I am something of a creature of habit, and therefore will stick to the same blogs for a long time. Lately though, I've been trying to expand my reading, and so I've been following links from the sites I frequent to language-oriented sites that I haven't visited before. In the process, I have discovered 1) Language Log really needs to update their links. At least half the sites on their sidebar are defunct. 2) Web 2.0 is significantly more difficult to pull off when the topic requires technical expertise. Grammar is a case in point. Many people think they know something about grammar because they studied a foreign language, or had a Catholic nun teach them 8th-grade English, or own a copy of Strunk and White. Such people assume they know what they are talking about when they give language advice, but, alas, they often overrate their expertise. The site Pain in the English acknowledges this problem, but I don't think their solution is particularly effective. If you go there, you will find a variety of questions about "gray areas" of English usage. The questions are moderated but not the comments. It's vaguely like Ask a Linguist, but anyone can respond. As the site's proprietors explain here,

The idea here is that interesting questions should inspire interesting answers and comments. As long as we quality-control questions, we should not have to quality control comments.

The virtue of that scheme is that the questions do tend to be legitimately interesting. Many of the comments, though, make me want to scream. Some contain sound advice, but others are laughably wrong. And the problem for the ordinary user is that there's no way to distinguish between the two categories of comments. This question, for example, is about subject-operator inversion in sentences like, "We must acquire funding. Only then can we achieve our goals." The best of the responses more or less get it right, although they don't tell the full story. (The inversion is triggered by the fronted adverbial, "only then.") But there are also assertions such as, "It's the subjunctive mood, expressing a condition of unreality. We teach it in Spanish as well." There are two problems with this comment. First, the sentence isn't in the subjunctive mood. Can is a modal verb, of course, and so the sentence is not a plain indicative, but that doesn't make it subjunctive, no matter what mood would be used in Spanish. The English subjunctive, in the traditional formulation, includes the so-called present subjunctive, as in, "The teacher insisted that each essay be error free." And the past subjunctive, as in, "I wish she were happy." Some works, e.g., The Cambridge Grammar of the English Language argue that the second is not a true subjunctive, but that's not relevant here. No grammars I'm aware of consider can a marker of the subjunctive. (As an aside, I've noticed that teachers of foreign languages--at least of the romance languages--often assume that grammatical categories map transparently from the language they teach onto English. This may ultimately stem from the slavish adherence of traditional English grammar books to the categories of Latin grammar, but it could also be simply that the only grammar they learn is that of the foreign language, since formal grammar has largely disappeared from native-language English classes.) The second problem is that the remark is no explanation at all for why the inversion occurs in the sentence. Even if we revised the statement to say that the sentence used a non-indicative mood, that isn't why inversion occurs. We find the subject-operator inversion even in a sentence with a plain indicative mood: (1) Only then was he able to achieve his goals. Now I certainly don't object to forums where everyone can help answer such questions. But to be really useful, we need to have some way of sorting out the helpful explanations from the bad ones. And someone who knows little about English grammar has no real way of evaluating assertions such as the one above. What a site like Pain in the English really needs is a moderation system. I think it would be interesting to try to combine the general rating systems like the one on Digg with moderation by a pool of vetted experts. Have two independent ratings, one by people with demonstrated technical competence, and another by general readers. Each rating would act as a check on problems that the other group might not be sensitive to. After all, experts have a tendency to go into more detail than general readers want or need, and general readers have a problem distinguishing the accurate answers from the crank or discredited ones.