I’m reading around the topic of the normativity of meaning at the moment (by which the mere civilians among the readership of this blog should understand the rulishness of language and the nature of the implicit standards of correctness that apply to language). Being philosophy, much of the current debate proceeds without any helpful examples, but we could take well known disagreements over the correct usage of ‘disinterested’ or ‘refute’ as examples (properly speaking, these words do not mean the same as ‘uninterested’ and ‘deny’ -look it up!).
Now the issue at the heart of current thinking on the normativity of language is whether semantic norms are properly action guiding or prescriptive. It is the prescriptivity of moral norms which is thought to be difficult to accommodate within a materialist naturalistic world-view. Moral norms have ‘to be doneness’ and ‘not to be doneness’ built right in to them, and this seems to contrast with plain old ordinary facts which just ‘tell it like it is’ they don’t of themselves say what is to be done about it.
Anyway, one suggestion with norms of meaning is that there must be norms of meaning because there are standards of correct use. Given what the word means, you can correctly apply ‘green’ to something if and only if it is green. You can correctly claim to be ‘disinterested’ in the proceedings, only if you have no (financial) stake in them and so on. But it has been objected to this proposal that these are but mere semantic facts. They state what you can and cannot do with words, but not what you ought to do with them. For instance, if you have an interest in the proceedings, it is perhaps encumbent on you morally not to say you are disinterested in them, but this is not a semantic obligation. If you want to lie, then you should say you are disinterested. In other words, the philosophers who push this line (Anandi Hattiangadi, Paul Boghossian, Asa Wikforss and Katrin Gluer) claim the relevant semantic standards are not normative because they are not prescriptive. The prescription to tell the truth and not lie is a moral one that bears on how we should speak and the semantic standards of correctness themselves, to the extent that they seem prescriptive, merely inherit this prescriptive force from contact with these other norms. The sceptics are quite right about this. We should not look for the source of the putative normativity of meaning in the moral prescription to tell the truth. But actually, this is not the most plausible possible source of normativity anyway.
It’s worth dredging up the examples, because they show that many people take a prescriptive attitude toward semantic facts. Various people, such as Lynn Truss, H W Fowler, William Safire and others have earned fame and fortune categorising linguistic mistakes, such as the difference between ‘infer’ and ‘imply’, ‘between you and I’, where to put an apostrophe, or what the phrase ‘to beg the question’ actually means.
On the other hand, professional linguists take a notoriously latitudinarian line on these matters. Partly, I think, linguists do this because they take the prescriptivist arguments to be based in unsound historical conjectures or from applying grammatical rules of Latin to English grammar. Partly, they do this because they don’t really think it’s possible for a native speaker to err about the grammar of their own language. They don’t really believe in languages as such, that is, as shared public communication systems with rules we must learn. For these linguists, we are born with a native language acquisition device which is ‘calibrated’ through exposure to the utterances of people in our environment to produce roughly similar utterances in our turn. But this situation is one of closely resembling individual idiolects, and not one common public language. On this picture, you can’t really get anything wrong since, when your language faculty is functioning in favourable conditions, what you say basically defines your idiolect as such.
As far as I see from my reading so far, the sceptics about semantic normativity (i.e. Wikforss et al.) don’t consider actual prescriptivist causes like ‘between you and I’. Instead, they consider principles like ‘if ‘horse’ means horse, then ‘horse’ is correctly applied to an object, x, if and only if x is a horse’. Sure, they claim, if I don’t apply ‘horse’ to all horses and only to horses, I don’t apply it correctly, but so what? Why should I apply it that way? Moral reasons? Prudential reasons? Instrumental reasons? All of these are external to language and the rule governing ‘horse. If the principle were normative, it would itself provide action guiding direction.
Now, I don’t know whether to call this a cunning rhetorical ploy or a large gap in the argument, but since they’re choosing examples that haven’t traditionally elicited any prescriptive sentiments, it’s impossible to tell whether we mutely go along with scepticism over a rule like the rule for ‘horse’ because linguistic rules are merely descriptive or because the prescription is so widely understood and endorsed that seriously diverging from correct use is not a credible possibility. (of course, the sceptic will say that if language is genuinely normative, it will be prescriptive for words like ‘horse’ too, so it shouldn’t matter what word we choose as an example. I accept this as a general point, but there seems to be a genuine issue as to whether we would recognise prescriptivity when it presents itself. Unless we can be sure we’d know it if were there, it does matter what example we consider).
The standard prescriptivist causes demonstrate that people take linguistic rules to be action governing. Divergence from linguistic prescriptions is standardly taken as a sign either of ignorance or stupidity. But people can be wrong. What gives these rules their authority? I mean, the sceptics may be right after all, because maybe these prescriptivist sentiments are just subjective and really only reflect inherent conservativeness. The game is just about up with ‘to beg the question’ after all. It is progressively harder and harder to maintain that it is wrong to take this phrase to mean ‘to prompt the question’. Language changes. ‘Sophisticated’ used to mean ‘sophistical.’ ‘Nice’ used to mean ‘precise.’ How is the evidence of change compatible with strict prescriptive rules?
Here’s an anecdote that bears on the general point. I visited Montreal for two weeks in 1994 in the middle of summer, around the time of the world cup soccer final and the Montreal ‘juste pour rire’ comedy festival. I was quite unprepared for how hot it gets there. My preconceptions of Canada were quite blown away. Anyway, I was going for a walk one day while I was there and decided to take my shoes off. It’s not at all remarkable in New Zealand to see people going barefoot, even in town. But in Montreal it definitely is! As I walked along, some guy drove past and shouted ‘where’s your shoes?’ (in English). I was gobsmacked. Who slows down and rolls down their window to shout at a complete stranger ‘where’s your shoes?’ I mean, what’s it to him? Why doesn’t he just mind his own business? Here we clearly have norm-enforcing behaviour. We have judgement and reproach, but do we have an actual rule like, ‘thou shalt not go barefoot’? Well, maybe, but it seems an odd sort of rule to me. The explanation I prefer is that people enforce things that aren’t rules and this shoe business was an example of it. And other examples would be the Lynn Truss apostrophe nazis, and all of us when we worry about declining standards of grammar and diction all round.
So my reason for thinking that language and meaning may not be normative is that language changes, and I can’t see how this dynamism is compatible with there being genuine prescriptions, and that the appearance of prescriptive attitudes among speakers is adequately explained by our tendancy to conformism, conservativeness and general antipathy to others’ being different.
That said, if there’s something to semantic normativity, then it lies in our belief that others should be like us and talk like us. If this is right, obviously linguists who privilege individual idiolects over public languages won’t find it. And philosophers who consider languages as abstract objects (set theoretical mappings from vocabularies to objects and sets of objects) won’t find it, since if it is there, the normativity lies not in the constitutive facts about language but in the social fact that a group of people, to communicate, must choose one out of the many logically possible languages to communicate in.