A Comprehensive Analysis of Wh-Movement in Interrogative Sentences in English

CONTENTS

1 Introduction

2 Brief Overview of Government and Binding Theory

3 What is Wh-Movement?

4 Typology of Wh-Movement and Movement at LF

5 Wh-Movement in Interrogative Sentences in English
5.1. Subject/Auxiliary-Inversion
5.2. Constraints on Wh-Movement at S-structure

6 PROBLEM ANALYSIS: Why are Certain Constructions that Violate the Wh-Island Constraint Acceptable?

7 Sketch of Wh-Movement in the Minimalist Program

8 Conclusion

References

1 Introduction

This essay is going to provide an in-depth account for the phenomenon of wh -movement in English along the lines of Chomsky’s Government and Binding theory (GB), a transformational approach in which wh -movement is described as constituent movement. The analysis will mainly focus on wh -movement in interrogative sentences. First, the theory of Government and Binding is being briefly presented and the term “wh -movement” is going to be defined. Subsequently, the different types of wh -movement occurring in languages are being presented together with wh -movement at LF. This introduction of the topic is going to be followed by an overview of prominent constraints that account for illicit wh -movement. The focus of analysis is going to be on Ross’s (1967) island constraints and Chomsky’s (1977) Subjacency. The description of wh -movement in interrogative sentences and its constraints will provide the theoretical foundation for the subsequent problem analysis: Specifically, this analysis addresses the question why there are some sentences that involve the violation of wh -islands but are acceptable for English native speakers. At the end of this essay, it is going to be briefly shown how wh -movement is being analyzed in the framework of the Minimalist Program, the successor of GB and current most prominent theory of Universal Grammar.

This essay was written during my studies abroad at the University of Queensland in Brisbane, Australia. It was an assignment for the postgraduate course “Generative Syntax” which dealed with Government and Binding Theory. The analysis contained in this essay goes far beyond the material covered in the classes of the course. This essay provides a much deeper and advanced understanding of wh-movement in the framework of GB.

2 Brief Overview of Government and Binding Theory

Government and Binding theory is a transformational approach of the theory of Universal Grammar that was developed by Noam Chomsky in the 1980s. It is often also referred to as Principles and Parameters theory as it stipulates that all languages share universal structural principles together with various parameters on a language-specific basis.^[1] GB was first introduced by Chomsky in his “Lectures on Government and Binding” in 1981.

GB considers Universal Grammar to consist of two components: levels of representation and systems of constraints (cf. Black 1999:2). There are four different levels of how language can be represented: At first there is a mental lexicon which stores the lexical items (= words) with their idiosyncratic features that amongst other things provide information on their syntactic category and subcategorisation (reveals what complements a predicate takes). Predicates, which typically are verbs, encode their semantic relationship to their arguments by assigning θ-roles (theta-roles). At D-structure (D for “deep”) the lexical items are arranged together along the lines of their properties. The structural relations between the items obey the rules of X-bar theory, which regulates how specifiers, adjuncts and complements must be arranged around a head X within a phrase XP (cf. Carnie 2007:188). The derivation at S-structure (S for “surface”), which is the product of the application of transformational rules to D-structure (e.g. rules for wh -movement), is what a speaker actually says (cf. Horrocks 1987:94). The movement rules are subsumed under the general concept “move α”. The transformation to S-structure underlies several constraints, such as Subjacency, Binding Theory, and Control Theory. S-structure is factored into Phonological Form (PF), which is the interface with phonology, and Logical Form (LF), which is the semantic interpretation mechanism where matters such as quantifier scope are determined (cf. Black 1999:2). The different levels of representation can be illustrated with the T-model:

Abbildung in dieser Leseprobe nicht enthalten

3 What is Wh -Movement?

wh -movement (also being referred to as “wh -fronting” and “wh -extraction”) is a syntactic feature by which interrogative words appear at the beginning of an interrogative sentence or clause. In Linguistics, these interrogative words are often called wh -words, as in English most of them start with “wh”. Wh -words can appear in direct as well as indirect questions:

(1) a. What book did you read?

b. I wonder where she is gone.

As we can see, wh -words can either constitute a phrase, as in (1b), or be part of a phrase, as in (1a) where “what” is part of the NP “what book”. A wh -phrase containing a wh -word can be an NP, PP or AdvP.

In GB, wh -movement is considered as constituent movement, where the wh -phrase originates in D-structure, moves up to a higher position in the sentence and leaves behind a silent trace^[2]. The D-structure for (1a) for example would be (2):

(2) [CP You read [NP what book]]

Applying wh -movement, in S-structure we get (1a). The landing site of wh -movement, i.e. the position in which a moved wh -phrase appears in S-structure, is the specifier position of the complementizer phrase, [Spec,CP]. Empirical evidence for this is being provided by Middle English and the variety of English currently spoken in Belfast, which allow the complementizer that to immediately follow wh -phrases (by contrast, modern Standard English does not do so):

(3) a. First the behoueth to knowe why that suche a solitary lyf was ordeyned.

“First, it behoves thee to know why such a solitary life was ordained.”

(example from Middle English; adapted from Santorini 2007)

b. I wonder which dish that they picked.

(example from Belfast English; adapted from Santorini 2007)

According to Chomsky (1973), wh -movement is triggered by the feature [+WH] under COMP (= complementizer). Such a [+WH] COMP attracts a wh -phrase to join it in the specifier position, i.e. [Spec,CP]. In an indirect question, whether COMP in the embedded CP has the feature [+WH] or [-WH], depends on what kind of complement the matrix predicate selects. For example, in (4), the matrix predicate “know” selects a [+WH] complement, whereas in (5) the matrix predicate “think” selects a [-WH] complement, so that here no wh -movement to [Spec,CP] of the embedded CP can take place.

(4) a. I know that John read the Bible.

b. I know what John read.

(5) a. I think that John read the Bible.

b. * I think what John read.

Apart from interrogative sentences, there are two further instances of wh -movement in English: relative clauses and topicalization. Relative clauses are constructed very similar to interrogative sentences, namely by moving a wh -phrase to a [Spec,CP] position. The following two sentence pairs illustrate the similarities, with each (a) being a interrogative sentence and each (b) being a relative clause:

(6) a. [CP [ Where ]i [IP did you [VP go ti on the weekend]]]

b. This is [NP the place [CP [ where ]i [IP you [VP went ti on the weekend]]]]

(7) a. [CP [ Whose french fries ]i [IP did you [VP steal ti ]]]

b. This is [NP the guy [CP [ whose french fries ]i [IP you [VP stole ti ]]]

Both interrogative sentences and relative clauses are introduced by wh -phrases that have moved from a lower position up to [Spec,CP].

Another instance of wh -movement is topicalization which refers to the fronting of a phrase to the beginning of the sentence for the purpose of emphasis, as the following sentences illustrate:

(8) a. The students should be better prepared for next class.

b. [CP [PP For next class ]i [IP the students should [VP be better prepared ti ]]]

(9) a. We like Mary’s parents. We don’t like John’s parents.

b. We like Mary’s parents. [CP [ John’s parents ]i [IP we don’t [VP like ti ]]]

Thus, topicalization can be analyzed as moving a phrase to [Spec,CP]. Although no wh -phrase is involved, Chomsky (1977) argues to subsume topicalization under wh -movement, as here movement takes place to [Spec,CP] that is constrained by Ross’s (1967) island constraints^[3]. For example, a phrase is not allowed to be topicalized out of a complex noun phrase:

(10) * [CP [ John ]i [IP he made [NP the claim [CP that he could beat ti ]]]].

4 Typology of Wh -Movement and Movement at LF

As Richards (2001) notes, across languages three different types of wh -movement are being distinguished:

i.) Overt wh -movement (also called “obligatory”^[4] ) which entails the movement of all wh -phrases at S-structure. A language bearing this kind of wh -movement is Bulgarian, as the following example demonstrates:

Kogo kakvo e pital Ivan?

whom what AUX asked Ivan

„Who did Ivan ask what?“

(adapted from Bošković 1995a:13 according to Richards 2001:1)

ii.) Covert wh -movement (also called “non-existent”^[5] ) which entails that at S-structure no wh -phrases move at all, but stay in their syntactic position of origin. Wh -movement only occurs at LF, i.e. it is phonologically not realized but covert. Japanese is such a language, as the following example demonstrates:

Taroo-ga dare-ni nani-o ageta no?

Taroo-NOM who-DAT what-ACC gave-Q

“Who did Taroo give what?”

(taken from Richards 2001:1)

iii.) Partly overt wh -movement which entails that at S-structure one wh -phrase is moved and the rest is left in situ (the other wh -phrases finally move at LF). English is such a language, as the following example demonstrates:

[...]

^[1] For example, embedding is a universal principle, whereas constituent order is a parameter as it differs from language to language (e.g. SVO in English, SOV in Japanese).

^[2] Carnie (2007:324) provides empirical evidence for the existence of traces by pointing out to wanna -contraction. For example, “I want to kiss” can be changed to “I wanna kiss” which has the same meaning. Carnie argues that such a wanna-contraction is not possible in a sentence like ii):

i) Who do you want ti to kiss the puppy?
ii) * Who do you wanna kiss the puppy?

Carnie argues that sentence ii) is ungrammatical, as the silent trace ti intervenes between “want” and “to” so that adjacency between these two words is blocked, making contraction impossible.

^[3] See section 5 for a comprehensive analysis of Ross’s (1967) island constraints.

^[4] cf. Lasnik/Saito 1992:1

^[5] THE SAME