I will write my opinions on this matter, though somewhat more disconnected from Grice/Strawson’s article. For more info on In Defense of a Dogma, please check both mentioned posts (which are great BTW). Anyway, some of my opinions are in this post, and probably there will be some more in the following…
First, I agree with Quine that analytic/synthetic distinction in a certain sense is problematic. That is, this distinction often goes (or went) together with a view which I think one can safely characterize as wrong (though I’m sure there will be people with different oppinion). In that (wrong) view, words like common nouns refer to concepts, and concepts are further reducible to a combination of some other concepts. So, for example, concept of BACHELOR is reducible to (or in its “structure” contains) concepts UNMARRIED and MALE. Different philosophical views can differ in the details here, one view might use for example genus/differentia view where MALE is genus and UNMARRIED is differentia (which as I was told by Brandon from Siris in another post, was the standard view in the times in which Kant worked and which came from “Wolff’s scholasticized Leibnizianism”); or might imagine the concepts as simply being a set of necessary and sufficient conditions in terms of other concepts (which I guess was standard for conceptual atomism of Russell and of first half of 20th century empiricism in general).
Anyway, to this structure of the complex concept which is made from simpler concept, in the realm of language there will be a corresponding definition. So, to the structure of the concept BACHELOR, there will be a corresponding definition “bachelor is an unmarried male”. In those views then, to learn what bachelor means is to grasp its definition, and we end up in situation where ‘bachelor’ means ‘unmarried man’. The judgment that any bachelor is an unmarried male, or the proposition that any bachelor is an unmarried male, are then taken as analytic, having in mind those particular types of theories. In those theories, analytic truths are also a priori, as the definition ‘A is B’ has to be grasped by the mind to know what A means, and once that it is grasped it never ‘leaves the mind’ so to say. Or we can say, that semantic facts which are important for the sentence being true, in the case of analytical sentences, given this theory, are in our possession, given simply that we are competent speakers of the language.
In this sense then analytic/synthetic distinction is a cover under which we have specific theory. It is in this sense (or with this burden) that I don’t think that distinction works as I don’t think those kind of theories work.
We can put forward another sense of ‘analytic’ though, where we can say that a sentence is true solely in virtue of certain linguistic facts.
To start with an example… In this sense ‘Hesperus is Phosphorus’ is analytic because both names refer to the same object – Venus. And that they refer to the same object is a part of the linguistic facts – which include the history of the usage of the names, going back to and including the baptizing of something which appeared as intentional content of intentional acts of people in certain linguistic community. That is, in simple words, in certain society people were seeing Venus in the morning sky, baptized it Phosphorus, were seeing also Venus in the evening sky and baptized in Hesperus. This with addition of the other linguistic facts of the usage of those names until their use today (in which use I take part with this post) is enough facts to proclaim the sentence true. We don’t need anything else.
In this sense however (to contrast it with that previous sense) linguistic and other facts needed to determine the truth value of the sentence doesn’t have to be possessed by us. By the way, that is why I didn’t use more standard “it is true in virtue of its meaning” – given the standard meaning/reference distinction, it doesn’t seem right to say that Phosphorus is Hesperus, if we limit ourselves just in the realm of meaning, and exclude the reference. The meaning might determine the reference, but that is further linguistic fact which doesn’t seem to be included in the meaning of ‘Hesperus’ or ‘Phosphorus’. So, instead, it is better to say that a sentence is analytic if it is true in virtue of certain linguistic facts. If you ask me what those linguistic facts would be, my first thought probably would be a) linguistic facts about the words used in that sentence and b) the facts about linguistic practices performed in certain way by certain combining of the worlds.
In this way analytic sentences don’t have to be a priori. ‘Analytic’ here is doing quite a different work than it was when related to the specific theories that were mentioned. There, analytic, as I can see, was supposed to serve to point to independence of the truth of the sentence from empirical matters, and it was either supposed to be one kind of a priori, or merely identical with a priori. However in this other sense, it is used to point to independence of the truth of the sentence from empirical matters other than certain linguistic facts.
Having given this place to the the analytic/synthetic distinction, is it the distinction to which Strawson and Grice point to when they say that that there is obviously difference between “My neighbor’s three-year-old understands Russell’s Theory of Types.” and “My neighbor’s three-year-old is an adult.”? I don’t think so, as they say that we can’t understand what the later could mean (which means I guess that we can’t imagine what could count as making this sentence true). So, what they are pointing, seems to me have more to do with some kind of a priori/a posteriori distinction.
Is the distinction that they are drawing related to the classical view of concepts? It might be.
Is the matter settled if we agree that Grice and Strawson were really defending a dogma which is wrong (related to the bad theories mentioned)? Not even close! Well, at least me as a good idealist, I have to believe that we can get to new truths about the world solely by thought, and express those truths and the conclusions through words of which we don’t know anything else but their meaning. Hope that I will be able to put something sensible on this in the next post.