You are on page 1of 6

Syntax Course Material 2020-2021-7

Meeting 7

Complex Verb Forms

Introduction

So far we have been talking about sentences with simple verb forms, such as: She bought a
book, Jane saw Mary, He will eat apples, etc. As we know verb forms are not as simple as these
examples. There are sentences with complex verb forms such as: Jane has eaten the apples, Jane
will be eating the apples, or Jane could have been eating the apples. In this section we will
tackle such complex verb forms.

Let’s examine this sentence:


(1) Jane might have been eating apples.

There are two possible analyses that we can propose to deal with the complex verb form of (1).
The first is to assume that the sentence has a complex auxiliary, so the assumption is called
Complex Aux Hypothesis. The second analysis is to assume that the complex verb form is a
recursive VP, hence it is called a Recursive VP Hypothesis.

Complex Aux Hypothesis

If we assume that the sentence in (1) has a complex Aux, the analysis in a tree diagram will be as
follows.

(2) S

NP Aux VP

N M have been V NP

Jane might have been eating apples

This diagram shows that the Aux is complex, consisting of three elements; a modal might, and
auxiliary components have and been. The other proposal is what we have called the Recursive
VP Hypothesis. This hypothesis assumes that the VP is recursive, that is, it occurs repeatedly,
rather than the Aux is complex. Therefore the tree diagram of sentence (1) in Recursive VP
Hypothesis will look like the following.
(3) S

NP M VP

N V VP
[+Aux]
V VP
[+Aux]
V NP
[-Aux]
N

Jane might have been eating apples

From the two proposals of the possible analysis of complex verb forms like (3) above, which
one can give a more accurate account of the structure of (3)? To answer this question we need
some data. At the first meeting of this class we presented various forms of phrases: noun phrase
(NP), verb phrase (VP), adjective phrase (AP), and prepositional phrase (PP) among others. At
the meeting we also talked about evidence (i.e tests) of phrasehood, they are movement test
(only phrases can be moved), co-ordination test (only phrases can be coordinated), proform test
(only phrases can be replaced by proforms), and omission test (only phrases can be omitted).
Now we will use the omission test to evaluate the validity of the two analyses above. The
omission test is VP deletion test. This means a VP can be deleted, meaning a sequence of words
qualifies as a VP if it can be omitted from a sentence. Consider the sentences below.

(4) Jane could [VP eat apples] and Mary could [VP Ø] too.

This is an example of simple VP deletion, where the second VP (eat apples) is deleted because it
is considered redundant, for its meaning can be recovered from the meaning of the earlier VP.
Now consider the more complex VP forms below.

(5) a. Jane could have been eating apples and Mary could have been eating apples too.
b. Jane could have been eating apples and Mary could have been ----- too.
c. Jane could have been eating apples and Mary could have ----- too.
d. Jane could have been eating apples and Mary could ------ too.

Let’s now see how the two hypotheses above account for the deletions in (5b), (5c), and (5d).
Both the Complex Aux Hypothesis and the Recursive VP Hypothesis can account for the
deletion in (5b), for the two tree diagrams above treat eating apples as a phrase. Things are
different in (5c) and (5d). In (5c) the deleted word sequence is been eating apples. For the
Complex Aux Hypothesis this sequence is not a phrase because been is one of the component
parts of the Aux, while the sequence eating apples is a separate phrase contained (i.e. dominated)
by VP. Two or more words that are contained in two different phrases cannot combine (or
merge) to form a bigger phrase. Therefore this sentence should be considered ungrammatical in
the analysis of the Complex Aux Hypothesis. The fact that (5c) is grammatical suggests that the
complex Aux Hypothesis makes a wrong prediction. Things will be different when we analyze
the deletion in (5a) along the line of Recursive VP Hypothesis. In this hypothesis, the deleted
sequence been eating apples is a phrase as can be seen from the tree diagram in (3) above where
the sequence is a VP that serves as the complement of the V have. Therefore the deletion is
lawful, for only a phrase can be deleted according to the constituent tests presented before. The
result of this deletion is a grammatical sentence as predicted by the Recursive VP Hypothesis.
The same line of reasoning can be extended to account for the grammaticality of (5d). Again this
sentence is predicted ungrammatical by the Complex Aux Hypothesis because again the
sequence of have been eating apples is not considered a phrase, so it cannot be deleted. The fact
that the deletion results in a grammatical sentence shown by (5d) above suggests that the
sequence have been eating apples must be a phrase, and that the deletion is lawful. The
grammaticality of (5c) and (5d) confirms the superiority of Recursive VP Hypothesis over the
Complex Aux Hypothesis. And this is the line of reasoning that we will follow in our analysis of
complex verb forms.

Verb Form

The analysis of auxiliaries has presented two hypotheses, the Complex Aux Hypothesis and
the Recursive VP Hypothesis. We have also demonstrated through the data of VP deletion that
the Recursive VP Hypothesis is superior over the Complex Aux Hypothesis. We will soon see
that the order of affixes (i.e. suffixes) in complex verb forms also presents two competing
hypothesis; the Linear Order Hypothesis and the Affix Hopping Hypothesis. Take for example
the sentences below:

(6) a. Jane could have been cooking rice in the kitchen.


b. Jane could have cooked rice in the kitchen.
c. Jane could be cooking rice in the kitchen.
d. Jane has cooked rice in the kitchen.
e. Jane is cooking rice in the kitchen.

In the Linear Distribution Hypohtesis the order of suffixes in the Aux complex will be given
several possibilities as follows:

(7) Aux → (Modal) have be -en V -ing


(Modal) have –en
(Modal) V –ing
etc. (added with passives, this list will be much more complex)

There are several objections that can be raised regarding this analysis of Aux. First, it has been
demonstrated in the case of Complex Aux Hypothesis above that linear ordering of elements
within Aux complex fails to account for the grammaticality of deleted VPs in sentences like (5c)
and (5d) above. Second, we have to make many specific rules of possible orders of V + suffix to
account for various sentences that contain complex verb forms; and third, the analysis fails to
make the key generalization, for example, that a verb must take an –ed/en suffix when it is
preceded by have.
The Affix Hopping Hypothesis

There are two basic assumptions underlying the Affix Hopping Hypothesis.
1. The orders of [have -en] and [be -ing) are each considered as single lexical entries,
meaning that each of these lexical entries consists of two morphemes; i.e a base + an
affix.
2. There is a so called Affix Hopping Rule that says: When you encounter the sequence of
‘affix – verb’, move the affix to the immediate right of the verb.

Let’s have a look of how this Affix Hopping Rule works. Consider the sentence:

Jane might have been cleaning the room.

Assuming that in the Affix Hopping Hypothesis the sequence of have -en and be -ing are treated
as single lexical entries, the tree diagram for the above sentence must be as follows:

NP Aux VP

N M V VP
[+Aux]
V VP
[+Aux]
V NP
[-Aux]
D N

Jane might have -en be -ing clean the room

As we can see in this tree diagram, the output of these phrase structure rules is an ungrammatical
sentence (i.e. grammatically ill-formed). What does Jane might have-en be-ing clean the room
mean? In the framework of Transformational Grammar this is what we call a deep structure or
D-structure (a stucture as the output of phrase structure rules or PS rules). This structure is
considered abstract. We don’t use it in actual speech. What we use in actual speech is a surface
structure or an S-structure, a structure resulted from transformational rules. The
transformational rule used to bring the deep structure to the surface structure is Affix Hopping
rule, which is informally stated as follows:

Affix Hopping Rule


When there is a sequence of ‘affix – verb’, move the affix to the immediate right of the verb.

Applying this transformational rule to the output of the tree diagram above, we come up with a
tree diagram of a surface structure as follows.
S

NP Aux VP

N M V VP
[+Aux]
V VP
[+Aux]
V NP
[-Aux]
D N

Jane might have be-en clean-ing the room

As we can see in this diagram now the output structure is grammatical already. This final surface
structure is reached after the operation of Affix Hopping rule which hops the suffix -en to the
immediate right of the verb be and affix -ing to the immediate right of verb clean. We have to
make it clear that the Affix Hopping rule is not a phrase structure (PS) rule. A PS rule states the
mother-daughter-sister relation; while Affix Hopping changes the structure (a structure-changing
rule). In the framework of Transformational Grammar, Affix Hopping rule is called a
transformation rule.
The adoption of the notion of Affix Hopping brings a big consequence to the theory of
sentence structure analysis in which we have to assume that there must be two types of structure;
deep structure and surface structure. These two types of structure are connected with
transformational rules or simply called transformations in the framework as follows:

Surface structure (S-structure)

Transformational rules (transformations)

Deep structure (D-structure)

In short we can say that D-structure is the product of PS rules. It is abstract in the sense it is not
used in actual speech or writing. The D-structure as the product of PS rules sometimes doesn’t
look like a grammatical sentence. S-structure is the product of transformational rules. It is
derived from the D-structure by the use of transformations. It is this structure that is used in
actual speech or writing.
In the next meeting we will take a closer look at how the theory of D-structure –
Transformations – S-structure works in the analysis of sentences with complex verb forms.

Exercises
Analyze in tree diagrams the structure of the following sentences along the line of Recursive VP
Hypothesis and Affix Hopping (AH) Hypothesis.

1. The students are playing the new game.


2. My brother has moved the desks
3. The farmers will have been plowing their fields.
4. Jane might be telling Kathie a story about ghosts.
5. Jane has said that Kathie is cleaning her room.

ooooo000ooooo

You might also like