In order to make sure that the rules of X-bar theory don’t overgenerate and produce structures for sentences that don’t exist (in English, or any other language), we introduced the Theta Criterion. This addition to our theory of grammar matches up the noun phrases in a sentence with the number and kind of participants the verb needs to form a complete thought. The Theta Criterion blocks the sentences in (1a) and (1c), while allowing (1b).
(1a) *Ellie gave to Carl
(1b) Ellie gave her book to Carl
(1c) *Ellie gave her book her button to Carl
And this process of verbs assigning θ-roles to the noun phrases that appear next to them in sentences seems to apply locally, according to the VP-Internal Subject Hypothesis (VPISH). That is, the constituents that participate in the action represented by the verb begin life inside the verb phrase. Then, depending on what type of sentence they’re in — and which language — they either stay put or move up into a higher part of the tree. This accounts for things like floating quantifiers and VSO word order.
But it isn’t only verbs that need to be paired up with other phrases, for the sentence as a whole to make sense. More specifically, it isn’t only verbs that need subjects; plenty of other kinds of phrase need a subject too, including adjectives, nouns, and even prepositions! Consider the copular sentences below, which use the linking verb “be” to connect subjects to all kinds of different categories of phrase.
(2a) Carl is [AP grumpy]
(2b) Ellie is [NP a kind woman]
(2c) Russell is [PP on the porch]
None of these would mean anything without a subject. Fundamentally, this is because all of these phrases are behaving predicatively. That is, they express properties which can come out as either true or false when applied to an individual (like Carl or Ellie). Like ordinary verbs, they come packaged with a need for a subject.
Now, if we suppose that all predicates are created equal, it might be that the subjects in these sentences are generated low, too. We could generalize and extend VPISH, to become the Predicate-Internal Subject Hypothesis (PISH). It’s a nice idea, since it jibes with the notion that within and across languages, sentences are all mostly the same. It would be even nicer if we had evidence to back that idea up.
Luckily, we do! In the form of what are called small clauses. To get an idea of what small clauses are, let’s consider clauses in general. Put simply, a clause is a combination of one subject and one predicate. In English, this is usually a noun phrase, followed by some verb phrase. Of course, clauses often have more going on inside them than that. Take the sentence in (3), which involves more than just the subject “Kevin” and the predicate “scared.” It also contains the modal verb “will,” which usually locates an event in the future, the auxiliary verb “have,” which picks out the end point of an action, and the auxiliary “be,” which marks this sentence as passive.
(3) Kevin will have been scared
To put it another way, this clause is a big clause, since it contains information about modality, aspect, and voice, superimposed on top of the subject-predicate pair. A small clause, by analogy, has none of that — no tense, nor inflection of any kind. So what does this look like? Well, consider the sentences in (4).
(4a) Carl finds [SC [NP Russell] [AP annoying]]
(4b) Dug considers [SC [NP them] [NP friends]]
(4c) Ellie saw [SC [NP him] [PP in the house]]
In each one, a noun phrase is paired up with some predicate, forming a clause which then behaves as the direct object of the main verb (i.e., “find,” “consider,” and “see”). These clauses are small, since they lack any kind of inflection.
And because they lack any inflection, it doesn’t really make sense to label them as Inflectional Phrases. But, if they’re not IPs, what are they? If we take the PISH seriously, we have an easy way of accounting for the structures in these sentences. The clause “Russell annoying” is actually an adjective phrase, “them friends” is a noun phrase, and “him in the house” is a prepositional phrase.
(5a) [AP Russell annoying]
(5b) [NP them friends]
(5c) [PP him in the house]
If each of the constituents above had been in a copular sentence, the subjects would have moved out, as in (6). But since there’s no IP near enough to move to in each case, the subjects stay inside the predicate.
(6a) Russell is annoying
(6b) They are friends
(6c) He is in the house
So PISH manages to unify the different kinds of clauses we find in English, while also describing the structures of sentences like those in (4).
By capturing the varying word orders found amongst the world’s languages, and tying together otherwise unconnected ideas, the Predicate-Internal Subject Hypothesis has come to form one of the cornerstones of modern syntactic theory. And though there are still a few tweaks to be made, you can bet we’ll be seeing more of it soon!
So how about it? What do you all think? Let us know below, and we’ll be happy to talk with you about where exactly subjects really belong, or other syntax topics. There’s a lot of interesting stuff to say, and we want to hear what interests you!
Previous Topic: Next Topic:
Sound Wave Hissy Fit Coming soon!