Andrew Radford - Linguistics An Introduction [Second Edition]

But now contrast the bracketed clause in (388) with the clause in (390):

(390)

Das Buch hat der Adrian gelesen

The book has the Adrian read

‘The book, Adrian has read’

322

senten ces

There are three important differences between the two. Firstly, the clause in (388) contains the complementiser dass ‘that’ (because it is a complement clause, here serving as the complement of the verb weiss ‘know’), but that in (390) doesn’t (because it isn’t a complement clause). Secondly, the auxiliary hat ‘has’ is positioned at the end of the clause in (388), but in front of the subject der Adrian in (390). And thirdly, the complement das Buch ‘the book’ is positioned immediately in front of the verb gelesen ‘read’ in (388), but in front of the auxiliary hat ‘has’ in (390). How can we account for the change in word order between (388) and (390)?

Given our framework, the obvious analysis is to say that those constituents

which have changed their position in (390) relative to the position they occupy in

(388) have undergone movement. Thus, the auxiliary hat ‘has’ originates at the end of the clause (as in 389) but is then moved into the complementiser position at the beginning of the clause – precisely as happens in the case of auxiliary

inversion in English; and the DP das Buch ‘the book’ is preposed from its original complement position immediately in front of the verb gelesen ‘read’ and moved into the specifier position within CP (in much the same way that topic phrases are in English). As a result, (390) will be derived as in (391):

(391)

CP

DP

C'

Das Buch

C

TP

hat

DP

T'

der Adrian

VP T

hat

DP V

das Buch

gelesen

Here, we see that the auxiliary hat ‘has’ originates in T and moves to C, and the DP

das Buch ‘the book’ originates in complement position within VP and moves into specifier position within CP.

Now consider the following sentence:

(392)

Der Adrian hat das Buch gelesen

The Adrian has the book read

‘Adrian has read the book’

Since the auxiliary hat ‘has’ doesn’t occupy its normal position at the end of the clause here, it seems once again to have moved from T to C. And this time, the subject der Adrian is positioned in front of the auxiliary, so seems to have moved from specifier position in TP into specifier position within CP. This means that

(392) has the derivation in (393):

Syntactic variation

323

(393)

CP

DP

C'

Der Adrian

C

TP

hat

DP

T'

der Adrian

VP

T

hat

DP

V

das Buch

gelesen

This structure shows that the auxiliary hat ‘has’ has moved from T to C, and the subject der Adrian has moved from spec-TP to spec-CP.

Next consider (394):

(394)

Welches Buch las der Adrian?

which

book read the Adrian

‘Which book did Adrian read?’

What’s going on here? It seems clear that the operator phrase welches Buch ‘which book’ has moved into the specifier position within CP (as in English). But how does the verb come to be positioned after it and in front of the subject der Adrian?

The obvious answer is that (much as in Early Modern English), the verb moves out of the head V position in VP, into the head T position in TP, and from there into the head C position in CP, as indicated in (395):

(395)

CP

DP

C'

Welches Buch

C

TP

las

DP

T'

der Adrian

VP

T

las

DP

V

(III)

welches Buch

las

(I)

(II)

Movement (I) in (395) is head movement of the verb las ‘read’ from V to T; movement (II) is again head movement of the verb las ‘read’ from T to C; and movement (III) is wh-operator movement of the DP welches Buch ‘which book’

from complement position within VP into specifier position within CP. Since the verb las can move from V to T and from there to C, it follows that both T and C must

324

senten ces

be strong in finite clauses in German (and hence have to be filled at some stage of the derivation). Note that in consequence of the head movement constraint (381) (or the Economy Principle requiring ‘short’ moves), the verb las cannot move directly from V to C, but rather must move first to T, and then from T to C.

An interesting property which the German CPs in (391), (393) and (395) share is that in each case the specifier position within CP must be filled – though this is not true of (389) where dass ‘that’ appears to have no specifier. This means that where the head C of CP is filled by a preposed verb or auxiliary (as in 391, 393 and 395), CP must have a specifier.

The assumption that clauses in which C is occupied by a preposed verb or

auxiliary require a specifier has interesting implications for how we analyse yes–no questions such as (396):

(396)

Las der Adrian das Buch?

Read the Adrian the book

‘Did Adrian read the book?’

Here, the overall clause (like all clauses in German) is a CP, and the head C

position of CP is filled by the preposed verb las ‘read’. If we posit that CPs headed by a preposed verb or auxiliary require a specifier, how can we account for the fact that there appears to be no CP-specifier preceding the verb las in (396)? Recall that in section 20 we suggested that yes–no questions contain an abstract question operator ? which occupies the specifier position within CP, and which is required if a sentence is to be interpreted as a question. This being so, (396) will have the derivation in (397):

(397)

CP

ADV

C'

?

C

TP

las

DP

T'

der Adrian

VP

T

las

DP

V

das Buch

las

(I)

(II)

The verb las originates in the head V position of VP and then moves from there firstly into the head T position of TP, and then into the head C position of CP (since C is strong in all finite main clauses in German and so always has to be filled). The requirement for the specifier position within CP to be filled where C contains a preposed verb or auxiliary is satisfied by the null question operator ? which occupies spec-CP, and which serves to mark the clause as a yes–no question.

Syntactic variation

325

Our discussion of structural variation in this section has important implications for the development of a theory of grammar. In previous sections, we have assumed that principles of Universal Grammar (UG) determine that certain aspects of

syntactic structure are invariant across languages (e.g. every phrase or clause is a projection of a head; clauses are universally CPs; questions universally contain an interrogative operator in spec-CP; subjects are universally positioned in spec-TP; categories can universally be overt or covert, etc.). But in this section, we have seen that there is a certain amount of structural variation across languages and language varieties, and that this can be characterised in terms of a set of binary parameters.

This leads us towards the Principles and Parameters Theory (PPT) developed by Noam Chomsky and many others over the past three decades, in which those

aspects of syntactic structure which are invariant across languages are attributable to principles of UG, while those aspects of structure which vary from one language to another are described in terms of a set of (binary) parameters (exercises 2 and 3).

Exercises

1.

Discuss the structure of the following sentences in African American