Word
Grammar
New
Perspectives
on a
Theory
of
Language Structure
This page intentionally left blank
Word
Grammar
New
Perspectives
on a
Theory
of
Language Structure
edited
by
Kensei Sugayama
and
Richard
Hudson
continuum
Continuum
The
Tower Building
15
East 26th Street
11

in
writing
from
the
publishers.
First
published
2006
British Library Cataloguing-in-Publication Data
A
catalogue
record
for
this
book
is
available
from
the
British Library.
ISBN:
0-8264-8645-2 (hardback)
Library
of
Congress Cataloguing-in-Publication Data
To
come
Typeset
by
BookEns

This page intentionally left blank
Contents
Contributors
Prefac e
Kensei
Sugayama
Introduction
1.
What
is
Word
Grammar?
Richard
Hudson
1.
A
Brief Overview
of the
Theory
2.
Historical Background
3. The
Cognitive Network
4.
Default
Inheritance
5. The
Language Network
6.
The

Chet Creider
and
Richard Hudson
1.
Introduction
2.
The
Data
3. The
Analysis
of
Case Agreement
4.
Non-Existent Entities
in
Cognition
and in
Language
5.
Extensions
to
Other Parts
of
Grammar
6.
Comparison
with
PRO and pro
7.
Comparison

21
24
27
28
33
35
35
35
41
42
46
49
50
52
54
54
56
viii
WORD
GRAMMAR:
PERSPECTIVES
ON
LANGUAGE STRUCTURE
3. Eat in
English
4.
Taberu
in
Japanese
5.

6.
Semantics
of the Be To
Construction
7.
Should
To be
Counted
as
Part
of the
Lexical Item?
8. A
Word
Grammar Analysis
of the Be To
Construction
9.
Conclusion
5.
Linking
in
Word
Grammar
Jasper
Holmes
1.
Linking
in
Word

Grammar Approach
to
Code-Mixing
4.
Word
Order
in
Mixed
and
Monolingual 'Subordinate' Clauses
5.
Summary
and
Conclusion
7.
Word
Grammar
Surface
Structures
and
HPSG
Order
Domains
Takafumi
Maekawa
1.
Introduction
2.
A
Word

Introduction
2.
Structural Heads
58
60
63
67
67
68
69
70
71
72
75
77
81
83
83
103
114
117
117
118
121
128
139
145
145
146
154

Words
10.
Attributive
Adjectives
11.
Determiner Phrases
12. The type
of
Construction
13.
Inside-out Interrogatives
14.
'Empty Categories'
15.
Coordination
16.
Correlatives
17.
Dependency Types
18.
Conclusion
9.
Factoring
Out the
Subject Dependency
Nikolas
Gisborne
1.
Introduction
2.

191
199
204
204
205
210
216
222
225
227
229
This page intentionally left blank
Contributors
RICHARD
HUDSON
is
Professor Emeritus
of
Linguistics
at
University College
London.
His
research
interest
is the
theory
of
language structure;
his

in
educational linguistics.
Website:
www.
phon.
ucl.
ac.
uk/home/dick/home.
hrm
Email:
dick@linguistics.
ucl.
ac. uk
KENSEI
SUGAYAMA,
Professor
of
English Linguistics
at
Kobe
City
University
of
Foreign Studies. Research interests: English Syntax,
Word
Grammar, Lexical
Semantics
and
General Linguistics.
Major

in
Word
Grammar' (1999).
The
Kobe
City
University
Journal
50. 7;
Scope
of
Modern
Linguistics
(2000,
Tokyo:
Eihosha);
Studies
in
Word
Grammar
(2003, Kobe: Research Institute
of
Foreign Studies,
KCUFS).
Email:
ken@inst.
kobe-cufs.
ac.
jp
CHET

in
Norwegian
(1987,
University
of
Trondheim
Working
Papers
in
Linguistics
4); The
Syntax
of the
Nilotic
Languages:
Themes
and
variations
(1989, Berlin: Dietrich
Reimer);
A
Grammar
of
Nandi
(1989, with
J. T.
Creider, Hamburg:
Helmut
Buske);
A

of
English grammar.
Email:
a.
rosta@v21.
me. uk
NIKOLAS
GISBORNE
is a
lecturer
in the
Department
of
Linguistics
and
English
Language
at the
University
of
Edinburgh.
His
research interests
are in
lexical
semantics
and
syntax,
and
their interaction

lexicography, education
and IT.
Teaching
and
research interests
include
syntax
and
semantics, lexical structure, corpuses
and
other
IT
applications
(linguistics
in
computing, computing
in
linguistics),
language
in
education
and in
society,
the
history
of
English
and
English
as a

encyclopedia' (2000,
with
Richard Hudson),
in
B.
Peeters (ed.
)
The
Lexicon-Encyclopedia
Interface
(Amsterdam: Elsevier);
'Constructions
in
Word
Grammar' (2005,
with
Richard Hudson)
in
Jan-Ola
Ostman
and
Mirjam
Fried (eds)
Construction
Grammars:
Cognitive
Grounding
and
Theoretical
Extensions

Computer
brauchst'
es ja
nicht
zeigen.
":
because
+
German main clause word
order'
International
Journal
of
Bilingualism
8. 2
(2004),
pp.
127-44.
Email:
evieppler@hotmail.
com
TAKAFUMI
MAEKAWA,
PhD
student, Department
of
Language
and
Linguistics,
University

3-6.
Email:
maekawa@btinternet.
com
Preface
Thi s
volume comes
from
a
three-year (April 2002-March 2005) research
project
on
Word
Grammar supported
by the
Japan
Society
for the
Promotion
of
Science,
the
goal
of
which
is to
bring together
Word
Grammar linguists
whose

papers
was
planned
so as to
introduce
the
readers into this theory
and to
include
a
diversity
of
languages,
to
which
the
theory
is
shown
to be
applicable, along
with
critique
from
different
theoretical orientations.
In
September 1994 Professor Richard Hudson,
the
founder

has
been writing
in a
very
engaging
and
informative
way for
about
two
quarters
of a
century
in the
world linguistics scene.
Word
Grammar
is a
theory
of
language structure which Richard Hudson,
now
Emeritus Professor
of
Linguistics
at
University College London,
has
been
building since

Dependency Grammar,
his own
invention.
Word
Grammar
fills
a gap in the
study
of
dependency theory. Dependency
theory
may not
belong
to the
mainstream
in the
Western
World,
especially
not
in
America,
but it is
gaining more
and
more attention, which
it
certainly
deserves.
In

has
been introduced into virtually
all
modern
linguistic
theories.
In
most grammars, dependency
and
constituency
are
used
simultaneously.
However, this adduces
the
risk
of
making these grammars
too
powerful.
WG's challenge
is to
eliminate constituency
from
grammar except
in
coordinate structures, although certain dependency grammars, especially
the
German ones,
refuse

Systemic
Grammar,
North
Holland,
1971);
and his
second book
was
about Daughter-Dependency Grammar
(Arguments
for a
Non-transforma-
tional
Grammar,
University
of
Chicago Press, 1976).
As the
latter
tide
indicates,
Chomsky's transformational grammar
was
very
much
'in the
air',
and
both
books accepted

by
Paul Schachter
in
Language
54,
348-76).
His
exploration
of
various general ideas that
hadn't
come together
became
an
alternative coherent theory called
Word
Grammar,
first
described
in the
1984 book
Word
Grammar
and
subsequently improved
and
revised
in
the
1990 book

the
publication
of
Richard
Hudson's
Word
Grammar
(1984)
and
this volume
is
more than
two
decades
(21
years
to be
precise).
The
intervening
years
have seen impressive developments
in
this theory
by the WG
grammarians
as
well
as
those

John
Anderson
and
other users
of
Dependency Grammar,
via
Daughter-Dependency Grammar;
a
reaction against Systemic Grammar
where
word-word dependencies
are
mediated
by the
features
of the
mother
phrase.
)
• It does not use phrase structure - e. g. it does not recognize a noun phrase
as
the
subject
of a
clause, though these phrases
are
implicit
in the
dependency structure. (This

current Transformational Grammar.
)
• It
uses default inheritance,
as a
very
general
way of
capturing
the
contrast
between
'basic'
or
'underlying' patterns
and
'exceptions'
or
'transforma-
tions' - e.
g.
by
default,
English words
follow
the
word they depend
on, but
exceptionall y
subjects

(i.
e. all
links
in the
network) have equal status, though some
may for
pragmatic
reasons
be
harder
to
override than others. (From Lakoff
and
early
Cognitive Linguistics, supported
by
work
in
sociolinguistics).
• It presents language as a network of knowledge, linking concepts about
words , their meanings, etc.
- e.
g.
twig
is
linked
to the
meaning 'twig',
to the
form

able
to
incorporate
sociolinguisti
c
facts
- e.
g.
the
speaker
of
jazzed
is an
American. (From
Sociolinguistics).
In
this theory, word-word dependency
is a key
concept,
upon
which
the
syntax
and
semantics
of a
sentence
build.
Dependents
of a

in WG, as
shown
in
Figure
1:
Figure
1
xvi
WORD
GRAMMAR:
PERSPECTIVES
ON
LANGUAGE
STRUCTURE
Contributors
to
this volume
are
primarily
WG
grammarians across
the
world
who
participated
in the
research organized
by
myself,
and I am

how
powerful
WG is to
offer
analyses
for
linguistic
phenomena
in
various languages.
The
papers
we
have collected come
from
varying
perspectives
(formal,
lexical-semantic, morphological, syntactic,
semantic)
and
include work
on a
number
of
languages, including English,
Ancient Greek, Japanese
and
German. Phenomena studied include verbal
inflection,

with
a
chapter
on WG by
Richard
Hudson
which serves
to
introduce
the
newest
version
of WG. The
subsequent chapters
are
organized into
two
sections:
Part
I:
Word
Grammar Approaches
to
Linguistic Analysis:
its
explanatory
power
and
applications
Part

varied picture
of the
possibilities
of WG.
In
Chapter
2,
Creider
and
Hudson
provide
a
theory
of
covert elements,
which
is a hot
issue
in
linguistics. Since
WG has
hitherto denied
the
existence
of
any
covert elements
in
syntax,
it has to

predicted
from
the
word
on
which
it
depends.
The
analysis
given
is
interesting because
the
argument
is
linked
to
dependency.
It is
more
sophisticated than
the
simple
and
undefined Chomskyan notion
of PRO
element.
In
Chapter

quite unique
in
allowing
its
verbs
to
miss
out
their complements
on the
condition that
the
speaker assumes that they
are
known
to the
addressee.
The
reason seems
to be
that
in the
semantic structure
of
the
sentences, there
has to be a
semantic argument which should
be, but is
not, mapped onto

the
relation between
a
word
and the
construction.
In
Chapter
4, he
attempts
to
characterize
the be to
construction within
the WG
framework.
He has
shown that
a
morphological, syntactic
and
semantic analysis
of be in the be to construction provides evidence for the category of be in this
construction. Namely,
be is an
instance
of
modal verb
in
terms

semantic arguments
in the WG
approach. Under
the
WG
account, both thematic
and
linking properties
are
determined
at
both
the
specific
and the
general level.
This
is
obviously
an
advantage.
In
Chapter
6,
Eppler
draws
on
experimental studies concerning
the
code-

units like phrasal constituents.
The
Null-Hypothesis, then, formulated
in
WG
terms, assumes that each word
in a
switched dependency
satisfies
the
constraints imposed
on it by its own
language.
The
material
is
taken
from
English/German conversations
of
Jewish refugees
in
London.
Maekawa
continues
the
sequence
in
this collection
towards

Constructional
HPSG
approach,
and the
Linearization
HPSG
analysis.
Maekawa,
a
HPSG
linguist, argues that
the
approaches within
WG and the
Constructional
HPSG
have some problems
in
dealing
with
the
relevant
facts,
but
that Linearization
HPSG
provides
a
straightforward account
of

chapters that examine
two
theoretical
key
concepts
in
WG,
head
and
dependency.
They
are
intended
to
help
us
progress
a
few
steps
forward
in
revising
and
improving
the
current
WG,
together
with

claims that every phrase
has
both
a
distributional head
and a
structural head, although
he
agrees that normally
the
same
word
is
both distributional
and
structural head
of a
phrase. Finally,
Gisborne's Chapter
9
then challenges
Hudson's
classification
of
dependencies.
The
diversification
of
heads
(different

on a
review
of the
subject-of
dependency, distinguishing between
two
kinds
of
subjects,
which
seems promising.
Gisborne's
thesis
is
that word
order
is
governed
not
only
by
syntactic
information
but
also
by
discourse-presentational
facts.
I
hope this short overview

look
forward
to
future
volumes that
will
further
develop this cooperation.
The
editors
gratefully
acknowledge
the
work
and
assistance
of all
those
contributors whose papers
are
incorporated
in
this volume, including
one
non-
WG
linguist
who
contributed papers
from

Daiwa Anglo-Japanese Foundation, whose
assistance
we
gratefully
acknowledge here.
In
addition,
we owe a
special debt
of
gratitude
to
Jenny Lovel
for
assisting
with
preparation
of
this volume
in her
normal professional manner.
We
alone accept responsibility
for all
errors
in
the
presentation
of
data

—
(2004,
July
1-last
update),
'Word
Grammar',
(Word Grammar),
Available:
www.
phon.
ucl.
ac.
uk/home/dick/wg.
htm
(Accessed:
18
April
2005).
—
(in
preparation),
Advances
in
Word Grammar.
Oxford:
Oxford
University
Press.
Pollard,

Research
Institute
of
Foreign
Studies,
KCUFS.
Tesniere,
Lucien
(1959),
Elements
de
Syntaxe Structurale.
Paris:
Klincksieck.
Introduction
This page intentionally left blank
1
What
is
Word
Grammar?
RICHARD
HUDSON
Abstract
The
chapter
summarizes
the
Word
Grammar

3. 3
Modularity;
4.
Default
inheritance;
5. The
language
network;
6. The
utterance
network;
7.
Morphology;
8.
Syntax;
9.
Semantics;
10.
Processing;
and 11.
Conclusions.
1.
A
Brief Overview
of the
Theory
Word
Grammar
(WG)
is a

been addressed
at all are
phonology
and
language acquisition (but even
here
see van
Langendonck 1987).
The aim of
this article
is
breadth rather than
depth,
in the
hope
of
showing
how
far-reaching
the
theory's tenets
are.
Although
the
roots
of WG lie
firmly
in
linguistics,
and

theory
has
been developed
from
the
start
with
the aim of
integrating
all
aspects
of
language
into
a
single dieory which
is
also compatible
with
what
is
known about general
cognition.
This
may
turn
out not to be
possible,
but to the
extent that

language structure. However,
our
assumptions make
a
great deal
of
difference
when approaching these
facts,
so it
is
possible
to
arrive
at
radically
different
analyses according
to
whether
we
assume
that language
is a
unique module
of the
mind,
or
that
it is

and
largely
successful,
as we
shall
see
below.
4
WORD
GRAMMAR:
PERSPECTIVES
ON
LANGUAGE
STRUCTURE
As
the
theory's name suggests,
the
central unit
of
analysis
is the
word, which
is
central
to all
kinds
of
analysis:
• Grammar. Words are the only units of syntax (section 8), as sentence

of
syntax,
but
also
the
smallest.
In
contrast
with
Chomskyan linguistics,
syntactic
structures
do
not,
and
cannot, separate stems
and
inflections,
so
WG is an
example
of
morphology-free
syntax
(Zwicky
1992: 354).
Unlike
syntax,
morphology (section
7) is

syntax
and
semantics,
giving
a
radically 'lexical' semantics.
As
will
appear
in
section
9, a
rather
unexpected
effect
of
basing semantic structure
on
single words
is a
kind
of
phrase structure
in the
semantics.
• Situation. We shall see in section 6 that words are the basic units for
contextua l analysis
(in
terms
of

related
concept
G
relationship
of C to
notation
in
diagram
cycled
the
word
/
the
word
to
the
morpheme
{cycle}
the
word-form
{cycle+ed}
the
concept
'ride-bike'
the
concept
'event
e'
the
lexeme

line
triangle
resting
on
CYCLE
'speaker'
'time'
As can be seen in this diagram, cycled is the meeting point for ten
relationships which
are
detailed
in
Table
1.
These
relationships
are all
quite
traditional (syntactic, morphological, semantic, lexical
and
contextual),
and
traditional names
are
used where they exist,
but the
diagram uses notation
which
is
peculiar

meaning.
All
these elements, including
the
words themselves,
are
'concepts'
in the
standard sense; thus
a WG
diagram
is an
attempt
to
model
a
small
part
of the total
conceptual network
of a
typical speaker.
2.
Historical
Background
The
theory described
in
this article
is the

have preserved some
of the
most
fundamental
ideas
- the
central place
of the
word,
the
idea that language
is a
network,
the
role
of
default inheritance,
the
clear separation
of
syntax
and
semantics,
the
integration
of
sentence
and
utterance structure.
The

and
students.
The
following brief history
may
be
helpful
in
showing
how the
ideas that
are now
called
'Word
Grammar'
developed during
my
academic
life.
The
1960s.
My PhD
analysis
of
Beja used
the
theory being developed
by
Halliday (1961) under
the

conditions,
I
published
the
first
generative version
of
Halliday's Systemic
Grammar
(Hudson
1970).
This
theory
has a
very large network (the 'system
network')
at its
heart,
and
networks also loomed large
at tihat time in the
Stratificational
Grammar
of
Lamb (1966; Bennett 1994). Another reason
why
stratificational
grammar
was
important

the
addition
of
word-word dependencies under
the
influence
of
Anderson (1971);
the
theory
was
called 'Daughter-Dependency
Grammar' (Hudson 1976). Meanwhile
I was
teaching sociolinguistics
and
becoming increasingly interested
in
cognitive science (especially
default
inheritance
systems
and
frames)
and the
closely related
field of
lexical
semantics (especially
Fillmore's

1981):
everything
in the
grammar
is
'lexical'
in the
sense that
it is tied to
word-sized units (including word classes).
The
1980s.
All
these influences combined
in the first
version
of
Word
Grammar (Hudson 1984),
a
cognitive theory
of
language
as a
network
which
contains both 'the grammar'
and
'the lexicon'
and

of
continental depen-
dency theory
is
evident
but the
dependency structures were richer than those
allowed
in
'classical' dependency grammar (Robinson 1970)
-
more like
the
functional
structures
of
Lexical Functional Grammar (Kaplan
and
Bresnan
1982). Bresnan's earlier argument (1978) that grammar should
be
compatible
with
a
psychologically plausible parser also suggested
the
need
for a
parsing
algorithm, which

as a
detailed application
to
large
areas
of
English morphology,
syntax
and
semantics.
The
1990s.
Since
the
publication
of EWG
there have been some important
changes
in the
theory, ranging
from
the
general theory
of
default
inheritance,
through matters
of
syntactic theory
(with

will
be
described
below.
WG has
also
been
applied
to a
wider range
of
topics than previously: