Wh-movement and first language acquisition

1)  Introduction  3

2)  Wh-movement  4

2.1.  Characteristics of Wh-Movement  4

2.2.  Wh-Movement as Copying and Deletion  5

2.3.  Wh-Movement and the EPP Feature  7

2.4.  Restrictions on Wh-Movement  8

3)  Wh-Movement and First Language Acquisition  10

3.1.  The Acquisition of Wh-Movement  10

3.2.  Wh-Movement and UG  12

3.3.  Acquisition of Long Distance Movement and Direct-Object Wh-Questions  13

3.4.  Summary  15

4)  Conclusion  16

5)  Bibliography  17

6)  List of Abbreviations  18

7)  Appendix  19

Introduction 1)

In the first part of the paper I want to present the characteristics of wh-movement and want to show that there are good reasons to assume that the wh-expressions originate as complements at the end of the sentence and are moved to spec-CP in order to satisfy the [WH] and [EPP] feature of C. Furthermore, I want to present different kind of evidence - empirical, semantic, theoretical and evidence from language acquisition - that wh-movement is a syntactic phenomenon involving copying and deletion. These aspects can also be studied in connection with first language acquisition and this will be my topic in the second part of the paper. Here I want to ask the question how wh-movement is acquired by English speaking children and if one can find signs that Universal Grammar is involved in the process. Furthermore, I want to show how one can explain that the mechanisms triggering wh- movementdo not function in an early stage of language acquisition.

I think it is justified to combine syntactic theory with data from first language acquisition since this can be a very productive for both fields of study. Finding data from language acquisition to support one’s syntactic theory always gives a solid basis for the argumentation. On the other hand, it is easier to know what to look for in language acquisition when one has the current syntactic research in mind. I want to begin the first part of the paper with a look at the most important characteristics of wh-movement.

3

2) Wh-movement

2.1. Characteristics of Wh-Movement

Wh-movement is a syntactic phenomenon whereby interrogative wh-words appear at the beginning of interrogative sentences. The term wh-movement is due to the fact that most English interrogative words start with wh-, for example what, where, why, when etc. However, how is also treated as an wh-word in sentences like How are you? The term wh- movementis applied universally, even when the interrogative words of a given language (such as French) do not start with wh-. Since the landing site of wh-movement is an A-bar position, this type of movement is also referred to as A´-movement (Culicover 1997:179).

The following examples show that it is justified to assume that wh-expressions originally occupy a complement position at the end of the sentence; in the following two examples it is the complement position of the verb (Radford 2004:153): (1) What languages can you speak? You can speak what languages? (2) Which one would you like? You would like which one?

However, the wh-expression can also have other functions within the sentence (Borsley 1999:188):

(3) Who did Bob see? – object of a verb (4) Who did he give the book to? – object of a preposition (5) Who did Hobbs consider to be a fool? – subject of an infinitive (6) Who do you think saw Hobbs? – subject of a finite VP In (1) and (2) we have seen that the wh-expression can sometimes remain at the end of the sentence, but typically, the fronting in questions is obligatory:

(7) Whom does Susan love? * Susan loves whom?

However, sentence (7*) may occur in natural language as an echo-question: (8) Speaker A: I just heard that Susan loves Peter.

(9) Speaker B: What? Susan loves whom?

4

After having thought about the nature of wh-words and their function within the sentence, I now want to take a closer look at the movement operation as such. One interesting observation is that very often the respective wh-word is not moved on its own, but that it carries material with it. In (14) for instance, the entire QP which girl is moved: (14) Which girl does he like? (see attachment 7.1.) It looks as if the noun girl was pied-piped to move with the quantifier which into Spec-CP position - much like the Pied Piper of Hamelin attracted rats and children to follow him. This is why the phenomenon is referred to as pied-piping. In terms of quantifiers, pied-piping is obligatory - a sentence like *Which does he like girl? would be ungrammatical. Radford argues that this has to do with the nature of specifier positions which have to be occupied by maximal projections (2004:164). Unlike head movement, wh-movement always involves movement of maximal projections and so the whole QP which girl has to be moved in order to form a grammatical sentence. To the question why no other maximal projections can be moved (like the for example the VP resulting in * Like which girl does he?) Chomsky formulated the UG Principle of Convergence which says that when an item moves, it carries along with it `just enough material for convergence` (informally cited in Radford 2004:165). However, there are cases where pied-piping can be optional. In English, this may happen when the wh-word is the object of a preposition:

(10) To whom did she reveal her secret?

(11) Whom did she reveal her secret to?

Sentence (11) is a case of preposition stranding which means that in informal styles of English the preposition can be left orphaned at the end of the sentence (Radford 2004:167).

2.2. Wh-Movement as Copying and Deletion

Much evidence can be found that wh-movement is an operation involving copying and deletion with usually a null-spell out of the copy in the PF-component. First of all, this idea is supported by the phenomenon of wanna-contraction. In English, a sentence like I want to leave can be shortened into I wanna leave. However, this is not possible when wh-movement has taken place (t standing for trace):

(12) Who do you want t to kiss the puppy?

(13) * Who do you wanna kiss the puppy?

5

The theory of traces offers a nice explanation for the ungrammaticality of (13), namely that the copy of the moved wh-word stands in the way between want and to, so that both words cannot merge with each other (Carnie 2002:287).

The theory of copying and deletion can also be defended on empirical grounds, as we see by looking at the following examples:

(14) What proof that he was implicated have you found?

(15) What proof have you found that he was implicated?

Here we have the possibility of a split or discontinuous spell out since we see that a CP belonging to the moved constituent can be spelled out in one position and that the remainder of the constituent can be spelled out in another (Radford 2004:157). Without the assumption of movement as a copying and deletion operation, the phenomenon of split spell out would be difficult to explain.

Another strong argument can be given from a semantic point of view. Let us consider the meaning of the following sentence:

(16) Joe wonders which picture of himself Jim has bought.

In this sentence the reflexive anaphor himself can either refer to Joe or to Jim. However, from a syntactic point of view we can not account for the dual interpretation of himself since a reflexive has to be c-commanded by a local antecedent and Jim does obviously not c-command himself. Chomsky argues that in the semantic component, the sentence looks like this: [CP[TP Joe wonders [CP which picture of himself [TP Jim bought which picture of himself]]] Although the italicised copy of which picture of himself gets deleted in the PF component, the possibility of himself referring to Jim can be attributed to the fact that the italicised himself is c-commanded by Jim in the LF representation - which is where the binding conditions actually apply (Radford 2004:158).

Finally, the last interesting source of evidence comes from the study of language acquisition (Radford 2004:128). In the following examples, Sam, 33 month old, successfully produces auxiliary copying structures, but has obviously not yet mastered the deletion of the original auxiliary in T (although no wh-movement is involved, the principle is the same):

6