Extensions to

Syntax
Chapter 8
Table of contents

- OT and syntax.
- OT-based syntax: general assumptions.
- Defining the input of syntax.
- Defining Gen for syntax.
- Defining the syntactic constraint in inventory.
- The structure of extended verbal projections in English.
- Wh-movement and subject-auxiliary inversion.
- Do-support.
-Typological consequences.
- The general typology.
- Variations on subject-verb inversion
OT and Syntax
 Introduction

 This chapter is about applying OT in syntax, it looks into the results of

OT within syntax, it emphasizes that OT is a theory of grammar and
not only restricted to phonology which means it can be applied to non-
phonological phenomena.-Reference to Grimshaw's paper (1997) on
English auxiliaries as the foundation for the syntactic extension of OT.
 OT AND SYNTAX
 Comparison between 'Minimalist' syntactic theory and OT-based syntax, highlighting
both similarities and differences, for example: Parameters vs. Constraints: Minimalist
Syntax: Assumes universal principles with language-specific parameters that can be
switched on or off. OT-based Syntax: Utilizes universally ranked constraints, offering a
different way to achieve typological diversity by reranking constraints rather than
switching parameters.

 Universal Principles: Both theories acknowledge a core set of universal principles that
govern linguistic structures. Typological goals: Both aim to account for linguistic
typology by explaining the diversity of language structures and patterns.
 Parameters in Syntax

 Explanation of the traditional view of parameters in Principles-and-

Parameters Theory in which it is assumed that grammatical principles
are universal, in the sense of being inviolable.Contrast with OT's
approach where constraints are universal and violable, emphasizing
that constraints are not switched off but dominated.
Constraint Reranking vs. Parameter Setting

 Languages achieve typological diversity through either switching off

parameters or dominating constraints.Emphasis on the key difference:
dominated constraints can still manifest their effects in OT.
 Economy in Syntax

 The concept of 'economy' in 'Minimalist' syntactic theory,

emphasizes the minimally necessary conditions for well-formedness.
Being in parallel with OT's 'do only when necessary' principle and
the interaction of faithfulness and well-formedness constraints.
 Economy in OT vs. Minimalist Syntax

 The soft nature of economy in OT: OT allows flexibility and

adaptability compared to the hard system in Minimalist syntax
referring to the strict principles governing the derivation of
linguistic structures. -Reference to ongoing attempts by
syntacticians to explain economy using OT assumptions.
 OT-based syntax: general assumptions
 Universality, violability: Constraints are universal and can be violated

 Strict domination: Constraints are hierarchically ranked in a grammar

 Freedom of analysis: the ability to generate any structure.

 Harmony: Candidate outputs are ranked by the grammar according to

their relative harmony with the constraint hierarchy.
Defining the input of
syntax
Input
• a lexical head plus its argument structure.
• an assignment of lexical heads to its arguments.
• a specification of the associated tense and
semantically meaningful auxiliaries
For Example:

Mary did the vocal parts

Do (x,y)
What did Mary say?

say (x, y)
x = Mary
y = what
tense = past
Containment
1- competing candidates are evaluated as analyses of the
same lexical material.
2- competing candidates to be generated for a single input
must be semantically equivalent.
Defining Gen for syntax
 Grimshaw defines Gen as a function generating all
possible analyses of an input within the universal structural
requirements of X′ theory.

X′ theory (respected under analysis by Gen): each node

must be a good projection of a lower one, if a lower one is
present.
Extended projection:

A unit consisting of a lexical head and its

projection plus all functional projections erected
over the lexical projection.
Operations allowed under analysis by Gen
- introducing (extended) projections conforming to X¢ theory
- introducing functional heads as they do not appear in the input, due to their
lack of semantic content (e.g. the complementizer that and do-support in
English)
- introducing empty elements (traces, etc.), as well as their coindexa- tions with
other elements
moving lexical elements.
What did Mary say?
Various candidate analyses for {say (x, y), x  Mary, y  what,
tense  past)
a- [IP Mary [VP said what]]
b- [CP what [IP Mary [VP said t ]]]
c- [CP what saidi [IP Mary [VP ei t ]]]
d-[CP what didi [IP Mary ei [VP say t ]]]
8.2.4
Defining the syntactic
constraint inventory
Operator in Specifier (Op-Spec)
Syntactic operators must be in specifier position.

 The primary function of this constraint is to force the movement of

Wh-elements (which are syntactic operators) into the specifier
position of some extended projection typically Spec-of-CP.
 Wh-movement into Spec-of-CP
produces a chain of effects in
English syntax, which are known as
subject–auxiliary inversion and
do-support.
Obligatory Heads (Ob-Hd)
A projection has a head.

 For verbal projections, Verb Phrase and its extended projections IP

and CP, this implies that each must have a verb as its head.

 Satisfaction of this constraint may require the movement or insertion

of a verb to become the head of an extended verbal projection.

 For example, Ob-Hd may trigger subject–auxiliary inversion, the

movement of an auxiliary to become the head of CP.
 The auxiliary verb will undergoes
head-movement to CP, from its
original position in IP, as has been
indicated by its trace ei.

 Crucially, both the auxiliary verb and

its trace function as proper heads in
their respective extended projections,
IP and CP.
 Each of these analyses violates Ob-Hd, however, since each has
one projection that is not headed: CP in (13a) and IP in (13b).
What are the ‘costs’, stated in terms of
constraint violations, of satisfying Op-
Spec and Ob-Hd?
 The notion of faithfulness, as we have used it throughout the book, involves
evaluation of the degree of identity between the input and output.

 it seems entirely logical to conceive of an anti-movement constraint as the

syntactic counterpart of (phonological) Linearity as both militate against
differences in the order of elements in the input and the output.

 Along the same lines, a syntactic constraint ‘do not insert’ is a plausible
counterpart of Dependence.
Economy of Movement (Stay)
Trace is not allowed.

 The function of this constraint is to militate against any movement per

se, or in the case of gradient violation, to select the shortest
movement, the one that has the minimal number of intermediate
steps.
 The chain of a Wh-element and its trace ti in (a) incurs one violation of
Stay, while the chain of an Aux-element and its two traces ei in (b) incurs
two violations:
No Movement of a Lexical Head (No-Lex-Mvt)
A lexical head cannot move.

 The effect of this constraint is that a lexical head (for example, a

lexical verb) must stay in the projection that it heads (e.g. the VP).
  The configuration below presents the typical case of violation
of No-Lex-Mvt

Note: No-Lex-Mvt is a special case of the general anti-movement

constraint Stay. That is, every violation of No-Lex-Mvt implies a violation
of Stay, but not conversely.
Full Interpretation (Full-Int)
Lexical conceptual structure is parsed.

 The key function of this constraint is to ban semantically empty auxiliary

verbs, as in do-support in English.
 In a sentence such as What did Mary say? the verb do is semantically
empty, functioning merely as an auxiliary for the lexical verb say.

 It will be assumed that do in such sentences is lacking from the input,

and inserted from the lexicon into candidate analyses by Gen. ( Gen is
‘semantically neutral’)
 Observe that the lexical verb do is a theta-marker in a sentence such as
Mary did the vocal parts. This can be represented as below:

do (x, y)

 Under do-support, this lexical conceptual structure of the verb do is not

parsed, hence the sentence What did Mary say? incurs a violation of Full-
Int.
 Full-Int expresses a faithfulness requirement with
respect to lexical elements, which is violated if
the output ‘suppresses’ their theta-marking and
argumenttaking properties.
 8.3
The Structure of Extended Verbal
Projections in English
8.3.1. Wh-movement and subject auxilary inversion
a. Mary will say too much.
b. *Will Mary say too much.

- Each projection is headed by an appropriate head

- VP is headed by lexical verb “say”
- IP is headed by the auxiliary verb “will”
- Since no heads are missing from this structure, there is no need for
an additional CP (see the structure below) :

This structure has unnecessary complexities in comparison to (a) : -It

has an additional extended projection CP
- It has an additional movement to fill the head of CP
Conclusion : Such avoidance of unnecessary structure and movement is
what Chomsky’s Minimalist programme refers to as ‘economy’

.
.
.
-In OT, ‘economy’= ‘do only when necessary’= inversion in the absence of any structural necessity leads
to a fatal violation of the anti-movement constraint STAY.

Conclusion : - No inversion takes place in declarative sentences because of lack of necessity .

- Both OP-SPEC and OB-HD are satisfied without movement
- Next consider interrogatives, where Wh-movement is obligatory, in combination with inversion as in
(24a):
a. What will Mary say?
b. *What Mary will say?
c. *Will Mary say what?
d. *Mary will say what?
Conclusion : -Wh-movement without inversion causes ungrammaticality and so does the lack of movement.
- Subject- auxiliary inversion occurs if and only if wh-movement applies. Conversly, inversion
is impossible elsewhere.

.
1- What makes wh-movement obligatory ?
2- What functional projections are involved in wh- movement and inversion?

-Wh-movement itself is obligatory due to high-ranking Op-Spec requiring that syntactic operators occupy a
specifier position.
- Wh-phrases are operators, and hence must move into some spec position.
- Issue : But to which spec position can a wh-phrase in sentences such as 24 move?

Let us consider these structures, both of which have wh-movement to CP:

 Conclusion

-Both movements naturally take their toll as they occur at the expense of violations of STAY, the
general anti-movement constraint. This motivates the following ranking :
Inversion only under Wh-movement.
STAY< Op-Spec, Ob-Hd.
- Op-Spec must dominate Stay in order for Wh-movement to take place; with the reverse ranking
movement would be blocked by Stay.
-And Ob-Hd must also dominate Stay in order for head movement of the auxiliary verb to take
place
 Economy and OT
- Let us now see how the property of ‘economy’ follows from OT analysis :

● Derivational Economy ● Representational Economy

- No unnecessary movement is enforced directly by - No empty projections is achieved without a

STAY constraint explicitly militating against ‘unnecessary
- Head movement of the auxiliary occurs whenever projections’
necessary - Instead, exclusion of superfluous extended
- Naturally, any movement of the auxiliary that yields projections such as CP is achieved indirectly, by
no gains with respect to OB-HD will be blocked by cooperation of two constraints, Ob-Hd and Stay.
STAY = that is in declarative sentences.
8.3.2
Do-support
The distribution of ‘Do’

Secondly, do cannot cooccur with

Firstly, do is obligatorily present in other auxiliary verbs in
simple interrogative sentences interrogatives:

a. What did Mary say? a. What will Mary say?

b. *What Mary said? b. *What does Mary will say?
c. *What will Mary do say?

Why? It is the head of CP (OB-HD). Why? For reasons of economy.

The distribution of ‘Do’

Fourthly, the occurrence of

Thirdly, do-support is impossible in auxiliary do is ruled out in
positive declarative sentences: declarative
sentences that already contain
a. Mary said much. another auxiliary verb
b. *Mary did say much.
a. What will Mary say?
b. *What does Mary will say?
Why? Superfluous (against economy). c. *What will Mary do say?
The distribution of ‘Do’

The generalization about do-

Fifthly, auxiliary do cannot cooccur support (Chomsky 1957,
with itself, even in interrogatives: 1991)

a. What did Mary say? The auxiliary do is possible only

b. *What did Mary do say? when it is necessary.
‘do only when
necessary’
Here, ‘do’ literally refers to ‘do-support’.
The analysis

the auxiliary do is a semantically

empty verb

It only serves as the head of extended

projections

The semantic emptiness of The suppression of the lexical–

auxiliary do. conceptual properties of its
source.
 Lexical verb ‘do’ = ‘to act’ = minimal
semantic content

‘Did you do your homework?’

‘do’ has the smallest lexical conceptual

Semantically Lexical structure of its own, then using its empty
Empty. Verb. verb causes the minimal suppression of
lexical
conceptual structure.
 Well-formedness
Do-support
constraint (OB-HD)

Do-support Violates FULL-INT

OP-SPEC

OB-HD
(b) Avoided violating FULL-INT but at the cost of OB-HD.
(e) to avoid violations of OB-HD by not creating extended projections. However, it violated OP-
Spec
The mising output candidate

No-Lex-Mvt
The difference between ‘do’ and the other
auxiliaries:

Do-support Other auxiliaries (can, will, may)

- Semantically empty. - Semantic and functional content.
- Not part of the input. - part of the input.
In this tableau FULL-INT dominates STAY

Why (a) is the optimal candidate although it has two

violations of STAY in comparison with (B)?
simple declarative sentences that lack an auxiliary

This type of sentence is self-supporting because there is no need for

IP nor CP (no wh-elements)
do-support never evidence that Full-Int
cooccurs with itself. dominates Stay.
Wh-objects VS. Wh-subjects

Wh-subjects Wh-objects
- can satisfy Op-Spec by staying in - must move out of VP to find a proper specifier
Spec-of-VP position.
- do-support is not needed.
Typological
consequences
 Assumptions underlying the OT approach to typology:

Constraints are universal while their While typological variation due to

ranking is language-specific. constraint reranking is considerable, it is
not unlimited.
The general typology

Syntactic constraint :

Op-Spec/ Ob-Hd/ and Stay

Restricting rerankings to a set of three core constraints, we find six logically possible
rankings:
Stay dominates both well formedness
constraints Op-Spec and Ob-Hd.
 Stay still dominates Op-Spec, and this
produces a similar lack of Wh-movement.
Op-Spec dominates Stay. Stay,
in its turn, dominates Ob-Hd
Op-Spec and Ob-Hd on top of Stay
8.4.2
Variations on subject- verb
inversion
❖Recapitulation:

Op-Spec, Ob-Hd >>Full-Int, Stay

❑This ranking is shown to be appropriate in Tableau (38)

 An alternative structure results from inverting the subject and verb

Candidate 47b

❖ Wh-movement satisfies Op-Spec,

❖ Ob-Hd the lexical verb, movement out of the VP satisfies Ob-Hd (head of (CP))

❖ Avoids do-support.

Then why should it be ruled out?

The answer resides in the undominated constraint,

(No-Lex-Mvt)
No Movement of a Lexical Head
A lexical head cannot move.

We thus extend tableau (38) with the new candidate (47b = 48e)
❑ Tableau (48):

o 1- Lexical Verb Movement Blocked

o 2- The head Position of CP Must be Filled

o 3- Insertion of "do" Form

 English thus prefers violations of Full-Int to violations of No Lex-Mvt.
Meaning
Inversion involves the auxiliary, not the lexical verb

Comparing with Dutch:

No-Lex-Mvt ranks below FullInt

Meaning
Inversion involves the lexical verb not the auxiliary

❖ Wh-movement is supported by the movement of the lexical verb to

head position of CP, thus violating No-Lex-Mvt.
❑ The following tableau contains the exact Dutch analogues of candidates occurring
in the English tableau (48):
❑ Tableau(50)

❖ The Dutch language is incompatible with a semantically empty auxiliary: it

prefers moving the lexical verb rather than inserting ‘do’ support.
Thank you for your attention!

Zahira Dafir
Roqaya Elhaouzi
Houda Amarir
Wissal Mhirig Khadija
Hajjar Ibtissam Oussou
Hiba Elhariz.