Martha ate a kumquat
Our simplest sentences so far have had a single verb as both the head and the only element in the predicate. Verbs that can do so and make sense are called intransitive verbs. Verbs that make sense only with some kind of object of their action specified are called transitive verbs. The person or thing to which the action of the verb is done by a transitive verb is a kind of complement of the verb (think "makes complete") called a direct object. These are not really distinct kinds of words but kind of sentence functions. What we really have are intransitive and transitive predicates. In the latter, the same kind of words and phrases (mostly NPs) that can serve as subjects can serve as complements--e.g., direct object.
Some verbs can be used in both intransitive and transitive predicates:
(1) George walks.
(2) George walks the dog.
(3) Sarah shakes.
(4) Sarah shakes her booty.
(5) Martha ate.
(6) Martha always ate a kumquat.
(7) The dog ate George.
(8) Her booty shakes.
(9) A kumquat shakes its booty.
In diagramming sentence (6), we have a predicate (VP) with a specifier (always), so our first division will be between that specifier and the rest of the predicate, the V-bar. That, in turn, is divided between the verb itself and its complement NP. And the NP is divided into a determiner and noun. Going the whole hog, we'll draw in real bars instead of using apostrophes, and we'll show the N-bar elements in noun phrases even when this adds nothing to the representation of the sentence at hand. We won't keep this up.
Transitive vs. Intransitive. The distinction between the transitive and intransitive uses is not usually a problem for native speakers. For one thing, modern English finds it easy to convert intransitive verbs to transitive uses, just as it can convert nouns to verbs and vice versa, creating new uses for old words:
(10) The teacher grammared the class into submission.
Exercise 1: Transitive vs. IntransitiveSay whether the underlined verbs are used in intransitive predicates or (B) transitive predicates:
Don't be the first in your block to use an intransitive verb in a transitive predicate or vice versa, for it can drive purists crazy. Our ability to decipher sentences depends partly on our knowledge of what the norms of language are, and violations of those norms calls attention to itself--as one might want to do with sentence (10), of course, for comic effect.. The norms which say that certain verbs are mainly used in transitive ways and others in intransitive ways are in a different category from the rules which govern whether a sentence makes conventional sense. We may say , for example, that grammar, George, and a rose all "stink," but only George goes well with "walks the dog," and only a rose goes well with "blooms." Even so, most grammarians would say that sentences like these are grammatical--probably nonsense, but grammatical:
(11) A rose is walking.
(12) George is blooming.
Linguists say things like this because they like to make a sharp distinction between the rules of syntax and the kind of meaning-based (semantic restriction which tells us that the subject NP of "walks" should be animate and mobile. One good reason for making this distinction is that we can, in fact, imagine occasions on which we might say sentence (11)--our love is a red, red rose and out walking her poodle--or sentence (12)--George is blooming under the tutelage of his wise professor. Another reason is that we can use the rules of syntax even when dealing with nonsense words--"'Twas brillig, and the slithy toves did gyre and gimble in the wabe." One gortch "stinks" and two gortches "stink" no matter what a gortch may be.
What, a Question. When we want to make a question out of a transitive predicate, the rule is same as for intransitive subjects. We use DO-SUPPORT in such cases as well:
But what if we are unsure what Martha ate? That means our starting sentence is equivalent to (14), but sentence (15) sounds odd at best:
(13) Did Martha eat the kumquat?
What we actually say, of course, is this:
(14) ?Martha ate what?
(15) *Did Martha eat what?
(16) What did Martha eat?
The process involved here is called WH-movement, and it has applied along with INVERSION to give us sentence (15). [I find it helps to think of this process as WHOOPEE-MOVEMENT, but this doesn't seem to have caught on.] It is part of how we get to all the WH-questions, including those beginning with how, which counts as sort of an honorary WH-word.. When the WH-word involved is the subject noun phrase, WH-MOVEMENT has the effect of cancelling out INVERSION as far as the surface result goes, so that we use sentence (16) rather than sentence (17):
(16) Who ate the kumquat?
(17) *Did who eat the kumquat?
(18) Who did the monster eat?
As sentence (18) shows, this exception to the rule that interrogative sentences show INVERSION applies only when we have a subject who. A who (or whom or what) that begins life as an object of the verb will still show INVERSION. Many of the other WH-interrogative pronouns (including how) can be thought of as beginning life inside the predicate as adverbials. We will also encounter this process at work again when we talk about relative clauses.
Diagramming WH-Questions. When we diagrammed questions earlier, our phrase trees were a bit less complex. We represented a question as having a Q-marker which took an independent clause as its complement. INVERSION was shown have moved the tense/modal element under the Q-marker, leave a t (for trace) behind:
When we diagram WH-questions, INVERSION operates the same way, but WH-MOVEMENT has also applied to move the WH-word in front of that tense, leaving another trace behind. We can think of it now as occupying a specifier position in a phrase headed by the Q-marker. In the diagramming conventions followed here, that phrase is labelled CP, and the position occupied by the Q-marker is the C which is its head. These abbreviations stand for complementizer phrase and complementizer, for reasons we'll touch on much later. In the meantime, we have a CP which WH-MOVEMENT has given a specifier and a C-bar consisting the the C (with the Q-marker) as its head and an IP as the complement. We can imagine sentence (19) as the question form of a declarative sentence like (20), though the interrogative what in (20) will almost always trigger question movements:
(19) What did Martha eat?
(20) Martha ate what.
The head of our CP in sentence (19) is a question-marker (+Q) occupying the complementizer slot, with sentence (20) as the IP which is its complement.
This IP in turn can be broken down into its specifier (its subject NP) and the rest of the sentence:
We break down the rest of the sentence into the tense element (I), which is marked for past tense, and the verb phrase which is its complement.
And as usual, the verb phrase is broken down into the verb and its complement NP:
The question marker triggers INVERSION, which moves the tense around the subject, and DO-INSERTION, which supplies a do to carry the tense. A trace element is left behind:
The what triggers WH-MOVEMENT, which moves that interrogative into the specifier slot for the CP, which we might as well divide into this specifier and the rest of the CP. This last movement brings us to our phrase structure tree for the sentence we started with.
How Much of This Will be on the
There are intransitive predicates, whose verb does not require a complement, and transitive predicates, whose verb has at least one complement, traditionally called a direct object. We have just begun our discussion of predicates, but by adding in WH-Movement we have now covered all the main interrogative sentence types, and you should be able to identify the form of almost any English sentence. Diagramming questions has forced us to learn another new phrase type, the CP. Fortunately, under X-bar theory, all phrases have very much the same structure, so adding one more shouldn't really terrify you.