|Legendre, Géraldine, Jane Grimshaw, and Sten Vikner, ed. (2001) Optimality-
Theoretic Syntax. MIT Press, xviii+548pp, paperback ISBN 0-262-62138-X,
Jonathan White, Högskolan Dalarna, Sweden
Chapter 1: An introduction to Optimality Theory in Syntax (Legendre)
This introduction sets out the rationale behind Optimality theory and its
application to syntax. In generative syntax, economy principles are strictly
applied. That is, if one is violated, that derivation is finished ("crashes"
in Chomsky's 1993 terms). In Optimality Theory (OT), some constraints are
seen as being more important than others. This is achieved by hierarchically
ordering them. A derivation that violates some low ranking constraints is
preferred over one that violates a high ranking one, but no low ranking
ones. This is the general framework the remaining papers in the volume are
Chapter 2: Competition between Syntax and Morphology (Ackema and Neeleman)
This chapter deals with the fact that a particular meaning can be encoded as
a full syntactic structure ("drive a truck") or as a morphological compound
("truck driver"). The proposal is that the syntactic and morphological
components of grammar compete. It is the presence of the affix in the input
which is crucial, in that it requires a morphological host. If there is no
affix present, the syntactic path is taken.
Chapter 3: Markedness and subject choice in OT (Aissen)
This paper seeks to explain the fact that certain types of subject are more
marked than others. This markedness may be expressed through morphology, for
example a subject may appear as an ergative among other options. The
solution offered is to adopt a solution from OT Phonology, namely Harmony of
the input and output. Underlying this are hierarchies of person and animacy,
and of semantic role and grammatical role. The advantage of this approach
according to Aissen is that it eliminates the need for language-particular
Chapter 4: Optimality and ineffability (Bakovic and Keer)
This deals with a classic case of optionality in syntax, the Complementizer
in English. The solution is to rank markedness constraints and faithfulness
ones. The Complementizer is optional if the faithfulness constraints are
ranked above the markedness ones, since, thereby, different inputs will have
different outputs. If the markedness ones are ranked higher, the
Complementizer is obligatory since the output will be the same no matter
what the input.
Chapter 5: The emergence of the unmarked pronoun (Bresnan)
Personal pronouns may appear as full pronouns, weak forms or as clitics.
Harmony constraints are argued to govern the choice, with the pronoun form
linked to the properties of anaphoricity, topic-hood and agreement. Reduced
forms only appear if the pronoun is a topic; person and number features are
only present if the pronoun is overt.
Chapter 6: Binding and discourse prominence: reconstruction in "focus"
Scrambling is shown to have mixed properties. It acts as a wh-movement-type
operation, in that displays reconstruction effects, while it is similar to
raising in admitting new binding relations in the target position.
Scrambling as a phenomenon is explained by the interaction of discourse
constraints that old and prominent information should precede new and
non-prominent information with the syntactic one that subjects precede
non-subjects. The binding facts stem from constraints that the binder must
precede the element it binds and also that the former must outrank the
latter on a functional hierarchy.
Chapter 7: The emergence of unmarked word order (Costa)
This paper examines how unmarked word order in languages comes about. The
explanation is that syntactic constraints like the requirement that
languages have subjects interact with discourse information. Discourse
constraints outrank syntactic ones, so that focusing an element affects word
order. If no focus is required, the syntactic constraints ensure the
Chapter 8: Optional clitic positions and the Lexicon (Grimshaw)
Two separate facts are explained in this chapter. Firstly, the actual clitic
forms that are used depend on faithfulness constraints. That is, the input
is matched with a particular clitic form. Then, the order of clitics is
derived through a set of language-specific positional constraints.
Chapter 9: Masked second-position effects and the linearization of
functional features (Legendre)
The verb second effect is compared to the clitic second effect in many
languages. Legendre seeks a phonologically-based, rather than a
syntactically-based, explanation for both. Specifically, two constraints
interact, one that the element must not be realised at the edge of a
prosodic domain, and the other that the element must be realised at the left
edge of the domain of the relevant syntactic head. If the first constraint
outranks the second, the element appears in second position. With the other
order of the constraints, the element appears at the front of the clause.
Chapter 10: Order preservation, parallel movement and the emergence of the
C-command relations between elements are preserved across levels of
syntactic representation. This constraint interacts with those that require
or prohibit movement, such as the wh-criterion. If such a criterion may be
satisfied by two different structures, one involving movement and one with
no movement, the structure that preserves the c-command relations between
elements is chosen.
Chapter 11: Cross-linguistic typologies in OT (Samek-Lodovici)
SVO and VSO orders are derived through the interaction of constraints on
movement and phrase structure. Language-particular choices determine which
orders are possible, and in which situations. The effect of discourse is one
factor which affects typologies.
Chapter 12: Form and function in the typology of grammatical voice systems
The relative markedness of particular grammatical functions is examined by
Sells. For instance, in an active construction, the agent is more prominent
than a patient. In a passive, the situation is reversed. Prominence
hierarchies ensure that the output stays faithful to the input.
Chapter 13: Constraints on null pronouns (Speas)
The environments where null pronouns are allowed are argued to be governed
by a number of constraints, those relating to binding and control are
particularly relevant. The relative ranking of these constraints allows us
to derive the facts of many languages such as English (the null pronoun can
only be the subject of a non-finite clause), Spanish (the null pronoun can
be any subject) and Thai (the null pronoun can be any subject or object).
Chapter 14: V-to-I movement and "do"-insertion in OT (Vikner)
The position of a finite verb with respect to sentential adverbials and to
negation is analysed from the perspective of a number of languages.
Language-specific constraint rankings explain the different orders.
Chapter 15: Bi-directional optimization and the theory of anaphora (Wilson)
Two different situation are argued to exist in the theory of anaphora. One
is that Wilson calls an interpretive competition, where one syntactic
structure allows two interpretations. In English, this situation arises when
an anaphor can be associated with one of two NPs. There, the closest NP is
the one chosen. The second situation is an expressive competition, where one
interpretation may be realised in two syntactic structures. In Icelandic, a
single NP may be associated with either a pronoun or an anaphor, depending
on the syntactic situation.
Chapter 16: Case patterns (Woolford)
There are some examples where a simple view of Case, that it is licensed by
a head, has problems, namely when a particular Case is strongly dependent on
other factors. For instance, in Hindi, the case of the object is strongly
dependent on the case of the subject. Burzio's generalisation also comes
under this heading, in that case depends on the transitivity of the verb.
Woolford derives these problematic examples through a combination of
constraints against marked Cases and those ensuring faithfulness to lexical
requirements of Case.
All in all this volume is a very impressive presentation of current research
in Optimality-Theoretic Syntax. It does, though, show up the fact that the
approach is a new one. The issue that many of the papers come back to is
what the input to the Evaluation process actually comprises. Here the debate
that is on-going in syntactic theory in general is apparent. Within the
functional tradition, the role of discourse in determining syntactic form is
highly important (Givón 1997). Whereas within the generative tradition, this
is not recognised to the same degree. As a result of this, we can argue that
Optimality Theory as applied to syntax is really a meta-theory. That is, it
is simply a particular view about how to formulate syntactic theory. There
can be functional syntactic OT accounts of phenomena, and there can be more
generative accounts (Newmeyer 2000). This should not be seen as a criticism
of the approach, but it is an important point that needs to be made.
Chomsky, N. 1993. A Minimalist Program for Linguistic Theory. In Hale and
Keyser (eds.). The view from Building 20. Cambridge, Mass. MIT Press.
Givón, T. 1997. Syntax. Amsterdam, John Benjamins.
Newmeyer, F. 2000. Optimality and functionality. Rutgers Optimality Theory
REVIEWER'S RESEARCH INTERESTS
Phrase structure, syntax and semantics of adverbials, interfaces between
syntax and semantics and between syntax and morphology.