The DP Hypothesis in Generative Syntax | Noun Phrase Structure

[00:00:05]Now we're going to look a bit more in-depth at the structure of these noun phrases.

[00:00:11]And in particular, where it is that we're going to put these elements, these determiners, that frequently occur at the left edge of the noun phrase.

[00:00:20]So elements like "the" or "a" or "this" or "these" or "those" or "that".

[00:00:26]A question that any syntactician would ask about these is where exactly in the structure they are.

[00:00:34]And there are two hypotheses that one could entertain.

[00:00:39]The first hypothesis would be to suppose that that determiner is in the specifier position of the noun phrase.

[00:00:48]One reason one might think that the determiner is in the specifier position of the noun phrase is that it is in complementary distribution with these subject-type noun phrases, the possessors inside the noun phrase which look as though they should be inside the specifier position.

[00:01:07]So we get things like "the girls' admiration of the picture" or "the admiration of the picture" and we see they're in complementary distribution, because you can have one... that is to say you can have one or the other, but not both together.

[00:01:23]So you can have "Michael's book" or "the book", but you can't have "Michael's the book" or "the Michael's book" in English.

[00:01:31]So determiners and these possessive noun phrases are in complementary distribution, and the typical explanation for complementary distribution is that there's a single position, and these two different items are competing for that position, so you can have one but not the other.

[00:01:47]So that would give us a structure where the determiner is in the specifier position of the noun phrase.

[00:01:54]But if we just put the determiner on its own in the specifier position of the noun phrase, we actually have a structure which is not consistent with x-bar theory as we've been looking at it so far, because what our x-bar schema has given us is structures where you have the head of the phrase, it may or may not have a complement, it may or may not have a specifier, but those... the complement and the specifier, where they occur, are phrases; not just heads but full phrases.

[00:02:25]So we can't just have the determiner as the specifier of the noun phrase.

[00:02:30]One way we could deal with just is just to say "this is a case where the determiner, it's just a limit case, where the determiner is projecting a whole phrase, a whole DP, it just happens that it doesn't have a complement and it doesn't have the specifier, and so the whole phrase consists only of the determiner.

[00:02:51]And that would resolve the problem for x-bar theory.

[00:02:55]But it has worried people that in all of these cases with these determiners, so with "the" and "a" and "this" and "that" and "those" and "these", all of these we're only seeing the determiner; we're not seeing it with any other material.

[00:03:10]So that looks a little bit suspicious if we're saying that it's a full phrase.

[00:03:16]And that's one of the reasons why people have come up with an alternative hypothesis about the structure of these phrases with nouns in them.

[00:03:25]The second hypothesis, then, that one might have about the structure of nominals and in particular as to where the determiner is, is to say that rather than having the determiner project a determiner phrase which is the specifier of the noun phrase, instead, to suppose that the determiner is the head that takes the noun phrase as its complement.

[00:03:51]The effect of that would be, if we're sticking with x-bar theory, would be to say that the structures which up until now we've been calling noun phrases are in fact determiner phrases.

[00:04:03]They're projections of the determiner.

[00:04:05]That determiner takes the noun phrase as its complement.

[00:04:10]The idea that what we've been calling noun phrases up till now are really determiner phrases, that is that the determiner takes a noun phrase as its complement, and projects a determiner phrase, has been called "The DP Hypothesis".

[00:04:24]If that hypothesis is correct, we actually have a very close correspondence between the structure of sentences and the structure of noun phrases.

[00:04:33]So in the case of sentences, they're built up around verb phrases... we have a verb which projects a verb phrase, and then that verb phrase is the complement of a functional head "Inflection" or "Infl" or "I".

[00:04:50]In the case now of the DP, so in the nominal domain, we have a noun phrase projected by a noun.

[00:04:59]That noun phrase is then the complement of a functional head, a D in this case, that takes the noun phrase as its complement.

[00:05:08]So the noun corresponds to the verb, the D corresponds to I.

[00:05:13]We have a functional category taking a projection of a lexical category as its complement in both cases.

[00:05:22]At this point, we should go back and consider how we're going to analyse cases where instead of just a determiner, we have a possessor.

[00:05:31]We could say, or we might think of saying, that well, we have evidence for the DP hypothesis where we've got a determiner like "the" or "this", but in the case of where we have a possessor, that that possessive phrase is in the specifier of the NP; that's how we started out.

[00:05:47]But that would mean that where we have the possessor, we have a whole NP, whereas when we have the determiner, we have a DP.

[00:05:57]That would mean that where we have "Mary's collection of mushrooms" with a possessor, that would be an NP, and "the collection of mushrooms" with a determiner, would be a DP.

[00:06:08]That would give you different categories.

[00:06:09]But that's the wrong answer, because those two cases have exactly the same distribution.

[00:06:16]So phrases where you've got a possessor and phrases where you just have a determiner, overall they have the same distribution, and we want to say they have the same category, so we'd like them both to be DPs.

[00:06:29]How is that going to work out in the case of possessors?

[00:06:32]Well, one hypothesis here is to say that that clitic "'s", the thing that expresses possession, to take that to be a D. That would mean that in the case of "Mary's collection of mushrooms", that "'s" would be the D, where Mary would actually occur in the specifier position of the DP.

[00:06:59]This analysis would still give us an explanation for the complementary distribution for determiners like "this" and "that" and possessives like "Mary's".

[00:07:09]The whole possessor is not occupying the same position as the determiner, but that little "'s" element is.

[00:07:16]So you can have one or the other, but not both, so we still retain that explanation.

[00:07:22]It's also completely consistent with this that the possessor doesn't have to be just a word like "Mary"; that specifier position is a position for phrases, and indeed we can have full phrases there, so as well as "Mary's collection of mushrooms" we can have something like "My friend's collection of mushrooms" as well as "Joanna's hat" we can have "the Duchess of Cambridge's hat", so the possessor is, as we expect under this analysis, can be a full phrase itself, because it's occupying a specifier position.

[00:07:58]We've seen that, just as in verb phrases, so too in noun phrases, it seems that there can be arguments of the head.

[00:08:08]So, a verb like "hate" takes two arguments: an experiencer and a theme.

[00:08:16]And the experiencer is what appears as the subject, whereas the theme is what appears as the object.

[00:08:24]So in a sentence like "the cat hates the dog" we can only interpret it as "the cat" experiencing some emotion with respect to "the dog"; this sentence doesn't tell us anything about how the dog feels about the cat.

[00:08:38]In exactly the same way, when we have an nominal, we have something like "the cat's hatred of the dog", and again that initial noun phrase expresses the experiencer, it's the cat that's experiencing the emotion, and it's the dog that's arousing the emotion, and that's fixed for those two positions.

[00:09:01]In the case of verb phrases, we captured this by saying that lexical items are stored in the mental lexicon of speakers along with the information about the phrases that they select, and those phrases that they select are also phrases that they assign thematic roles to.

[00:09:23]So that information is found in the lexical tree associated with each item.

[00:09:30]So we've seen that there's an elementary tree for "hate" that has a complement, we would now say "DP", and that would express the theme, the thing that is hated, and a specifier DP, and that is going to be the position for the DP which expresses the entity that experiences the emotion.

[00:09:56]Since we see exactly those same arguments in the case of the nominal, we'd want to say that the noun "hatred" also has an elementary tree where we see the phrases that it selects.

[00:10:11]In this case it selects the PP headed by "of" as its complement, and then it also selects a specifier which is what expresses the experiencer in this case, that's where you're going to find the DP corresponding to the person or entity experiencing the emotion.

[00:10:30]So that will give us an elementary tree for the noun phrase with a specifier and a complement.

[00:10:34]We've seen, then, to build up a sentence, you construct it then from these elementary trees.

[00:10:40]You've got the VP, then for the verb "hate".

[00:10:44]It's got slots for DPs, so you'll substitute arguments into those slots, and then we also have the functional head I. It has a slot for a VP, so those combine, and you have the I taking the VP as its complement.

[00:11:03]And then, as a final step, the specifier of the VP moves to become the specifier of the IP.

[00:11:13]We build up the nominal in exactly the same way, so you have an elementary tree for the noun, it has slots for the arguments (in this case a PP and a DP), so you put elements into those argument slots.

[00:11:28]Just as the VP then combined with the elementary tree associated with I, so this noun phrase is going to combine with the elementary tree associated with the determiner, with the D.

[00:11:40]So, those will then get... those will then combine and we'll have a structure for a... projected by the D, the D-bar with the determiner and the noun phrase in this case.

[00:11:55]But now it looks like we have a problem, because we've got the wrong order.

[00:12:00]We've got "'s", and then the subject and then the rest of the N-bar.

[00:12:10]That is to say, we've got something like "'s cat hatred of the dog", and that's clearly the wrong order, but if you think about it that's exactly the same problem that we had in the case of the sentence.

[00:12:23]We had reason to think that the subject of the sentence originates in the VP, but that puts the subject on the wrong side of that inflectional element; it would come after a modal like "will", whereas we've seen that it comes before.

[00:12:35]The way that we solved that in the case of the VP is to suppose that the subject of the VP moves to become the subject of the IP, and we can use exactly the same solution in the case of the nominal.

[00:12:47]So we'd say that the subject of the NP moves to become the subject of the DP, and that then will give us the right order for the DP, and it will also give us a tree which reflects the selectional properties of the heads in that structure, the fact that that possessor noun phrase is actually an argument of the noun.