grammatical foundations

Test your knowledge

Exercise 4

Give the X-bar structure of the following sentences and explain how the DPs receive thematic roles and Case.

(1)a John’s message arrived.

bDavid made the ball roll to the wall.

cDavid rolled the ball to the wall.

dJohn sank Jim’s boat.

eJim’s boat sank.

fJim’s boat was sunk.

gBill caught a bird.

hThe bird was caught.

iSam coughed.

jJohn sent a message to Mary.

kMary was sent a message.

lJim took his shoes off.

mJim took off his shoes.

nJohn thinks that Jim knows that Mary gave his book to Jane.

241

Chapter 7 - Complementiser Phrases

In (2a) the clausal complement of the verb know has obviously moved from its normal complement position behind the verb to a position at the front of the sentence. As the complementiser still precedes this clause, it can be assumed to form a constituent with it. In (2b) the pronoun it is used to replace the clause that Stan saved the world. As this includes both the complementiser and the IP, we conclude that these must form a constituent. Finally, in (2c) two clauses are coordinated. That both clauses begin with a complementiser demonstrates that the complementiser forms a constituent with the following IP.

That the complementiser is not inside the IP itself can be demonstrated by the fact that it is possible to pronominalise the IP without the complementiser and we can coordinate two IPs separate from the complementiser:

(3)a we expect there to be trouble, but we are not hoping for it

b I wondered if there would be trouble and I could be involved

The first case in (3) is fairly straightforward. The it replaces the non-finite clause there to be trouble and given that the complementiser precedes the it the pronoun replaces just the IP. The second case needs a little explanation. What is coordinated here are the two IPs there would be trouble and I could be involved. In some clauses the complementiser does not have to appear overtly, however:

(4)I think (that) his eyes were blue

One possible analysis, which we will argue in favour of a little later, is that when there is no overt complementiser, there is an phonologically empty complementiser. This possibility then raises the question of what is being coordinated in (2b): is it just the IP, as we claimed, or an IP with an empty complementiser?:

(5)a if [there would be trouble] and [I could be involved] b [if there would be trouble] and [e I could be involved]

In the second case, the data obviously do not support the claim that the IP can be coordinated without the complementiser as both sides of the coordination contain complementisers. However, (5b) cannot be the correct analysis as the only complementiser that is allowed to be empty in this kind of situation is that and if cannot be omitted:

(6)a I think (that) his eyes were blue

b I wondered *(if) his eyes were blue

Thus, (5b) would only be possible if the empty complementiser were a version of that. But this is not possible as, in this case, a clause beginning with if cannot coordinate with a clause beginning with that:

(7)*I wondered [if there would be trouble] and [that I could be involved]

Thus, the only viable analysis is (5a).

The final outcome of this discussion, then, is that the structure of the clause is:

(8)XP

C IP

244

The structure of CP

As the complementiser is a word and the IP is a phrase, we immediately see a potential head–complement relationship between them and if we apply X-bar principles to this situation the structure we expect is:

C©

C IP

If this is so, we expect the complementiser would demonstrate certain head-like properties and it is fairly easy to show that it does. For example, compare the following sentences:

(10)a I wonder [if Charles likes chocolate] b I think [that Charles likes chocolate]

The embedded clause in (10a) is interrogative as it can act as the complement of the verb wonder and this subcategorises for interrogative complements. The clause in (10b) is declarative as it can act as the complement of the verb think which subcategorises for declarative clauses. But the only difference between the two clauses lies with the complementisers. The IP in both cases is identical. This would suggest that the interrogative/declarative nature of the clause is fixed by the complementiser and not by anything inside the IP. In other words, it is the complementiser that provides the force of the sentence. As it is heads that provide the properties of the structures they head, this demonstrates that complementisers do have a head-like status within the clause.

Another claim made in (9) is that the IP is the complement of the complementiser. Can this be substantiated? One obvious relevant observation is that the IP follows the complementiser and as we know complements follow heads in English. Thus the claim that the complementiser is the head and the IP its complement accounts for this fact about English word order without resorting to stipulation. Furthermore, we have seen that functional heads such as determiners and inflections have a very limited range of possible complements: determiners always have NP complements and inflections always have v/VP complements. The element which follows a complementiser is always an IP and so this fits the pattern very well. Finally, note that different complementisers introduce different IPs. If and that both introduce finite IPs, while for introduces a non-finite IP. As heads select for their complements, this is again an indication that the complementisers are heads selecting for different types of IP complements. This is very similar to plural determiners selecting for plural NP complements and singular determiners selecting for singular ones.

In what follows, we will be assuming the structure in (9) as essentially correct, though we will see that some extension will be needed for elements that appear between the complementiser and the IP. We will start by discussing facts that do not concern these elements however, and so for the time being (9) provides us with an adequate model of the top part of the English clause.

245

Chapter 7 - Complementiser Phrases

2 The Clause as CP

Not all clauses are introduced by a complementiser. For example, subordinate declarative finite clauses may or may not be introduced by that and main clauses never have complementisers:

(11)a she said [(that) we should make the sandwiches] b (*that) we should make the sandwiches

What is the status of the clause when there is no complementiser? One possibility is that when there is no complementiser there is no CP and hence a clause without a complementiser has the status of an IP. For embedded clauses this is a problematic conclusion as it means that the verbs which select for such clauses must be able to take IP or CP complements. In other words, they subcategorise for a complement with the features [+F, –N]. But if this is so, we would predict that there should be verbs that select for only CP complements, i.e. complements with [+F, –N, –V] features, and those that select for only IP complements, with [+F, –N, +V] features.

But while there are many verbs which take clausal complements both with or without a complementiser, it is doubtful whether the other predicted verb types exist. It seems that we have to accept a generalisation that if a verb selects for a declarative finite IP complement, it also selects for a declarative finite CP complement. It is not easy to think how we can explain this generalisation when stated in this way. There is another possible view, however. This sees all these complements as being CPs, but sometimes the complementiser is filled with an overt that and sometimes it is filled by an unpronounced complementiser:

(12)a she said [CP that she wanted ham and pickle] b she said [CP she wanted ham and pickle]

The generalisation is now that all verbs which select for a finite declarative complement select for a CP. This is fairly easy to capture in terms of the notion of canonical structural realisation principles. The idea behind this is quite simple. Basically, certain arguments are canonically realised by certain categories. For example, themes are typically realised as DPs and locations as PPs. This is their ‘canonical realisation’. It may be that a certain degree of non-canonical realisation of arguments is possible, for example the nominal home can realise a goal argument usually realised by a PP:

(13)he went [PP to London]/[DP home]

All we need to say is that something with a propositional meaning is canonically realised as a CP and then it follows that if a verb takes a propositional complement, this will be realised as a CP. It follows from this that all finite declarative complement clauses will be CPs and hence that we must assume that sometimes the complementiser can be abstract, as in (12b). Non-finite complement clauses differ from this pattern quite substantially. Certain verbs take non-finite complements with an obligatory complementiser:

(14)a we were hoping [for the good weather to arrive soon] b *we were hoping [the good weather to arrive soon]

246

The Clause as CP

Verbs such as wish, prey, plead, demand, indicate, signal, etc. all seem to behave in this way. Obviously, for these verbs there is no question that they take CP complements.

Others take non-finite clause complements that never have a complementiser:

(15)a I tried [ - to spread the butter]

b *I tried [for - to spread the butter]

Verbs such as attempt, have (= be obliged), promise, wish, prey, plead, demand, etc. all behave like this.

Note that some of these verbs are in the other category as well. However there is a difference, verbs in the try category take non-finite complements with missing subjects and those in the hope category take non-finite complements with overt subjects. Thus there seems to be a correlation between when the complement clause has an overt subject and when it has an overt complementiser. We will go into this in more detail in the next chapter, but it can be argued that clauses with covert subjects must be CPs with a covert complementiser position:

(16)I attempted [CP - to cut the tomatoes]

One class of verb takes a non-finite clause complement that has an overt subject:

(17)he believes [Troy to be trouble]

In the next chapter we will argue that these are exceptional verbs and do not behave like the others in that they take IP non-finite complements. Exceptions aside however, the conclusion is that the majority of non-finite complement clauses seem to be, like the finite ones, CPs. Hence a general conclusion seems to be that complement clauses are always CPs.

This leaves main clauses. As pointed out in (11b), these never have overt complementisers. However, given that covert complementisers seem to be a possibility it is reasonable to ask whether main clauses are CPs which have an obligatory covert complementiser, or whether they are just IPs with no space for a complementiser. The issue is complicated unfortunately. On the one hand, there are some main clauses that have to be argued to be CPs, as we shall see a little later. Thus, on general grounds it seems reasonable to assume that all clauses are CPs. Moreover, if the role of the complementiser is to indicate the force of a sentence, and main clauses without complementisers have a force interpretation, then it might be argued that there must be a complementiser to provide this aspect of clausal semantics. On the other hand, most linguists accept that ‘exceptional clauses’ lack complementisers and these also have a force interpretation and so it seems that there is a way for this to be introduced in the absence of a complementiser, which undermines the argument that main clauses must have complementisers because they have a force interpretation.

If we assume that main clauses are CPs we need an explanation as to why their complementisers are obligatorily covert. But if we assume that main clauses are merely IPs, we must account for why the CP is obligatorily banned. All in all then, it is hard to decide on the issue. In this book, we will take the fairly standard view that all clauses are CP (except for the exceptions) and hence we assume that main clauses have obligatorily covert complementisers by a general principle.

247

Chapter 7 - Complementiser Phrases

3 Interrogative CPs

3.1Basic positions within the CP

So far we have looked mainly at declarative CPs. In this section we will turn our attention to interrogatives. There are a number of different types of interrogatives and most of them seem to make use of the CP in one way or another. The two most obvious ones are wh-questions and yes–no questions.

A wh-question is formed with the use of an interrogative pronoun, usually called a wh-element as in English they mostly begin with the letters ‘wh’, as in who, why, which, what, where, etc. How is an exception to this spelling convention, though from a linguistic point of view, the spelling is uninteresting and it is syntactic and semantic behaviour that are more important. From these points of view, how is just like the other ‘wh’-elements. In a wh-interrogative, the sentence begins with the wh-element:

(18)a who said that?

b what did you say?

c why do you say that?

Apart from (18a) where the wh-element replaces the subject, all other instances of wh- questions involve the wh-element being in front of the subject. Given that the subject’s surface position is the specifier of the IP, this indicates that the wh-element sits in a position outside the IP. Above the IP is the CP, so we may assume that the wh-element is situated in the CP. As wh-elements are phrases, probably DPs, though some of them have underlying distributions similar to PPs and therefore are either PPs or perhaps DPs non-canonically realising the prepositional role, they must occupy a phrasal position in the CP and the most obvious choice would be the specifier of the CP:

(19)CP

DP C©

he doing

In yes–no questions, there is no interrogative pronoun and the question is interpreted as asking about the truth of the proposition that it expresses. Hence they may be answered by “yes” (that’s true) or “no” (that’s not true):

(20)a will they ever stop singing? b have you seen my glasses? c did you see that?

Note that these questions are formed by placing an auxiliary verb in front of the subject, a phenomenon traditionally termed subject–auxiliary inversion. Again, given that the auxiliary is to the left of the subject, it seems to be outside of the IP and presumably somewhere in the CP. This time, however, the position needed is a word position and the obvious choice is the complementiser position itself:

248

Interrogative CPs

(21)CP C©

C IP

was it my fault

It makes perfect sense that wh-questions and yes–no questions involve the complementiser system as complementisers contribute the semantics of force to the sentence, but there is independent evidence for the validity of these analyses. To start, that inverted auxiliaries occupy the C position is supported by the observation that inverted auxiliaries and complementisers are in complementary distribution. The best evidence for this involves not interrogative clauses, though it is also true that we never get inverted auxiliaries and complementisers together in an interrogative, but conditional clauses. In English there are two types of conditional clause, one formed with an if complementiser and one formed with an inverted auxiliary:

(22)a [if he’s a government minister] then I’m the Queen of Sheba b [had I known about your allergy] I wouldn’t have sent flowers

That if conditionals do not involve auxiliary inversion demonstrates that the complementiser and the inverted auxiliary are in complementary distribution and therefore occupy the same position:

(23)a *[if had I known royalty was visiting] I would have combed my hair

b *[had if I been told the deadline] I would have typed the report yesterday

Moreover, with many wh-questions we also get auxiliary inversion:

(24)a what will you do?

b when should we meet? c who did you talk to?

As we can see the wh-element precedes the inverted auxiliary, supporting the assumption that the wh-element is in the specifier of CP and the inverted auxiliary is in the head position:

(25)CP

DP C©

Having established the structure of these interrogatives, let us spend some time discussing the processes involved in their formation.

249

Chapter 7 - Complementiser Phrases

3.2Wh-movement

Wh-elements come in a variety of forms and functions. All of them contain some element that has an interrogative interpretation, which may be a pronoun, a determiner or a degree adverb:

(26)a [DP what] do you think?

b [DP which present] did she get you? c [PP to whom] am I talking?

d [AP how fast] did they run?

We might assume that all of these phrases share a common ‘interrogative feature’ which determines their interpretation as question elements. Typically this feature is represented as [+wh] and is contrasted with [–wh] which has a declarative interpretation.

The main issue to be discussed is why wh-elements move to the specifier of the CP. We have seen that movement of DPs to the specifier of IP is motivated by the need to circumvent the actions of the Case filter. But this cannot be the motivation for wh-movements. This can be argued for from a number of perspectives. First note that wh-movement does not always involve DPs, unlike movements to spec IP. As the Case filter concerns only the distribution of DPs, it cannot be the motivation for all wh- movements and there must be another reason for why non-DP wh-elements, at least, undergo the process. Moreover, the Case filter cannot be the motivation for DP wh- elements moving to spec CP. This is because DP wh-elements move to spec CP from Case positions:

(27)a what1 do you know t1?

b who1 did you speak to t1?

c who1 do you think [t1 is the thief]?

The wh-element in (27a) moves from an object position, from the object of a preposition in (27b) and from a finite clause subject position in (27c). All these are Case positions, so the Case filter cannot explain why the movement took place.

There seem to be two interrelated motivations for wh-movement, one concerning the interpretation of the clause, and one concerning the interpretation of the wh- element itself. Let’s start with the clause. Obviously clauses are either interpreted as questions or not. Moreover, in embedded contexts the distribution of a clause will depend on its force as some verbs require an interrogative complement while others require a declarative one:

(28)a Andrew asked [if/*that the pears were ripe] b they didn’t think [that/*if the pears were ripe]

In the examples in (28) the force of the embedded clauses is obviously determined by the complementiser, as would be expected. However, in the following, it seems that the force of the clause is determined by the wh-element in specifier positions:

(29)Wendy wondered [when the pears would be ripe]

This is a little puzzling as phrases do not get their properties form their specifiers, but from their heads.

250