DS and SS - Carnie (2013)

Download as pdf or txt
Download as pdf or txt
You are on page 1of 3

290 Movement

The following sentence is from Modern Irish, which is a verb-subject-


object (VSO) word order language:
1) Phóg Máire an lucharachán.
Kissed Mary the leprechaun
“Mary kissed the leprechaun.”
In this sentence, the subject (a specifier) intervenes between the verb and
the object; this sentence cannot be generated by X-bar theory. (Try to draw
a tree where the specifier intervenes between the head and the complement
– it’s impossible.)
Now consider the following sentence from French:
2) Je mange souvent des pommes.
I eat often of.the apples
“I often eat apples.”
Souvent “often” intervenes between the verb and the object. If souvent
is an adjunct it is appearing between a head and its complement.
X-bar theory can’t draw the tree for this one either.
Finally think about the relationship between the auxiliary verb have and
its complement main verb in (3). In the last chapter, we claimed that the
participle is a complement to the auxiliary, yet here it is appearing separated
from that complement by the negative word not.
3) He has not eaten yet today.
This is surprising. X-bar theory requires that complements be adjacent to the
head that introduces them, but here we see three cases where that isn’t true.
In sum, X-bar theory under-generates because it does not produce all the
possible grammatical sentences in a language.
Although his concerns were based on very different problems than the
ones in (1–3), Chomsky (1957) observed that a phrase structure grammar
(such as X-bar theory) cannot generate all the sentences of a language. He
proposed that what was needed was a set of rules that change the structure
generated by phrase structure rules. These rules are called transformational
rules. Transformations take the output of X-bar rules (and other
transformations) and change them into different trees.
The model of grammar that we are suggesting here takes the form in
(4). You should read this like a flow chart. The derivation of a sentence
starts at the top, and what comes out at the bottom is your judgment about
the acceptability of that sentence.
Chapter 10: Head-to-Head Movement 291

4) The Computational Component


The Lexicon X-bar rules “the base”

D-structure
(constrained by the theta criterion)

Transformational rules

S-structure
(constrained by EPP)

Grammaticality judgments
X-bar theory and the lexicon conspire together to generate trees.
This conspiracy is called the base. The result of this tree generation is a level
we call D-structure (this used to be called Deep Structure, but for reasons
that need not concern us here, the name has changed to D-structure).
You will never pronounce or hear a D-structure. D-structure is also
sometimes called the underlying form or underlying representation (and is
similar in many ways to the underlying form found in phonology). The
theta criterion filters out ungrammatical sentences at D-structure.
D-structure is then subject to the transformational rules. These
transformational rules can move words around in the sentence. We’ve
actually already seen one of these transformational rules. In Chapter 7,
we looked briefly at T to C movement in subject-aux inversion constructions.
(In this chapter, we’re going to look in more detail at this rule.)
The output of a transformational rule is called the S-structure of a sentence.
The S-structure is filtered by the EPP, which ensures that the sentence
has a subject. What are left are grammatical sentences.
In the version of Chomskyan grammar we are considering here, we
will look at two different kinds of transformations: movement rules and
insertion rules. Movement rules move things around in the sentence.
Insertion rules put something new into the sentence. This chapter is about
one kind of movement rule: the rules that move one head into another, called
head-to-head movement. These transformational rules will allow us to
generate sentences like (1–3) above. X-bar theory by itself cannot produce
these structures.
292 Movement

Generative Power
Before we go any further and look at examples of transformations,
consider the power of this type of rule. A transformation is a rule that can
change the trees built by X-bar theory. If you think about it, you’ll see that
such a device is extremely powerful; in principle it could do anything.
For example, we could allow X-bar theory to generate sentences where
the word “snookums” appears after every word, then have a
transformation that deletes all instances of “snookums” (iv). (v) shows
the D-structure of such a sentence. (vi) would be the S-structure (output)
of the rule.
iv) “snookums”  Ø
v) I snookums built snookums the snookums house snookums.
vi) I built the house.
This is a crazy rule. No language has a rule like this. However, in
principle, there is no reason that rules of this kind couldn’t exist if we
allow transformations. We need to restrict the power of transformational
rules. We do this two ways:
vii) Rules must have a motivation.
viii) You cannot write a rule that will create a violation of an output
constraint.
As we go along we will consider specific ways to constrain
transformational rules so that they don’t over-generate.

1. VERB MOVEMENT (V  T)
1.1 French
Let’s return now to the problems we raised in the introduction to this
chapter. Let’s start with the sentence from French:
5) Je mange souvent des pommes.
I eat often of.the apples
“I often eat apples.”
In this sentence, an adjunct surprisingly appears between the head of VP
and its complement. Compare this sentence to the English sentence in (6):
6) I often eat apples.
In the English sentence, the adjunct does not intervene between the verb
and the complement. The tree for (6) would look like (7).

You might also like