2011-06-03

Island (syntax)

Wh-movement

Wh-movement (or wh-fronting or wh-extraction or A'-movement) is a syntactic phenomenon found in many languages around the world, in which interrogative words (sometimes called wh-words) or phrases show a special word order. Unlike ordinary phrases, such wh-words appear at the beginning of an interrogative clause. The term wh-movement is because most English interrogative words start with wh-, for example, what, where, why, etc. The term wh-movement tends to be applied to similar word order permutations in languages other than English as well, even when the interrogative words of a given language do not start with wh- (though some authors use the term A'-movement to avoid confusion).

Because of variation in analyses and terminology, wh-movement constructions are sometimes referred to as long-distance dependencies or unbounded dependencies. These names are most commonly used by linguists who work with non-transformational approaches like lexical functional grammar and head-driven phrase structure grammar.

According to Joseph Greenberg's linguistic universal No.12, "If a language has dominant order VSO in declarative sentences, it always puts interrogative words or phrases first in interrogative word questions; if it has dominant order SOV in declarative sentences, there is never such an invariant rule." Many SVO languages have overt wh-movement too, such as English, but some do not, such as Chinese. Languages with no overt wh-movement are referred to as wh-in-situ languages. In 1981, Huang argued that there must in fact be wh-movement in Chinese, though this movement is covert and unpronounced. His arguments were widely influential for theories of generative grammar.

Wh-movement in English

Wh-movement in English main clauses

English is one language that features wh-movement. For example, a declarative sentence in English featuring normal word order would be:

He buys bread.

The direct object, "bread", of the verb, "buy", normally follows the verb; however, when the direct object is replaced with a wh-word in order to form a question, the wh-word generally appears at the beginning of the sentence:

What does he buy?

In standard English main clause nonsubject questions, an auxiliary verb needs to follow a wh-word. If there is no auxiliary verb, a form of "do" must be used. The auxiliary verb (or a form of "do") occurs after the wh-word and before the subject:

He should buy bread.
What should he buy?

As mentioned above, wh-movement can extend over several clauses. Note the following example:

What does she think that I said that he buys?

The subject, which is normally at the beginning, can also be questioned. The question word also goes at the front, replacing the subject. There does not seem to be any movement here, but for uniformity of explanation it may be that there is also movement and a trace here too:

He buys bread.
Who buys bread?

There are three circumstances in which wh-movement does not occur in English: echo questions (to confirm what you thought you heard), quiz questions, and multiple questions, when there is already one wh-word at the front:

You bought what!?
George Orwell was born in which country?
Who bought what?

Wh-movement in English subordinate clauses

Wh-movement is also seen in subordinate clauses in English. Sentences of the kind below are sometimes called embedded (or indirect) questions.

I wonder what he bought.

rather than

*I wonder he bought what.

However, most varieties of English do not show the auxiliary do in such cases:

*I wonder what did he buy.

In most varieties of English, other auxiliaries remain in their normal position after the subject of the sentence:

I wonder what he should buy.

rather than

*I wonder what should he buy.

Examples like this show us that the wh-phrase does not necessarily occur at the front of the sentence, but sometimes occurs at the beginning of a subordinate clause.

Belfast English has been cited as an example of an English dialect where *I wonder what did he buy and *I wonder what should he buy are allowed. However, most North American and British English disallows these constructions.

Wh-movement in English relative clauses

Wh-words at the front of the clause are also seen in relative clauses in English:

I know the man [ who bought the cheese].
I know the woman [ whom John saw].
I know the place [ where John put the cheese].

In some of these examples the wh-word may be omitted:

I know the woman [John saw].
I know the place [John put the cheese].

The word that may also be used instead of a wh-word:

I know the man [ that bought the cheese].
I know the woman [ that John saw].
I know the place [ that John put the cheese].

Theoretical approaches to wh-movement

The name wh-movement comes from analyses in Generative Grammar where a wh-word begins at some other place in a sentence and moves to the front. Although wh-movement is a common name for this phenomenon, wh-movement is probably not the best name because there are a number of other elements in a sentence that show the special word order found in questions.

The details of wh-movement are very complex, particularly when English is compared to other languages with wh-movement. All modern theories of syntax have some part of the theory which deals with the correct formulation of the rules for wh-movement.

Because of debate about the best analysis of wh-movement, there is also some variation in the terminology for talking about the parts of a sentence that contains wh-movement. However, any theory will need to talk about

a.) the word or phrase which shows a special order, usually at the beginning of a clause. This word or phrase is sometimes called the filler or the moved element. b.) the position where the word or phrase would normally have appeared. This position is often called the gap, and in some theories of syntax there is a silent element called the trace , which occupies this position. The trace is sometimes indicated in a syntactic diagram of the sentence as t (for trace) or e (for empty). c.) the part of the sentence which is between the "filler" and the "gap". This part of the sentence is sometimes called the "dependency path".

In early transformational approaches to syntax, the analysis of wh-movement involved two syntactic levels -- deep structure and surface structure. The moved element occupies the position of the "gap" at deep structure. It undergoes a rule which moves it to a special position at the beginning of a clause. The structure of the sentence after the movement rule is called surface structure. In more modern approaches to syntax such as Minimalism, there is no special deep structure level, but words and phrases still undergo movement to arrive at their final position.

Theories that do not posit separate syntactic levels like deep structure and surface structure, such as lexical-functional grammar and head-driven phrase structure grammar, do not use movement rules as such in their analyses. Instead they speak of the dependency relationship between the "filler" and the "gap" in a sentence to account for the grammar of these sentences through restrictions on the feature structures on the sentences.

Wh-movement in other languages

Wh-movement is also found in many other languages around the world. Most European languages also place wh-words at the beginning of a clause, as in the following Spanish example:

'What did Juan buy?'

In this example qué is the object of the verb compró, but it appears at the beginning of the interrogative clause. In contrast, a normal object will follow the verb:

'Juan bought meat.'

Wh-movement is also found in many other languages around the world. In some languages, such as French, it is optional in certain matrix clauses.

Pied-piping

Pied-piping (first identified by John R. Ross) describes the situation where a phrase larger than a single wh-word occurs in the fronted position. In the case where the wh-word is a determiner such as which or whose, pied-piping refers to the wh-determiner's appearance sentence-initially along with its complement. For instance, in the following example, the entire phrase "which car" is moved:

Which car does he like t?

In the transformational analysis, the wh-word which moves to the beginning of the sentence, taking car, its complement, with it, much as the Pied Piper of Hamelin attracted rats and children to follow him, hence the term pied-piping.

In the case of determiners, pied-piping is obligatory. For instance, the following sentence would be ungrammatical:

*Which does he like t car?

However, there are cases where pied-piping can be optional. In English, this is often the case when a wh-word or phrase is the object of a preposition. For instance, the following two examples are both grammatical:

To whom did she reveal her secret t?
Whom did she reveal her secret to t?

The second example is a case of preposition stranding, which is possible in English, but not allowed in Latin or other Romance languages. For languages that use postpositions rather than prepositions, stranding is not allowed either.

Prescriptive grammarians often claim that preposition stranding should be avoided in English as well; however, in certain contexts obligatory pied-piping of prepositions in English may make a sentence feel artificial or stilted (e.g. "To whom are you talking?" rather than the more conventional "Who are you talking to?").

Some languages show a special word order in pied-piped phrases. This phenomenon is known as pied-piping with inversion or secondary wh-movement.

Extraction islands

In many cases, a wh-word can occur at the front of a sentence, regardless of how far away its canonical location is. For example:

Whom does Mary like t?
Whom does Bob know that Mary likes t?
Whom does Carl believe that Bob knows that Mary likes t?

In more technical terms, we can say that the dependency relation between the gap and its filler is unbounded in the sense that there is no upper bound on how deeply embedded within the given sentence the gap may appear. Consider the following three examples:

This is the book [which [John recommended ____ t ]].
This is the book [which [I think [John recommended ___ t ]]].
This is the book [which [I think [you said [John recommended ___ t ]]]].

In these examples, the NP the book functions as a filler and a gap in the embedded clause. As shown above, there is no grammatical limit on how many layers of embedding there should be to make a grammatical sentence. (If we don't attempt a much longer sentence with the embedding structure, that's probably because of processing constraints or psychological reasons, not because of the grammatical restriction of the long-distance dependencies.)

However there are cases in which this is not possible. Certain kinds of phrases do not seem to allow a gap. These phrases from which a wh-word cannot be extracted are referred to as extraction islands or simply islands. In addition to the islands listed below, regular that-clauses serving as complements to verbs may show island-like behavior if the matrix verb is not a so-called bridge verb (a verb permitted movement across it, hence the name, coined in Erteschik-Shir 1973). Non-bridge verbs include manner-of-speaking verbs, such as whisper or shout. Compare (the star indicates that the sentence is not acceptable):

How did you think he did __?
*How did you whisper he did __?

Adjunct islands

An adjunct island is a type of island formed from an adjunct clause. Wh-movement is not possible out of an adjunct clause. Adjunct clauses include clauses introduced by because, if, and when, as well as relative clauses. Some examples include:

You went home because you needed to do what?
*What did you go home because you needed to do t?
Alex likes the woman who wears what?
*What does Alex like the woman who wears t?

Wh-islands

A wh-island is an island that is created by an embedded sentence which is introduced by a wh-word. For instance, the clause "where Eric went to buy the gift" in the following example, is a wh-island:

John wonders where Eric went to buy the gift.

Wh-islands are weaker than adjunct islands since extraction is often awkward but not necessarily considered ungrammatical by all speakers.

Grammatical: John wonders where Eric went to buy what?
Questionable: ?What does John wonder where Eric went to buy t?

Complementizer-trace effect (or that-trace effect)

It is typically easier to extract objects rather than subjects from a clause, especially when an overt complementizer such as "that" or "for" is used. Take the following examples:

Who do you believe t saw Tom?
*Who do you believe that t saw Tom?
Whom do you believe Jim saw t?
Whom do you believe that Jim saw t?

Subject islands

Wh-movement does not appear to be possible out of phrases that appear in the subject position. This is particularly true for subject clauses. For instance, here is a sentence where the clause appears in a non-subject position (the predicative complement):

It is likely that John went home.

Here is the same sentence where the clause appears in the subject position:

That John went home is likely.

Notice that wh-movement can occur only in the clause that appears in the predicate position:

Grammatical: Where is it likely that John went t?
Ungrammatical: *Where is that John went t likely?

Left Branch islands

The left branch of an NP cannot be extracted.

Two subcases fall under this ban: attributive adjectival phrases (and their subparts) and possessors:

*How expensive did he buy [a _ boat]?
*Whose did she shake [_ hand]?

In these situations, pied-piping is obligatory.

Left branch islands are among the most cross-linguistically variable; while they exist in English, they are absent in many other languages (most famously, the Slavic languages).

The Coordinate Structure Constraint

Extraction of or out of a conjunct is ungrammatical:

*What did Sam eat [beans and _]? (The conjunct condition)
*What did Sam buy [a house next to _] and [a sailboat from Martha]?

Extraction out of a conjunct is possible only if this extraction affects all conjuncts equally (that is, if it occurs in an across-the-board manner):

Whom did Sam buy [a house next to _] and [a sailboat from _]?

The Complex Noun Phrase Constraint

Extraction out of complex noun phrase such as [a man who went on holiday] is ungrammatical:

*Where do you know [a man who went _]?

See also

Notes

References

  • Cheng, Lisa. 1997. On the Typology of Wh-Questions. Garland Publishing, Inc., New York and London.
  • Cheng, Lisa Lai-shen and Norbert Corver, eds., Wh-Movement: Moving On, The MIT Press, 2007
  • Chomsky, Noam, "On Wh-Movement", in Culicover, P. W., Wasow, Thomas, and Akmajian, Adrian (eds), Formal Syntax, New York, 1977
  • Erteschik-Shir, Nomi. 1973. On the Nature of Island Constraints. Ph. D. thesis, Massachusetts Institute of Technology, Cambridge, MA.
  • Greenberg, Joseph H. "Some universals of grammar with particular reference to the order of meaningful elements", in Greenberg, Joseph H. (ed.) Universals of Language, London: MIT Press, 73-113, 1963
  • Grosu, Alexander. 1973. On the Left Branch Condition. Linguistic Inquiry.
  • Ross, John R. Constraints on variables in syntax. Doctoral dissertation, Massachusetts Institute of Technology, 1967

External links






Retrieved from : http://en.wikipedia.org/wiki/Wh-movement#Extraction_islands
Posted by at