August 05, 2004

Content clauses are not necessarily complement clauses

Andrea Lafferty, the executive director of the Traditional Values Coalition (a conservative religious organization) was recently quoted here by Brian Leiter saying something that provides an excellent illustration of the rationale for a terminological distinction made in The Cambridge Grammar of the English Language. Ms Lafferty said:

"There's an arrogance in the scientific community that they know better than the average American."

The Cambridge Grammar refers to finite clause constituents like that they know better than the average American as content clauses, taking the term from the great 20th-century Danish grammarian Otto Jespersen. We don't call them ‘that-clauses’, and we don't call them ‘complement clauses’, and there are solid reasons for both decisions. Ms Lafferty's quote provides a good example to illustrate why the second of those decisions is correctly made.

There are two reasons we don't call constituents of this sort ‘that-clauses’. First, that would be a parochial term rather than a universally applicable one: other languages have constituents of what appear to be exactly the same type, but in Spanish they're marked with que and in German they're marked with dass and in Hindi they're marked with ki and so on. Second, in many contexts the word that is omissible, and it would seem perverse to name a constituent after the one word in it that is freely omissible without any change in the construction (omit any other word from that they know better than the average American and you get either an ungrammatical constituent or at best one with a different meaning).

But we also don't call them finite complement clauses, though many linguists would. The reason is that content clauses are often complements, but not always. Notice that there is no way we can say in general that the noun arrogance takes content-clause complements: it just isn't grammatical to say something like *His arrogance that everything will be all right amazed me. (Try replacing arrogance by assumption and note the difference.) You might want to say that Ms Lafferty's remark isn't grammatical either, but it surely comes close, and it's fully intelligible (Brian Leiter quoted it and discussed its content at length; he didn't say it was garbled and he couldn't understand it). So set aside the question of whether it's perfectly grammatical, and just consider how we can talk coherently about its structure, for it certainly has syntactic structure. We can relate it to something found elsewhere if we note that occasionally utterances like this are encountered:

What's up with you, that you're looking so miserable?
You must have been sitting awfully quietly, that he could could come in there and not notice you.

What's important about such examples is that the clause after the comma is not subordinate, in the sense of having the function of complement to some noun, verb, adjective, adverb, or preposition that licenses it. Ms Lafferty's remark can be regarded as illustrating the same sort of possibility. Whatever the exact details of the structure, the point is that the constituent that they know better than the average American is a content clause but it's not functioning as a complement clause in Ms Lafferty's sentence (we're leaving the matter of how it does function to be determined by future research). So the property of being a finite complement clause is distinct from the property of being a content clause, despite the fact that nearly all content clauses function as complements.

Posted by Geoffrey K. Pullum at August 5, 2004 04:59 PM