If you want to say that "sat on" is a transitive verb, you need to say it is transitive in the active and in the passive. Voice has no bearing on transitivity.
Just to be clear to HeartShape, I am disputing what Phaedrus claims above. My understanding is that voice
does have a bearing on transitivity. This seems to be a disagreement of technical definition. (Phaedrus—I don't mean to disrespect your expertise but I need to be convinced that your deep structure analysis of the dependence of complements on verbs is not simply a function of valency, as opposed to transitivity.)
That passage is on page 56 of the edition I have -- the second (1963); the first edition was published in 1947. Isn't it weird that they (there are two authors: Pence and Emery) speak of a "preposition-like" word and then say, in their explanation of "laughed at," that "at" is a preposition? It's especially weird, I think, because on page 43 there is a note with the same example ("They laughed at me"), and they state there that "at" is an adverb.
Yes!
So I can't tell whether they want to say that, in "laugh at," "at" is a preposition, an adverb, or a preposition-like thing that has no name.
I suppose they are saying that in terms of its form it's a preposition but functionally/semantically it is more 'attached' to the verb than to its complement
me.
But it is clear that they want to say that "laugh at" functions as a single verb.
I'm not sure that 'single verb' is the right way of putting it but yes. I'd say it's more accurate to say that
laughed at can be seen as a
single divalent predicator. That is, it has two arguments (
They and
me). Semantically, it seems to me that it's like a compound of two predicators (monovalent
laughed and divalent
at). It's like the verb
laughed 'borrows' valency from the preposition.
Two problems come to mind. First, if "laugh at" is a single verb, it should not be possible to insert an adverb right into the middle of it, yet we can say things like "They laughed scornfully at me." Compare: "She sat down hard on the chair."
Why should that not be possible? Would this problem disappear for you if it were labelled a 'compound predicator' instead of as a verb?
Second, if "laugh at" is a single verb, and "at" is not a preposition with respect to the direct object that follows, it should not be possible to prepose "at" together with the object at the front of a relative clause. But it is possible to do that. Not only can we say, "I was the person whom they laughed at," but we can also say, more formally, "I was the person at whom they laughed."
Yes. Very interesting observation. It does sound strange to me now that
at whom they laughed is possible. I'm just thinking out loud here but I wonder if this word order could be more a vestige of Latin grammar translation than anything else. I mean, if the
at is indeed more 'attached' to the verb
laughed than it is to the object
me, then why would this separation be made? I'm going to conclude that the reason why
whom they laughed at feels more comfortable than
at whom they laughed (in most informal speech at least) is precisely
because of the compound connection between the predicators
laughed and
at (or in other words the integrity of the 'single verb').
I think their case would have been sounder had they steered clear of prepositional verbs and stuck with phrasal verbs:
He looked up the word.
*[strike]He looked quickly up the word.[/strike]
That is the word which he looked up.
*[strike]That is the word up which he looked.[/strike]
The reason why *
That is the word up which he looked is not possible is that the particle
up, although it adds meaning, has no valency. It cannot therefore connect with
which like
at could in the same position.
But it's interesting why
*He looked quickly up the word is not possible. It seems that the intrusion of the adverbial
quickly between verb and particle somehow interferes with the transitivity, where it doesn't when relocated elsewhere, as in
He quickly looked up the word.