Wikipedia gets it half right
« previous post | next post »
I've been getting some mail about tricky cases involving the choice between who(ever) and whom(ever), in particular the question of which form to use in examples like
(1) Whoever/whomever you meet there is bound to be interesting.
versus
(2) Whoever/Whomever meets you there is bound to be helpful.
There are two factors at work here, one of usage and style and one of syntactic structure, but the big point is that for many speakers and writers. whomever is allowed (or required) in (1), but not allowed in (2). The Wikipedia page on Who gets this right, and correctly attributes the choice of whomever in (1) to the fact that the pronoun is the direct object of meet there. But it also says that whomever is the SUBJECT of is in (1), which is downright bizarre — and was absolutely baffling to my correspondent Ethan. And at first, to me, though now I think I now know what's going on.
First background fact: there are two now-standard systems for distributing the variants who(ever) vs. whom(ever) in English: an older system (system A), prescribed (often sternly) by many authorities, in which the m-variants are used whever the pronoun serves in the syntactic function of OBJECT (of a verb or preposition), and a more recent system (system B), in which who(ever) is the default, with the m-variants used only when a WH pronoun not only serves in the object function, but also is in a phrase with its governing element (a preposition, in the most common situations): in system B, the m-variants are used only with fronted prepositions (to whom). The two systems are reasonably well described in the Wikipedia article and many other places, including a Language Log posting of mine.
The second background fact is that a great many speakers use both systems A and B, distinguishing the variants on the basis of formality, emphasis, prosody, and the like. This posting is not about the choice between A and B — I'll take this up in a future posting — but about how system A works in a complex case.
Here's the Wikipedia page on this case:
in
• Whomever you meet there is bound to be interesting
‘‘Whomever’’ is in the accusative because it is the object of meet (cf. You meet her [or him]), even though it is also the subject of is (cf. She [or he] is bound to be interesting.)
This is half right. The WH pronoun is indeed the (direct) object of meet, fronted within the "free relative clause" whomever you meet there (and you is the subject within this clause), and that's why in system A the pronoun is whomever and not whoever (system B has whoever here). But the claim that whomever is the subject of is (that is, the subject of the main clause) is, on the face of it, preposterous: if the WH pronoun is the subject of the sentence, it should be whoever in system A (as well as system B). There's a conflict there — which is resolved by denying the claim that the WH pronoun is the subject of the main clause.
Instead, the whole free relative clause is the subject of the sentence, as you can see by looking at, among other things, how Subject-Auxiliary Inversion (SAI) and Subject-to-Object Raising (SOR) work for this example:
SAI: Is whomever you meet there bound to be interesting?
SSR: I expect whomever you meet there to be bound to be interesting.
So why did the Wikipedia writer(s) pick out whomever as the subject of the main clause?
Because of an unexamined theoretical assumption about syntax, namely that syntactic functions are functions of individual WORDS. The tradition of "school grammar" (with antecedents going back to Greece and Rome) sees syntax primarily as a matter of properties of individual words and relationships between individual words. So a subject is a single word; school grammars will then talk about the derivative concept of the "complete subject", which is the subject word in combination with all the words dependent on it (and all the words dependent on them, etc.), but the central and primary concepts are those of word properties and relations. (Most systems of "sentence diagramming" encode these assumptions graphically.)
Against this tradition, the idea of "(immediate) constituent structure" or "phrase structure" involved a major shift in viewpoint (although there are precedents from before the 20th century): properties of and relationships between linguistic expressions of any size, from words to multi-word phrases to clauses, became the central concepts in syntax, though properties and functions of head words of expressions continued to play an important role. In any case, with a very few exceptions, syntacticians now see the syntax of a language as primarily a matter of how these larger expressions (usually called "phrases", but embracing single-word phrases and expressions of clause rank) combine with one another, and not primarily a matter of how words combine with one another. This shift in view allows us to make sense of a great many phenomena that are hard to understand if we're focused on individual words.
But school grammar went on pretty much untouched by the development of phrase-structural and constructional ideas in linguistics. And that's where we are now.
Back to Whomever you meet there is bound to be interesting. If the subject of the sentence has to be a single word, what are our choices? Well, whomever, you, meet, and there (or some "understood" word, like person; school grammar can get a lot of mileage out of understood elements). If we stick to the actual words in the sentence, only whomever and meet are contenders: whomever if whomever you meet there is seen as consisting of a head whomever plus modifying elements; meet if whomever you meet there is seen as a clause, in which case it consists of a head verb plus syntactic arguments (you and whomever). But the subject of a sentence is (by definition in school grammar) a noun or pronoun (though in actual fact, things other than NPs can be subjects). So whomever it is!
An unfortunate consequence, but that's where traditional assumptions lead you.