Lexical-Functional Grammar

Lexical-Functional Grammar
Lecture 1: Motivations for LFG

Stephen Wechsler
University of Texas at Austin

---

Overview

• Psycholinguistic roots of LFG

• Nonconfigurationality

• Movement Paradoxes

• (Lexicalism— next time)

---

Chomsky’s (1965) competence hypothesis


The competence model should be a component of a
larger model of language use:

‘When we say that a sentence has a certain derivation with
respect to a particular generative grammar, we say nothing about
how the speaker or hearer might proceed, in some practical or
efficient way, to construct such as derivation. These questions
belong to the theory of language use–– the theory of
performance. No doubt, a reasonable model of language use will
incorporate, as a basic component, the generative grammar that
expresses the speaker-hearer’s knowledge of the language; but
this generative grammar does not, in itself, prescribe the
character or functioning of a perceptual model or a model of
speech production.’ (Chomsky 1965:9)

---

What sort of competence model can be embedded in a
successful model of language use?

2 important results of early psycholinguistic studies
(summarized in Fodor, Bever, and Garrett 1974):

• constituent structure is psychological y real

• there is no evidence that transformations are
psychological y real

Sentences purported to result from the application of
transformations are not harder or more time-consuming
to process than transformation-less sentences.

---

Responses to this challenge for transformations:

(i) Reject the competence hypothesis; instead,
language use involves agrammatical strategies (Fodor,
Bever, and Garrett 1974). Problem: If these strategies
violate postulates of UG, then what is UG a theory of?

(i ) Ignore the empirical evidence and abandon the
attempt to unify results from theoretical syntax and
psycholinguistic studies. (Chomsky 1980, p. 191ff)

(i i) Abandon transformations and attempt to develop a
competence model which is consistent with what we
know about language processing. LFG and HPSG grew
out of this third response.

---

The mental ‘processor’ must assign appropriate
grammatical relations and thematic roles to strings of
words: who did what to whom?

The active/passive alternation.
active:
Mary has kissed the frog.
passive: The frog was kissed (by Mary).

Transformation to derive passive (NP-movement):
S
S
==>
NP
Aux
VP
NP
Aux
VP
was
V
NP
PP
was
the frog
V
NP
PP
kissed
the frog by Mary
kissed
by Mary
e

---

Instead of a passive transformation, LFG posits active
and passive verb lexical entries with different mappings
between theta-roles and grammatical functions. The
realizations of grammatical functions are constant for a
given language (SUBJ is the NP daughter of S in English,
etc.). A lexical rule explains the regularity of passive.

kiss (active):

SUBJ=kisser, OBJ=kissee



=> (lexical rule)

kissed (passive):
SUBJ=kissee, (OBLby=kisser)

---

Lexical vs. syntactic rule: why it matters

transformational rule: output trees of A-movement are
infinitely many (John’s frog was kissed by Mary; John’s
mother’s frog was kissed by Mary; etc.) Hence these
trees are not stored in the mind. To decode the thematic
roles, the processor must perform operations specified by
transformations, resulting in greater derivational
complexity (which was disconfirmed).

lexical rule: captures a pattern over a finite domain,
namely the lexicon. Indeed, lexical rules have been
replaced by a non-derivational model for lexical
alternations (Lexical Mapping Theory; cp. Bresnan and
Kanerva 1989).

---

Nonconfigurationality (Bresnan 2001, ch. 1)

Morphology competes with syntax

‘across languages, there often appears to be an inverse
relation between the amount of grammatical information
expressed by words and the amount expressed by
phrases.’ p. 6.

How are Grammatical Relations signaled?

• English: by phrase structure

• Warlpiri: by morphology (case, agreement)

---

S
English


S

qp

(↑SUBJ) = ↓
↑= ↓

NP
VP
qpqp
two small children Aux
↑= ↓

|
VP

are
ei

V
(↑OBJ) = ↓

|
NP

chasingei

D
N


|
|


that
dog

Review question: Give empirical evidence for this
constituent structure.
E.g. show that the word string chasing that dog is a
constituent, while the string chasing that is not.

---