Duck has already covered Hacker's "Passing By The Naturalistic Turn: on Quine's cul-de-sac" in broad strokes, and done a generally satisfactory job of it. I just intend to prod at some particular passages that annoyed me, or which caused me to want to write something.
In the USA it is widely held that with Quine’s rejection of ‘the’ analytic/synthetic distinction, the possibility of philosophical or conceptual analysis collapses, the possibility of resolving philosophical questions by a priori argument and elucidation is foreclosed, and all good philosophers turn out to be closet scientists.
If philosophy is supposed to be "continuous with science" then what can it mean for some philosophers (and not others) to be "closet" scientists? I do not think Quine's claim was that the distinction between the subject-matters of philosophy and science was vague, but that there just was not any: The difference between philosophy and chemistry is the same as the difference between chemistry and physics -- they are in different areas of a university. (I believe Quine once referred to the distinction between philosophy and the natural sciences as "Useful to librarians, but not to philosophers" but I can't place the quote. I suspect it was somewhere in Quiddities.)
Hacker notes that Quine himself "offered an account of analytic truths" in The Roots of Reference. It's two pages long, ps.79/80. The upshot is that an "analytic sentence" is a sentence which one learns to use just by learning to affirm it. I quote the close of the passage, to give the flavor, the book is searchable on Amazon:
In Word and Object I defined a stimulus-analytic sentence as one to which every speaker is disposed to assent. The analytic sentences in the present sense are a subclass of those, and a somewhat nearer approximation to the analytic sentences uncritically so called. Even so, we have here no such radical cleavage between analytic and synthetic sentences as was called for by Carnap and other epistemologists. In learning our language each of us learns to count certain sentences, outright, as true; there are sentences whose truth is learned in that way by many of us, and there are sentences whose truth is learned that way by none of us. The former sentences are more nearly analytic than the latter. The analytic sentences are the ones whose truth is learned in that way by all of us; and these extreme cases do not differ notably from their neighbors, nor can we always say which ones they are.So there can be no dividing up of sentences into analytic and synthetic, since a) there are gradations of approximation to analyticity and b) it is not clear how to decide whether a sentence a speaker holds true is one whose meaning he learned while learning to assent to it, or if further inference (possibly involving non-analytic sentences) was needed to decide the sentence's truth (which he is presently disposed to assent to simply upon hearing the sentence). (Suppose you learned that Larry, Moe, and Curly were called "bachelors" by being taught to assent to the novel sentences "Larry is a bachelor", "Moe is a bachelor", and "Curly is a bachelor", which sentences are used to explain to you why the three share an apartment and throw wild parties each weekend. And suppose you knew they were men through their appearance. Then suppose that you realize that they are unmarried when you ask them where their wives are and they laugh at you. You might come to realize that "bachelors are unmarried men" at this point, depending on what else you'd picked up about the usage of the words. By Quine's standard, this sentence would not be analytic in your mouth -- you learned to assent to it some time after you were first able to form it.) So any attempt to separate sentences into the two categories of Hume's fork is going to fail -- there will be cases where it's just not clear which tine applies. And so any attempt to have philosophy consist in "clarifying" propositions partly by mean of Hume's fork is a dead program. It seems reasonable to deny that Quine is backsliding in Roots of Reference, contra the implication in Hacker's article on page two.
It also strikes me as worth noting that Quine handles the question of the analyticity of logical statements (on p. 80 of Roots of Reference) by just taking the highest common factor between quarreling logicians: The law of excluded middle "should be seen as synthetic" because denied by intuitionistic logicians, while "that an alternation is implied by its components" is declared analytic, because it is not a subject of disagreement. But there are paraconsistent logics which do reject the inference from A to AvB (which allows them to hold onto Disjunctive Syllogism as a rule of inference without risk of Explosion). So it would seem that Quine is wrong to count this as analytic -- there are logicians who find reason to object to it, and logics which disallow it. I take this to show that Quine's manner of distinguishing analytic from synthetic logical statement is ad hoc and pointless -- "analytic" here just means that there has not been a non-classical logician clever (or bored) enough to draw up a logic which denied some principle or other. Quine's just conjured up a useless standard for being "analytic" that lets him say (with many philosophers from the tradition) that logic is analytic. But this is mere wordplay. Quine is trying to find a way to talk in the traditional way despite having ruled out all of the plausible ways to actually get away with this.
And in respect of a priority, what goes for mathematics and logic goes too for such propositions as ‘red is more like orange than like yellow’ or ‘red is darker than pink’.
I should have thought that red and yellow were more alike than red and orange: red and yellow are both primary colors, while orange is a secondary color. I don't even know what to make of this example. It's weird. Perhaps schoolchildren are taught their colors differently in Britain, as they are taught to pronounce "Zee" as "Zed", and part of this is that colors' "likeness" to one another is measured by their spacing on a color-wheel. (So contrastive colors are maximally unlike, and the three primary colors are all equally like and unlike one another.)
Certain reds are not darker than certain pinks. Looking at my bookcase, I see two books whose spines are a soft red and hot-pink, respectively. The hot-pink looks darker to me. Is hot-pink not pink? Or consider the red of red cellophane -- is this darker than the pink of a healthy pink carnation?
I shouldn't think it would be hard to produce color-swatches which matched the shades of red and pink I have in view. In which case what Hacker says is false, or at least not true in all cases of "red" and "pink" -- if he wants to talk about colors which are somehow impossible to illustrate with color-swatches, then I don't know what he is talking about.
Is it a problem if a philosopher's examples of a priori truths appear to be empirically false? What does it tell us about the a priori/a posteriori distinction if we are not all agreed on where to draw the line -- is the a priori still a "purely" conceptual matter if we are not agreed on its extent?
Hacker: "Knowledge that Jack is taller than Jill is categorially unlike knowledge that red is darker than pink." I don't know what a "category" is supposed to be here, or how they are supposed to be distinguished. I had to look to confirm my suspicion that Hacker's pink-generalization was wrong, and I should have to look to see if Jack was shorter than Jill. What is the point in drawing up "categories" and apportioning various bits of knowledge to different ones? In one sense I can perfectly well understand the point of having "categories" for different areas of knowledge: Chemistry, psychology, biology etc. are separated as a division of labor tactic -- new departments are formed as certain areas of study start to take up too much of the resources of the previous department, and their interrelations shift as convenient (thus the development of chemistry and physics as separate disciplines towards the end of the nineteenth century, and the blurring of chemistry and biology with the arising of organic chemistry; one could also note the development of psychology and computer science as disciplines, both of which split from what had previously been called "philosophy" -- as "natural science" earlier developed out of "natural philosophy"). What the epistemological import of these accountant's divisions could be supposed to be, I haven't the foggiest. Mathematics has the pedigree of having been privileged by the Greeks, but I don't see that there is a firm boundary between recent progress in mathematics and theoretical physics or computer science -- and where there are researchers whose work covers multiple fields, I don't see that it makes any sense to suppose that they are mixing and matching a little bit from each discipline, which might be untangled by a philosopher (and then "represented perspicuously" by a tabulation of which of their practices was of which sort, and then whether or not each of their utterances had a sense).
There are distinctions in how we are able to make sense of things like rocks, things like pigs, and things like philosophers. The rock has no aims; the pig has piggish aims; the philosopher aims at doing what is right. But I don't think that these sorts of categorical distinctions can be what Hacker wants; at the least, I don't see that these sorts of distinctions include a priori/a posteriori, conceptual and empirical, grammatical and natural-scientific, which seem to be more the type of thing Hacker wants to trumpet.