 'Ontological Relativity'? How does it contrast with 'linguistic relativism'?'
I won't criticize you for not 'answering the question' as your essay is merely 'inspired' by it rather than being intended as a response.
I would not be at all surprised if there was, or could be interesting science to do on the way the brain is organized in relation to concept possession. Nor would I be surprised if there were interesting and theoretically useful correlations between what one could say about human beings and what one could say about, e.g. dogs. (Incidentally, in relation to your remark about the fox, it is extremely important for animals to be able to recognize members of their 'kind' - dogs rarely, if ever, attempt to mate with foxes.)
This would not lead me to conclude that dogs have 'concepts'. What may be two examples of a 'kind' from the point of view of neuroscience (neural templates) may be very different when viewed from the point of view of the philosophy of language. There is nothing surprising or remarkable about this. Each science, including philosophy, seeks to carve its subject 'at the joints' to use Plato's pregnant metaphor, but the very same things may have different 'joints' depending on which science is dealing with them.
However, we are not primarily concerned with the question whether animals have concepts, or whether, or to what extent language is necessary for concept possession.
Although neuroscience and the philosophy of language are two different disciplines, one would of course expect illuminating connections or even explanations. A priori, I don't know whether this will be the case or not. What is a mistake - and I think this may be a mistake you may have fallen into - is to assume that there must be a 1-1 mapping from one science onto the other.
One of the most important lessons from the later philosophy of Wittgenstein, and also a view widely shared by philosophers who would not describe themselves as 'Wittgensteinian' is the doctrine of *externalism* about meaning. On this view, concepts are not 'in the head', as Hilary Putnam bluntly puts it.
Consider the example of identical twins (unit 9). You know 'that nice man' as the helpful neighbour who recently moved in to the large white house at the end of the street. However, unknown to you, the man lives with his identical twin brother. In terms of your neural template, either of the twins would satisfy the conditions for being, 'that nice man' although you would of course revise the template if you discovered your false assumption.
On an externalist view, prior to discovering your false assumption, when you subsequently think or talk about 'that nice man', you *succeed* in referring to the man you met even though your neural template would have been exactly the same had you met the other twin instead. Your 'concept' of 'that nice man' is not just something in your head, it is complex pattern that includes your past behaviour and causal interactions with the world.
Again, I don't want to go too far with this because this is not the topic of the essay. The point is that it is wrong to assume that because it is the structure of our brains that enables us to 'refer' to things or 'mean' something, it follows that 'reference', 'meaning' can be explained purely at the level of the individual subject, without any reference to what is outside.
However, if the assumption is discarded, then the theory of neural templates could never be a substitute for, or successor to a theory of meaning.
A concept is much more than a mere recognitional capacity. Let's level the 'externalist' issue aside and consider how concepts work. Your description at the beginning of your essay gave the impression that a neural template is a kind of mental image, a picture which we compare with things we find in the world. However, there are many concepts which are not associated directly with a recognitional capacity. Concepts are embedded in a framework where each concept not only has 'criteria' for its application but also 'consequences' in terms of the other things that it might be part of a criterion for. This is fully consistent with the findings of neuroscience, and also with Quine's observation in 'Two Dogmas of Empiricism' about the way our conceptual scheme interacts with the world or experience as a whole, rather than individual bits being associated with individual bits of experience or parts of the world.
What externalism adds to this picture is that a conceptual scheme cannot be considered solipsistically. We share a conceptual scheme as users of a common language, irrespective of the fact that we do not all share the same knowledge or beliefs.
Quine is responsible for putting forward a sceptical challenge about meaning, which in 'Word and Object' he called 'indeterminacy of translation'. There is no way, even in principle, that we can rule out alternative translations from one language to another. Later, in response to Donald Davidson's contribution to the debate, he distinguished between two levels of 'indeterminacy', the level of individual concepts (indeterminacy of meaning) and the level of the basic apparatus for individuation and identity (indeterminacy of reference).
Ontological relativity is about the indeterminacy of reference. The only way to speak of the 'ontology' of an individual speaker or group of speakers is from the point of view of another language. We imagine that when we say 'rabbit' and look at a rabbit we have achieved a successful 'hook up' between a linguistic item and a physical piece of the world. On Quine's view, 'the world' and the 'pieces' that compose it is whatever our language says. In my example of a rabbit, I used the term 'rabbit' twice, once in quotes and once out of quotes. If I had picked the rabbit up in my hands and shaken it, I still would not have succeeded in fixing on the 'thing' independently of language.
What would the theory of neural templates say about this? Let's assume that there is a neural template for the word 'rabbit'. The very same physical structure, Quine would argue, would work just as well for 'collection of rabbit parts' provided that correlative adjustments were made in 'interpreting' the neural templates for the terms 'is similar to', 'is part of', 'is identical with'. Neural templates only work in a structure where they are connected with other templates. They 'interact with the world or experience as a whole'.
All the best,