American Sign Language: The Phonological Base

Abstract

This paper has the ambitious goal of outlining the phonological structures and processes we have analyzed in American Sign Language (ASL). In order to do this we have divided the paper into five parts. In section 1 we detail the types of sequential phenomena found in the production of individual signs, allowing us to argue that ASL signs are composed of sequences of phonological segments, just as are words in spoken languages. Section 2 provides the details of a segmental phonetic transcription system. Using the descriptions made available by the transcription system, Section 3 briefly discusses both paradigmatic and syntagmatic contrast in ASL signs. Section 4 deals with the various types of phonological processes at work in the language, processes remarkable in their similarity to phonological processes found in spoken languages. We conclude the paper with an overview of the major typed of phonological effects of ASL’s rich system of morphological processes.
We realize that the majority of readers will come to this paper with neither sign language proficiency nor a knowledge of sign language structure. As a result, many will encounter reference to ASL signs without knowing their form. Although we have been unable to illustrate all the examples, we hope we have provided sufficient illustrations to make the paper more accessible.

Full text

American Sign Language: The Phonological Base
Scott K. Liddell, Robert E. Johnson
Sign Language Studies, Volume 64, Fall 1989, pp. 195-277 (Article)
Published by Gallaudet University Press
DOI: 10.1353/sls.1989.0027
For additional information about this article
Access provided by Gallaudet University Library (2 Jun 2014 14:14 GMT)
http://muse.jhu.edu/journals/sls/summary/v1064/64.liddell.html

AMERICAN
SIGN LANGUAGE: THE
PHONOLOGICAL
BASE
Scott
K.
Liddell
and
Robert
E.
Johnson
Abstract
This
paper
has
the
ambitious
goal
of
outlining
the
phonological
structures
and
proc-
esses
we
have
analyzed
in
American
Sign
Language
(ASL).
In
order
to do
this
we
have
divided the paper
into
five
parts.
In
section
1
we
detail the
types
of
sequential
phenomena found
in
the
production
of
individual
signs,
allowing
us
to
argue
that
ASL
signs
are
composed
of sequences
of
phonological
segments,
just
as
are
words
in
spoken
languages.
Section
2
provides
the
details of
a
segmental
phonetic
tran-
scription
system.
Using the
descriptions
made
available
by
the
transcription
system,
Section
3
briefly
discusses
both
paradigmatic
and
syntagmatic
contrast
in
ASL
signs.
Section
4
deals
with
the
various
types
of
phonological
processes
at
work
in
the
language,
processes
remarkable
in
their similarity
to
phonological
processes
found
in
spoken
languages.
We
conclude
the
paper
with
an
overview
of
the major
typed
of
phonological
effects
of
ASL's
rich
system
of
morphological processes.
We
realize
that
the majority
of readers
will
come
to
this
paper
with
neither
sign
language proficiency
nor
a
knowledge
of sign
language
structure.
As
a
result,
many
will
encounter
reference
to
ASL
signs
without
knowing
their form.
Although
we
have
been
unable
to
illustrate
all
the
examples,
we
hope
we
have
provided sufficient
illustra-
tions
to
make the
paper
more
accessible.
1.
Sequential
Phenomena
in
Sign
Formation
1.1
Background
The
fact
that
all
spoken
languages
combine
meaningless
elements
to
form
meaningful
symbols
is
regarded
as
one
of
the
defining
features
of
human
language. Stokoe
(1960)
demonstrated that
ASL signs may
also
be
viewed
as
compositional
rather
than
holistic and
thereby provided
the
first
structural
evidence
that
ASL
should
be
regarded
as
a
language
rather
than
1989
Linstok
Press.
See
note
inside
front cover
ISSN
0302-1475
195

Liddell
&
Johnson
merely
a
gesture
system.
His
pioneering
work
has
had
a
profound
effect
on
all
subsequent research
into
ASL
structure.
He
proposed
that
a
sign
consists
of
three
parts
which
combine
simultaneously:
the
tab
(location
of
the
sign),
the
dez
(handshape),
and
the
sig
(the
movement). Influenced
by
the American
structuralists,
Stokoe
referred
to these
three
aspects
of
a
sign
as
'cheremes'.
He regarded
cheremes
as
meaningless
elements
which
combined
to
form
all
the
signs in
the
language,
in
a
manner
analogous
to
that
of
spoken
language
phonemes.
The
Stokoe model
has
been
adopted
almost
universally
by
sign
language
researchers.
The
most
recent
treatments
of
the
model
hold
signs
to
be
temporally unitary
phenomena,
composed
of
some
number
of
simultaneously occurring
gestural primes.
According
to
this
view
of
sign
structure,
the
entire
set of
gestures
comprising
a
sign
is
seen
to
be
analogous
to
the
set
of
articulatory
primes
that
comprise
a
single
segment
in
spoken
language
(Studdert-Kennedy
and
Lane
1980;
Klima
and
Bellugi
1979:85-194).
Differences
among
signs
are described
by
the substitution
of
primes
within
the
simultaneous
bundle.
Thus,
the
difference between
the
signs
MOTHER
(an
open
'5'
hand
touches
the
chin twice
with
the thumb)
and
FATHER
(an
open
'5'
hand touches
the forehead
twice
with
the
thumb)
is
described
as
a
difference
in
location
in
the
bundles
of
otherwise identical
primes.
Analogously,
the
difference between
[p]
and
[t] is
commonly
described
as
a
difference
in
the
place
of
articulation primes
in
bundles
of
otherwise
identical
primes.
Because of
this
view,
sign
languages
have
been
seen
to
be
unusual
in
that
meaning
is
attached
to
such
simultaneous
bundles
rather
than
to
sequences
of
such
bundles
as
it
is
in
spoken
languages.
In
this
model
of
sign
structure
(as
in
the
model of
spoken
language
segment
structure),
however,
the
claim
that
signs
are
simultaneous bundles
of
primes
is
not
a
claim
that there
are
no
sequential
events
within
signs.
It
is
a
claim
that
within
signs
sequential
organization
is
phonologically
unimportant.
Thus,
while
Stokoe
and
more
recent
workers recognize
sequence
within
signs,
they
typically
hold
it
to
be
phonologically
SLS
64

Phonological
Base
insignificant
(Stokoe
1960,
Battison
1978).
This
is
similar
to
the
recognition
that the
onset-closure
sequence
present
in
the
stops
[p]
and
[t]
is
phonologi-
cally
insignificant.
Liddell
(1984a)
argues
that
an
adequate
description
of many
phenomena
in
ASL
requires
the
recognition
of sequences of
primes,
and
demonstrates
that
such
sequences
are
capable of
signalling
contrast
among
signs.
Below
we
will
describe
several
descriptively
important
sequences
of
primes,
and
then
return
to
the
issue
of
contrast.
1.2
Handshapes
A
significant
number
of
signs
in
the
ASL
lexicon
are
produced
with
changes
in
handshape.
For
example,
UNDERSTAND
begins
with
an
S
handshape but
ends
with
a
1
handshape.
Figure
1.
UNDERSTAND
This
handshape
change
is
described
by
Stokoe
et
al.(1965)
as
a
unitary
movement
they
call
an
'opening
movement'
wherein
a
handshape
changes
from
a
'closed'
handshape
to
an
'open'
handshape.
Table
1
presents
a
SLS 64

198
Liddell
&
Johnson
SLS
64
sampling
of
signs
which
all
begin
with
an
S
handshape,
but
end
with
different handshapes.
Sign
Initial
Handshape
Final
Handshape
UNDERSTAND
S 1
THROW
S H
TWELVE
S V
SO-WHAT
S 0
FINGERSPELL-TO
S
4
GAMBLE
S
5
Table
1.
Signs
with
initial
S
handshape
changing to second
shape
Many
other
sequences
of
two
handshapes
occur
in
ASL
signs.
A
smaller
number
of
signs
are
produced
with
a
sequence
of
three
handshapes.
In
SHOCKED
the
handshape
sequence
is
S-C-S.
In
THINK-SAME-
THOUGHT
the
sequence
is
S-1-S.
In
GOVERNMENT
the
sequence
is 1-
bent
1-1.
1.3
Locations
It
is
quite common
for
the
hand to
move
from
one
location
to
another
location
during
the production
of
a
single
sign.
Such
relocations
occur
frequently
in
simple
signs
and are
especially
common
in
compound
signs,
almost
all
of
which
move
from
one
location
to
another.
The
sign
PARENTS
is
such
a
compound
sign,
derived
from
the
signs
MOTHER
and
FATHER.
It
begins
at
the
chin
(the
location
of
MOTHER)
and
then
moves
to
the forehead (the
location for
FATHER).
Table
2
lists
several
examples
of
signs
in which
the
location
of
the
hand
changes.

Phonological
Base
Figure
2.
PARENTS
Compounds
are
marked
with
a
'(C)'.
Because
Stokoe's
sign
schema
permits
a
sign
to
have
only
one
location,
his
notations
treat
relocations
in
simple
signs
as
complexes
of
movements.
Thus, for example, NAVY
might
be
said
to
be
located
at
the
left
side
of the
waist
and
then
to
move
to
the
right
and
make
contact.
The
actual location
at
the
right
side
of
the
waist
would
not
be
specified.
Compounds
are
treated
as
linked
notations
of
two
complete
signs,
each of
which
has
its
own
location.
Numerous
verbs
in
ASL
are
marked
for subject
and
object
agreement
and
typically
move
from
one
location to
another.
Table
3
shows
the
loca-
tions
involved
with
two
verbs
marked
for
agreement.
TELL
always
begins
in
contact
with
the
chin,
and
then
moves
to
a
location
which
reflects agree-
ment
with
its
object.
GIVE
agrees
with
both
its
subject and
object.
Its
initial
and
final
locations
are
determined
by
the
subject and
object agree-
ment
morphemes
which
are
inserted
into
the
verb
stem.
Subject and object
marking
is
not
capable
of being
represented
in
Stokoe's
notation
system.
We
will
discuss
this
in
more detail
in
section
1.6.
SLS 64

Liddell
&
Johnson
Sign
Initial
location
Final
Location
SANTA-CLAUS
chin chest
GOOD
chin
base
hand
NAVY
left
side
of
waist
right
side
of
waist
KING
left side of
chest right
side
of
waist
INDIAN
nose
side
of forehead
AHEM
chin chest
(C)
PARENTS
chin forehead
(C)
SON
forehead
forearm
(C)
PALE
chest face
(C)
BROTHER
forehead
base
hand
(C)
PROMISE
chin
base
hand
Table
2.
Initial
and final
locations
of
some
common
signs
Verb
SuI.
Initial
Loc.
Final
Loc.
TELL
1st
person
chin
chest
TELL 3rd
person(a)
chin place(a)
GIVE
1st
person
3rd
person(b) chest place(b)
GIVE
3rd
persona)
3rd
person(b)
place(a)
place(b)
GIVE
3rd
person(b)
1st
person
place(b) chest
Table
3.
Initial
and
final
locations for two
agreement
verbs
1.4
Movements
Stokoe's
original
observations
demonstrated
that
some
signs
require
movements to
be
carried
out
in
sequence. He
describes
CHICAGO
as
being
made
with
a
rightward
movement
followed
by
a
downward
move-
ment;
WHEN
with
a
circular
movement
followed
by
a
contacting
move-
ment;
YEAR
with
a
circular
movement
followed
by
a
contacting movement;
and
ALSO
with
a
contact movement
followed
by
a
rightward
movement,
then
another
contacting
movement.
Supalla and
Newport
(1978)
demonstrate
that
very
finely
detailed
differences
in
movement
could
distinguish
some
nouns
from
related
verbs.
Whereas Stokoe
et
al.
(1965)
reports the
existence
of
a
single sign
meaning
SLS
64

Phonological
Base
both
'sit'
and
'chair',
Supalla and
Newport
claim
that
SIT
and
CHAIR
are
separate
signs.
They
find
that
for
more
than
100
such
noun-verb
pairs, the
pattern
of
movement
of
the
noun
differs
in
predictable
ways
from
that
of
the
verb. They distinguish
these formational
differences
in
terms of
three
'manners
of
movement'.
They
describe
the
movement
of
the
sign
SIT
as
a
single,
unidirectional
movement
with
a
'hold
manner'
and
that
of
CHAIR
as
a
repeated,
unidirectional
movement
with
'restrained manner'
Figure
3a.
WHEN
Figure
3b.
ALSO
Their
manners
of
movement
demonstrate
a
significant
type
of
sequentiality
in
the
formation
of
signs.
The
sign
SIT
begins with
a
motion
toward
the
base
hand
and ends
with
the
two
hands
in
contact, but
not
moving.
A sign
ending
with
the
hands
immobile
is
said
to
have
'hold
manner'
at
the
end
of
the
sign. In
their
view
such
motionless
periods
are
as
important
in
providing
contrast
as
are
the
periods
of
movement.
They
note
(1978:96)
that
one
of
the
possible
implications
of
their
findings
is
that
signs
SLS
64

Liddell
&
Johnson
4a.
SIT
4b.
CHAIR
Figure
4.
The
movement
differences
between
SIT
and
CHAIR
may
have
sequential
internal
segments
rather
than
a
simultaneous
bundle
of
features.
1
1.5
Local
Movements
Local
movements
are
small
repeated
movements
of
the
fingers and
wrist
which
accompany
the
major
movements
of
the
hand.
For
example,
1 It
might
be
possible
to
argue
that
in
many cases,
the hold at the
end
of
a
sign
is
simply
the
physiological
result
of
making
contact
with
the
body.
It is
not
difficult
to
demonstrate
that
this
is
not
so.
The
sign
KNOW
moves
toward
the forehead,
makes
contact,
then stops briefly
in
contact
with
the
fore-
head.
It
can
be
described
as
ending
with
hold
manner.
Liddell
(1
984a)
reports the
occurrence
of
a
noncontacting
form
of
the sign
in
which
the
hand
approaches
but
does
not
touch
the
forehead
and
in
which the
sign
still
ends
with
hold
manner.
SLS
64

203
Phonological
Base
SLS
64
LONG-AGO
is
produced
with
a
'5'
hand
configuration
which
moves
back-
ward
to
a
hold
at
a
point
just
over
the
shoulder. During
the
backward
movement
itself
the
fingers
wiggle,
but
the
final
hold
is
produced
without
finger
wiggling.
Figure
5.
LONG-AGO
Thus
LONG-AGO
contains
the
sequence:
local
movement,
no
local
move-
ment.
In
other
signs, such
as
JUMP-FOR-JOY
the
wiggling
is
restricted
to
the
middle
of
the
sign
where
the
active
hand
does not
contact
the
base
hand.
This
produces
the
sequence:
no
local
movement,
local
movement,
no
local
movement.
1.6
Nonmanual
Signals.
Many
nonmanual
signals
involve
no
sequentiality.
For
example, the
combination
of
raised
eyebrows and backward
head
tilt
which
accompanies
topics (Liddell,
1977)
is
purely
configurational,
with no
internal
changes.
Some
nonmanual
signals, however,
are
produced
by
sequencing
nonmanual
activities. Some
such
nonmanual
signals
occur
as
part
of
lexical
items
and

204
Liddell
&
Johnson
SLS
64
others
occur
as
part
of morphological processes.
A
lexical
item
which
requires
a
sequence
of
nonmanual
activities
is
GIVE-IN.
During
the
initial
part
of
its
production
the
lips
are
closed
but
during
the
final
part
of
its
production
the
lips
are
open.
ALL-GONE,
on
the
other
hand,
begins
with
the
lips
apart
and
the
tongue
slightly
protruding
and
ends
with
the
lips
closed.
Sequences of
nonmanual
activities
are
also
important
as
part
of
morphological
processes.
Liddell
(1984b)
describes
a
sequence
of
nonmanual
activities
required
as
part
of
the
inflection
for
unrealized-
inceptive
aspect.
When
this
inflection
is
applied
to
a
verb,
specific,
predict-
able
changes occur
in
both
the
manual
and
nonmanual
portions
of
the
sign.
The
sequence
of
nonmanual
behaviors
associated
with
this
inflection
require
the
signer
to
inhale
through
the
mouth
while
rotating the
trunk,
and
then to
hold
the
resulting configuration during
the
final
portion
of
the
sign.
1.7
Contrast in
ASL.
We
have
illustrated
several
types
of
sequentiality
in
ASL
signing,
including
sequences
of
handshapes,
locations,
nonmanual
signals,
local
movements, and
movements
and
holds.
The
simultaneous
model
of
sign
structure
is
not
able to
represent
these
sequential
details
in
an
effective
way.
This
alone
argues
for
a
descriptive
device
which
is
able
to
represent
important
aspects
of
ASL
sequence.
Specifically,
given
that
signs
have
sequential
structure,
that
sequence
can
be
shown
to
correspond
to
phonological
segments
responsible
for
sequential contrast
of
the
sort
found
in
spoken
languages.
The
identifica-
tion
of
physical
sequence
in
the
linguistic
signal
provides
the
evidence
needed
to
argue
that
signs
are composed
of
sequenced,
abstract,
linguistic
segments.
Support
for
the
existence
of
such
linguistic
segments
comes,
in
part,
from
a
demonstration
that
ASL,
like
spoken
languages,
contains
pairs
of
signs
distinguished
only
by
differences
in
sign-internal
sequence.
It
has
become
traditional
in
treatments
of
ASL
structure
to
illustrate
minimal pairs' of
signs
as
a
demonstration
of phonological
contrast.

Phonological
Base
However,
because
the
simultaneous
model
of
sign
structure
dictates
that
signs
are composed
of
a
single,
simultaneous
bundle
of
gestural
features,
such
pairs
of
signs
are
able
to
demonstrate
only
simultaneous contrast
of
the
sort
found
within segments
in
spoken
languages.
Thus,
staying
for the
moment
with
the
notion
that
signs
are
simultaneous,
most
'minimal pairs'
of
signs
identified
in
the
literature
on
ASL
exhibit
contrasts analogous
to the
differences between
[p], [t],
and
[b].
They
are distinctions
of
one
feature
within
a
single,
co-occurrent bundle
of
features.
By
contrast,
in
spoken
language
analysis,
the notion of
'minimal
pair'
has typically
been
used
to
demonstrate
sequential
contrast.
Thus,
a
minimal
pair
is
usually
considered
to
be
two
words,
contrastive
in
meaning,
which
are identical
in
all
segments
except
one,
in
which
they differ
by
only
one
feature.
The
kinds
of
ASL
sequential
details
we
have
identified
above
provide
this
kind
of
evidence for
sequential
contrast.
THANK-YOU
and
BULLSHIT
are
minimal
pairs
in
this
sense.
Both
begin
with
identical
holds
produced
at
the
chin
and
move
to
holds
produced
at
a
location
about
six
inches
out
and
slightly
below
the
chin.
In
both
signs,
the
orientation
of
the
hand remains
constant,
with
the
palm
toward
the
face
and
the
wrist
toward
the
ground.
Thus,
from
the
perspective
of
movement,
location,
and
orientation
the
signs
have
identical
sequences.
They
differ
only
in
hand configuration
sequence.
Specifically,
during
the production
of
the
sign
THANK-YOU,
the
hand
configuration
begins
and
ends
as
a
'B'.
In
the
sign
BULLSHIT, however,
it
begins
as
a 'B'
and
ends
as
an
'S'.
In
Table
4
the
parts
of
THANK-YOU
and
BULLSHIT
are
aligned.
SLS
64

Liddell
&
Johnson
THANK-YOU
first
par
middle
part
last
art
movement hold move out
hold
location
chin
transitional
out
from
chin
orientation
palm
to
chin
transitional
palm
to
chin
hand
configuration
B
transitional
B
BULLSHIT
first
part
middle ipat
last
part
movement
hold
move out
hold
location
chin transitional
out
from
chin
orientation
palm
to
chin transitional
palm
to
chin
hand
configuration
B
transitional
S
Table
4.
Sequential
contrast
between minimal
pairs
THANK-YOU
and
BULLSHIT.
Although
true
minimal
pairs
such
as
these
are
not
abundant
in
ASL,
there
are
similar
pairs
that
demonstrate
sequential contrast
in
each
of
the
major
descriptive
parameters
of
signs.
Together
they
demonstrate
that
segments
function
to
signal
contrast
in
ASL
in
much
the
same
manner
as
in
spoken
languages,
and
suggest
that
the
description
of
segments
is
central
to
an
adequate
phonological
analysis
of
ASL signs.
In
addition,
the
value
of
a
segmental
description
in
the
analysis
of
the
phonological
and
morphological
processes
of
ASL
will
become
more
apparent
as we
proceed.
2. A
Phonetic
Transcription
System
for
ASL
At
first
glance,
it
may
seem
inappropriate
to
use
'phonology',
'phonetics',
and
other
seemingly
vocally-based
terms
in
referring to details
of
sign
language and
its
organization.
As
we
mentioned
earlier,
Stokoe's
work
explicitly
avoids
this
difficulty
by
positing
terminology
such
as
'chereme'
and
'cherology',
which
are
specific
to
sign
language.
Battison
(1974)
demonstrates
that
sign
language descriptions
contain
a
sub-lexical
level
of
analysis
that
appears
in
certain
ways
to
be
organizationally
and
SLS
64

Phonological
Base
functionally equivalent
to
the
level
of
phonology
in
spoken
languages.
He
argues
convincingly
that
standard
phonological
terminology
refers
appropriately
to
those
levels.
A
part
of what
we
will
be
demonstrating
in
this
paper
is
that
an
analysis
of
the
patterns
of
organization
of
sign
language
signals yields
levels
of
analysis
quite
similar
to
those
known
to
exist
in
spoken
language phonologies.
It
is
a
matter
of
historical accident
that,
during
the period
of develop-
ment
of
modern
linguistic
terminology,
all
languages
known
to
linguists
were
spoken
languages.
Even
so,
for
the
most
part,
phonological
terminology refers
to
the patterns
of
organization
of
linguistic
signals,
rather
than
to
the
signals
themselves. Thus,
the
vocal
reference
of
the
phone-
stem
in
words
such
as
phoneme
is
largely
unnecessary.
We
use
phonological
terminology
in
referring
to
the organization
of
sign
languages,
with
the
understanding
that
the
terminology here,
as
in
studies
of
other
languages, refers
to
general
principles of
organization
probably
found
in
all
languages
rather
than
to
the
specific
vocal
gestures
of
spoken
languages.
2.1
Goals
of
Transcription.
A
transcription
system
for
a
language or set of
languages
should
meet
the
dual
goals
of
at once providing
for
the
accurate
representation
of
the
detail
of
the
'facts'
of
a
language
and assuring
that
those
representations
are
useful
in
characterizing the
organization
of
the
facts.
We have
attempted
to
devise
a
system
that
provides
a
linkage
between
the
abstract
and
concrete
aspects
of
phonological
systems
without
committing
overwhelmingly
to
either.
Clearly,
an
adequate
system
of
tran-
scription
must
have
elements
of
both.
On
the concrete
end,
a
transcription
must account
for
all
the
linguistically
interesting
details of
the production
of
the
signal.
For
our
purposes,
such
phonetic transcription
will
be roughly
equivalent
in
its
concreteness
to
the
'systematic
phonetic
representations'
of
standard
generative
phonology
(Chomsky and
Halle
1968).
While
such
representations
must
account
for
a
great
deal
of
detail,
they exclude
(a)
linguistically
non-distinctive
differences
such
as
the
difference between
SILS
64

Liddell
&
Johnson
apical
and
laminal
[s]
in
English;
(b)
sequential detail
within
phonologically
functional units,
e.g.,
elimination
of
onset
and closure
information
from
the
description
of
English
stop
consonants;
(c)
detail
stemming from
universal
physiological
conditions;
(d)
detail stemming from
individual physiological
conditions,
e.g.,
absolute
voice
pitch;
and
(e)
traditionally
non-linguistic
detail
such
as
rate,
loudness, and
affect
marking
features.
On
the
abstract
end,
an
adequate notation
system
must
provide
descriptive
devices
that
permit
a
plausible
linkage
between
the
detailed
surface
representation
and
the
underlying
forms
of
the
individual
lexical
items
that
are
present
in
it.
Thus,
a
single
set of descriptive
devices
should
at
once
be
capable
of
characterizing
each
of
the
following:
(a)
the
phonological
shape
(underlying
form)
of
lexical
items;
(b)
the
phonological
aspects
of
the
morphology;
(c)
phonological
processes;
and
(d)
the
surface
forms
of
signs
in
running
signed
productions
(at
the
level
of
concreteness
specified
above). To
the
extent
that a
system
of
notation
succeeds
in
achieving
this
balance,
it
provides
phonetic
motivation for phonological
features
and
phonetic
plausibility
for
the
abstract structures
and processes
of
the
phonological
component.
That the
system
be
usable
is
a
second,
more pragmatic
goal
which
has
influenced
the
current
form of
our notation
system.
Thus,
while
sign
nota-
tions
will
ultimately
be
reducible
to
matrices
of
binary
phonological
features,
most
of
the
notations
presented
here
contain taxonomic
entries
that represent
bundles
of
such
features.
The
use
of
such
taxonomic
entries
is
primarily
a
matter
of
clerical
and typographical convenience,
reducing
the
number
of
symbols
required
to transcribe
a
sign.
2.2
Overview
of
Sign Structure.
2.2.1
Describing Segments.
The segment
is
the
central
element
in
our
view
of
the
structure
of
signs.
Thus,
the
representation
of segments
is
the
essential task
of
our
nota-
tion
system.
In
our
system,
each
segment
is
represented
individually and
SLS
64

Phonological
Base
signs
(and
discourse
strings) are
represented
as
strings
of
segments.
2
Segments
in
sign
languages
are
composed of
two
major
components.
One
describes
the
posture
of
the
hand; the
other
describes
its
activity.
A
description
of
the
posture
of the
hand
is
concerned
with
where
it
is,
how
it
is oriented,
how
its
own
movable
parts
are configured,
and
so
on. The
features
that
describe these details
are
collectively
called
articulatory
features.
We
refer
to
the
combination
of
articulatory
features needed
to
specify
a
given
posture
of
the
hand
as
an
articulatoryy
bundle'.
The articulatory
bundle contains
four major
clusters
of
features.
The
first
represents
the
hand
configuration,
i.e.,
the
state
of the
fingers and
thumb.
The second
cluster
represents
point
of
contact,
which
specifies
the
primary
location
with
respect
to
which
the
hand
is
located,
the
part
of
the
hand
that
points
to
or
contacts
that
location,
and
the
spatial
relationship
between
that
hand
part
and
that
location. The
third cluster
represents
facing,
which
is
composed
of
sets
of
features
specifying
a
second
location,
and
features
indicating
the part
of
the
hand
which
faces
that
location.
The
fourth
cluster
of features
in
the
articulatory
bundle,
orientation,
contains
features
specifying
a
plane
toward
which
a
part
of
the
hand
faces.
Orienta-
tion
features
distinguish
THING
(a
sequence
of movements
made
with
the
palm
up)
from
CHILDREN
(like
THING
but
with
the
palm down).
The
four
clusters,
all
taken
together,
describe
the posture
of
the
hand
at
a
parti-
cular
point
in
the
production of
a
sign.
They
do
not
describe
the
activity
of
the
hand.
The features
that
specify
the
activity
of the hand
during
the produc-
tion
of
the segment
are
grouped
into
a
separate
segmental
feature
bundle.
They
describe
whether
or
not
the
hand
is
moving.
and,
if
so,
in
what
2 In
actuality, discourse
strings
must
be
represented
as
several
simultaneous strings: one
for
each
hand,
since
each
produces
segments,
and
one
for
each
linguistically
independent
complex
of
torso.
head,
and facial behaviors.
For
the moment we
are
focusing
on
segments
and
strings
of
segments
produced by
a
single
hand.
SLS 64

Liddell
&
Johnson
manner.
The
elemental
work
of
this
class
of
features
is
to
distinguish
move-
ments from
holds.
Movements are
defined
as
periods
of
time during
which
some aspect
of
the
articulation
is
in
transition.
Holds
are
defined
as
periods
of
time
during
which
all
aspects
of
the
articulation
bundle
are
in
a
steady
state.
While
the
descriptive
work
of
the
segmental
features
is
to
detail
the
movement
of
the
articulators,
they
function
within
signed
strings
to
divide
the
flow
of
gestures into
segments.
By
definition,
then,
the features
that
distinguish
movements from
holds
also
define
the
segmental
structure
of
larger units
such
as
signs,
which
we
represent
as
strings
of juxtaposed
segments.
This
is
not
unlike
the manner
in
which
the
major
class
features
of
generative phonology
function.
In
spoken language phonology,
major
class
features
specify
phonetic
details
of
segments
such
as
spontaneous
voicing,
interruption
of
the
airstream,
and
syllabicity.
These
same
feature
values
distinguish
consonants
from
vowels
and
therefore
also
function
to
specify
the manner
in
which
the
flow
of speech
is
divided.
The
remaining features
in
the
segmental
bundle
specify
the
finer
detail
of
segments
such
as
contour
of
movement,
simultaneous
local
move-
ment
of
the
fingers,
and
precise
timing
information
such
as
length.
We
will
discuss
these
features
in
detail
below.
We
have
presented the
articulatory
bundles
and
segmental
bundles
separately,
and
in
fact
they function
independently
from
each
other
in
the
specification
of
entire
segments.
The
articulatory
features
combine
to
describe
postural
states.
By
definition,
movement
segments
are
those
during
which
there is
a
change
in
state
in
some
complex
of
articulatory
features,
and
hold
segments are
those
during
which
no
such
change
occurs.
Because they
involve
a
steady
state,
a
single
matrix
of
features
will
be
suffi-
cient
to describe
holds.
This matrix
will
contain
both
the
segmental
bundle
of
features
including
the
specification of fine
detail
of
the
segment
and
the
articulatory
bundle
of
features
describing
the
postural
state present
during
the production
of
the
hold
segment.
Movement
segments,
however,
present
another
problem.
During
a
movement
the
hand
changes
from
one
posture
to
another.
Thus,
because
SLS
64
210

Phonological
Base
our
articulatory
features represent
states,
our
system
requires
the
specifica-
tion
of
an
initial and
final
bundle
of
articulatory
features
to indicate
the
changes
during
the
production
of
the
segment.
Movement
segments
contain
one bundle
of
segmental
features
containing
the
specification
of
the
segment
type
and
the
fine
details
of
the
movement
and
two
bundles
of
articulatory
features,
the
first
of
which
specifies
the
postural
state
at
the
inception
of
the
movement
and
the
second
of
which
specifies
the
postural
state
of
the
hand
at
the
conclusion
of
the
movement. Hold segments
contain
one
articulatory
bundle; movement
segments contain
two
articulatory
bundles.
Both
hold
and
movement
segments
may
be
represented
by
matrices
of
features,
but
following
the
discussion
above
the
matrices
will
be
different.
The
hold
segment
would
correspond
to
be
a
straightforward
and
traditional
feature
matrix
as
in
Fig.
6,
while
the movement segment
will
have
one
set
of
segmental
specifications
and
two
sets
of articulatory
specifications,
as
in
Fig.
7.
I
segmental
| [
segmental
features
features
I I I I I
I
artic.
| | initial|
final
|
features
I I artic.
I
artic.
| |
Ifeatures|featuresl
|I I I
I_ I I
Figure
6.
A
hold matrix.
Figure
7. A
movement matrix.
An
apparent
alternative
solution
to
the
use
of
two
kinds
of
matrices
might be
to
use
only
hold matrices,
let
them define
segmental
structure,
and
have
movement take
place
as
a
result
of
transitions
from
one
state
to the
next.
As
we
present
more
detailed
descriptions
of
signs
it
will
become
apparent
that
independent
movement
features
are
necessary.
This
is
because
the
fine
details
of
movement
production
are
features
of
the
move-
ment
itself,
not
of
either
of
the
individual
articulatory
bundles.
For
example,
when
the
hand
moves
on
a
path
from
one location
to
another,
SLS 64

212
Liddell
&
Johnson
SLS 64
that
path
may
take
any
of
three
contours.
It
may
move
in
a
straight
line,
on
an
arc,
or
on
an
indirect
path
with
a
sharp
change
of
direction
in
the
middle.
These
differences
in
path
are
contrastive
and
therefore
must
be
recorded.
They
are not
a
feature
of
the
initial
articulatory
posture
nor
of
the
final
articulatory
posture
nor
of
both
at
once.
They
are
a
feature
of
the
period
of
time
during
which
the
hand
is
changing
from
the
initial
posture
to
the
final
posture.
Thus,
they
are details
of
the
movement
itself
and must
be
specified
independently
of
the
articulatory
information.
Considerations
presented
below
will
confirm
this
claim
of
independence
of
the
segmental
and
articulatory bundles
of
features.
2.2.2
Non-Manual
Behaviors.
The
segmental
structure
of
signs also
bears
on
the
representation
of
the
non-manual
behaviors
that
have
linguistic
function
in
ASL.
At
times,
non-manual
behaviors
clearly
have
functions
that
are
independent
of
the
segment.
Examples of
such
non-manual
behaviors
are
those
that
have
syntactic
function
and
those
that
have
clear
morphological
status.
Others
appear
to
be
tied
to
specific
segments
within
specific
signs
(Liddell
1984a).
In
both
cases,
although
possibly
independent
in
function,
the
behaviors
are
timed
to
the
production
of segments,
and
need
to
be
specified
in
the
tran-
scription
system.
The
exact
nature
of
this
specification
will
be
taken
up
later.
2.2.3
Describing
Sequences of
Segments.
In
the
view
of
sign
structure
presented
here,
individual
signs and
larger
constructions
are
all
composed
of
sequences
of
segments. Thus,
a
sign
or
a
piece
of
discourse
may
be
represented
as
a
sequence
of
hold
and
movement
matrices,
each
composed
of
the
appropriate
number
of
segmental
and
articulatory
bundles.
The
sign
GOOD,
for
example,
is
composed
of
three
segments:
a
hold,
a
movement,
and
a
hold
(see
Fig.
26b).
The
first
hold
occurs
with
the
finger pads
of
a
flat
hand
in
contact
with
the
chin.
For
convenience,
we
will
call
this complex
of
articulatory
information

213
Phonological
Base
SLS
64
posture
a'.
From
this hold,
the
hand
moves
outward and
downward
to
a
final
hold,
which
occurs
in
space
about
a
foot
in
front
of
the
sternum
with
the
same flat
hand
configuration
oriented
so
that
the
palm of
the
hand
is
facing
(roughly)
upward
and
the
tips
of
the
fingers
are
pointing
outward
at
about
a
forty-five
degree
angle. We
can
call
this
complex
of
articulatory
information 'posture
b'.
In
our
matrix
format
we
can
represent
this
sign
as
in
Fig.
8.
I I 1 1I
I
Hold
Movement
Hold
|
I I
I
I
|Posture
Posture Posture]Posture
a a b I b I
II I I
Figure
8.
Representation
of
feature
matrix
for
sign
GOOD.
Notice
that
in
the
representation
of
GOOD
the
initial
articulatory
specification
of
the
movement
segment
is
the
same
as
the
articulatory
speci-
fication
of
the
first hold
segment.
Similarly,
the
final
articulatory
specifica-
tion
of
the
movement
segment
is
the
same
as
the
articulatory
specification
of
the
second
hold
segment.
An
initial
posture
of
any
segment
in
a
string
is
identical
to
the
final
posture
of
the
preceding
segment.
This
is
true
by
definition
because
a
given
line
of
transcription
represents
a
sequence
of
behaviors
of
a
single
articulator,
which
can only
start
a
gesture
from
the
posture
in
which
it
terminated
the
preceding
gesture. From
this
perspective
it
is
unnecessary to
record
every
articulatory
bundle
of
information
because
(within
signs) two
articulatory
bundles
that
share
a
common
segmental
boundary
must
be identical.
This
observation stands
as
additional
evidence
for
the
independence
of
the
articulatory
features
from
the
segmental
features.
It
also
recom-
mends
the
use
of
an
autosegmental
representation
which
permits
the
attachment
of
single
clusters of
features
of
one sort
to
single
clusters
of
features
of
another
sort
(Goldsmith
1976,
McCarthy
1979), as
in
Fig.
9.

Liddell
&
Johnson
Hold
Move-
|
Hold
|
ment
I I I
I II I
I I I I
IPosturel IPosturel
a 1b
Figure
9.
Representation
of
autosegmental attachment
of
feature
bundles
of
the sign
GOOD.
Autosegmental
representations
of
the
sort
presented
in
Fig.
9, in
addition
to enhancing
clerical
economy,
provide
additional
support
for the
earlier
suggestion
that
the
articulatory
bundle
of
features
is
autonomous
in
function
from
the
segmental
bundle of features.
It
is
also
quite
likely
that
certain
of
the
clusters
of
features
within
the
articulatory
bundle
itself enjoy
a
similar
kind
of
autonomy,
particularly
at
the
lower
levels
of
the
phonology
where
the
independent
postural
and
movement
components
must
be
finely
timed
to
one
another.
Similarly,
there
may
be
more
autonomous
tiers
of
feature
clusters at
the
level
of
the
phonology
that
controls
the
production
of
fast
speech,
in
which
muscular
activities
and
postures
are
reinterpreted
and
produced
as
perceptually
and
productively
similar
(though
linguistically
different)
muscular
behaviors.
Autosegmental
analyses
of these
phenomena
may
prove
to
be
worthwhile.
For
our
purposes,
however,
it
is
sufficient
to
use
only
the
articulatory
and
segmental tiers,
together
with
a
tier
for
non-manual
behaviors.
A
number
of
the
combinations
of segments
that
may
occur
in ASL
signs
are
presented
in
Fig.
10.
SLS
64

Phonological
Base
I
a.
COLOR
b.
SIT
II |
1a
|
|a
| |
c.
OUR
l III IMI
cil
1I 1
d.
WHEN
181 II II
Ii
lIl
1W1
Figure
10
Signs
Illustrating
common
segment
combinations
SLS
64

Liddell
&
Johnson
e.
CONGRESS
NI
Wi
MI
N
i -1
CHI
Wj 1h i
.L, Lz I1
121
Iki
Ws IlyIs 12
M
s 1A
g.
JUMP
HI MI M
12;
lal
121
Figure
10
(cont)
Signs Illustrating
common
segment
combinations
SLS
64
216

Phonological
Base
2.2.4
Describing
Signs Requiring
Two
Hands
As
we
indicated
above,
many
signs
make
use
of
both
hands
as
articulators. From
a
phonetic
perspective,
each
hand
is
independent
of
the
other.
Moreover, the
hands
may
carry
different
phonetic information
at
a
given
moment. For
example, one
may
be
moving
while
the
other
is
not.
One
may
be
in
one location
or
orientation
or
hand configuration
while
the
other
hand
is
specified differently
for one
or more
of
these details.
As
one
might expect,
there
appear
to
be
fairly
strong conditions
on
the
nature
and
extent
of
the
simultaneous
articulation
of
two
segments
(Battison
1974,
1978),
so
the
two
hands
are not completely
independent
phonologically.
While
a
notation
system
may
ultimately
be
able
to
eliminate
certain
aspects
of
the
information
that
is
predictable
from
such
constraints
on
simultaneous
articulations,
it
is
useful
at
early
stages
of
analysis
to
be
able
to
represent
each
hand
in
its
full
phonetic
configuration.
From
the
perspective
of
the
segmental
notation
system
described
above,
there
is
no
difference
between the
productions
of one hand
and
those
of
the
other.
Given
this
and
their phonetic independence,
each
hand
must
be
represented
as
a
separate
string
of segmental
notations,
and
the
segments
of one
hand
must
be
attached
(for
timing
purposes)
to
the
co-
occurrent
segments
of
the
other
hand.
The
first
difficulty
encountered
in
the
representation
of
the
behaviors
of
both
hands
is
that
right
and
left
are
not
absolute
in
signing.
First,
left-
handed
and
right-handed
signers
sign
mirror
images
01the
same
sign
sequence
with
no
change
in
meaning.
A
notation
system
should
describe
both
the
left-handed,
left-dominant
and
the
right-handed, right-dominant
versions
identically.
Secondly,
certain
constructions
treat
spatial
locations
on
the
right
and
the
left
as
absolute.
A
notation
system
must
be
able
to
distinguish
right
from
left
under
these
conditions. Third, certain construc-
tions
allow
a
signer
to
meaningfully
alternate
between
right-dominant
and
left-dominant
signing.
The
notation
system
must
be
able
to describe
this
sort
of
alternation.
SLS 64
217

Liddell
&
Johnson
Strong
Hand
_
MI IHI
Weak Hand
IHI
Figure
lia.
FIRED, a
two-handed
sign
Figure
11b.
FIRED
in
which
the
strong
hand
moves
with
respect
to
the weak
hand
Padden
and
Perlmutter
(1984)
introduce
the terms 'strong'
and
'weak'
to
describe
the
active
hand
and
the
hand
it
acts
upon.
Adopting
those
terms
for
our
notation
system
will
permit
signs
to
be
specified
in
a
single
way
although
signed
in
mirror
image
by
right-
and
left-handed
signers.
We
have
chosen
to
use
two
vertically
stacked
strings
of segments
for
two
handed
signing.
The
top
line
represents
the strong
hand
and
the
bottom
line
represents
the
weak
hand.
In such
cases,
the
strong hand
is
understood
to
be
the
dominant
hand
of
the
signer.
Particular transcriptions of running
sign
will
need
to
be
marked
for
the
dominance
of
the
signer.
When
a
signer
shifts
from
expected-dominance
signing
to
opposite-dominance
signing
the
strong
label
will
be
shifted
to
the
bottom
line
and
the
weak
to
the
top
line.
In
those
instances
when
each
hand
is
actually
operating
independently,
the
top
line
will
be
right
for right-dominant
signers
or
left
for
left-dominant
signers.
SLS 64

219
Phonological
Base
SLS
64
It
appears
that
the
strong
hand
segments
function
as
the
central
organizing
elements
for
the
timing
of
strings
of
co-occurrent
segments.
Therefore,
the
segments
of
the
weak
hand
must
be
attached
to those
of
the
strong
hand.
Several
combinations
of
strong
and
weak
hands
within
signs
and
our
conventions
for
attaching them
are
presented
in
Figs.
11-13.
Strong
Hand
Weak Hand
1I
11
I'I
l
Jb|
1-1 1b
1
lal
JI
I L
Figure
12a.
LARGE,
a
two-handed
sign
in
which
the
two
hands
move
independently,
simultaneously,
and
symmetrically
Figure
12b.
LARGE
2.3
Detailed
Description of
Segmental
Bundles.
Segmental
feature
bundles
specify
the detail of movements
and
holds.
Each
such
cluster
defines
one
segment
in
the
string
of
gestures
in
the
tran-
scription of
a
running
signed
production.
Ultimately,
the segmental
bundle
will
contain
numerous
binary
features.
At
present
it
contains
five
slots
for
the
entry
of
taxonomic
symbols
representing
clusters
of
features.
The
five
types
of
entries
within
the
segmental bundle
are
laid
out
as
shown
in
Fig.
14.

Liddell
&
Johnson
Strong Hand ,
lal lbi lal
L
I L.I II
Weak
Hand
lbl
Il
a
bl
Figure
13a.
MAYBE,
a
two-handed
sign
Figure
13b.
MAYBE
in
which
the
strong
and
weak
hands perform
independent
movements
but
in
temporal alternation
major
class
j I
contour
j
plane
|
quality
j
Local
movement
j
Figure
14.
Organization
of
segmental
features
2.3.1
Major
Classes
of
Segments.
There are
two
major
classes
of
segments
in
ASL:
holds
and
move-
ments.
As
described
above,
a
movement
(M)
segment
is
characterized
by
a
change
in
one or
more
of
its
articulatory
features
and hold
(H)
segments
are
not.
Notice
that
not
all
movement
segments
involve
movement
from
one
location
to
another.
The
change
in
articulatory specification
may
occur
in
the
hand
configuration
(UNDERSTAND),
the
orientation
(START),
or
other
clusters
of
the
specification.
Such
non-path
movements do not
SLS
64

Phonological
Base
appear
to
have
a
phonological
status different
from
that
of
path
movements
(those
in
which
there
is
a
change
in
the point
of
contact
specification)
and
so
need
not be
distinguished
by
a
special
feature.
2.3.2
Contours
of
Movement.
Those
movement
segments
that
move on
a
path
between
two
loca-
tions
may
do
so
on
one
of
several
contours.
Straight
[str]
movements
traverse
a
direct, straight
path
between
two
points
(GOOD).
There
are
two
types
of indirect
contour
paths:
round
[md]
and
seven
[7].
The
seven
contour
describes
an
indirect path
that
is
sharply
angled
(CHICAGO).
The
round contour
describes
an
indirect
path
that
is
smooth.
Arcs
(OUR)
and
circles
(FACE)
both
describe
round
paths
but
are
distinguished
by
the
fact
that
an
arc begins
at
one
location
and
ends
at
another
whereas
a
circle
begins
at
a
point,
traverses
a
round path,
and
ends
at
its
beginning
point.
2.3.3
Contour
Planes.
When
a
path
movement
is
not
straight,
it
is
necessary
to
specify
an
additional
piece
of
information,
which
functions to
orient
the path.
The
entries
indicate
the
plane
upon
which
the
hand
travels
as
it
moves
between
points.
We
currently record
five
planes. The
horizontal
plane
[HP]
is
the
plane parallel
to
the
floor
(OUR).
The
vertical
plane
[VP]
is
that
plane
parallel
to
the front
of
the
torso (RAINBOW).
The surface
plane
[SP]
is
the
plane
parallel
to
the
surface
at
a
location
on
the
body
or hand
(FACE).
The
midline
plane
[MP]
is
a
plane
that
intersects
the
surface
plane
along
the
midsaggital
line
of
the
body (BLOUSE,
SIGN),
or
the plane
through
the
long
midline
of
the
bones
of the
arm or
the
hand (BASKET).
We
currently
use
the
designation
oblique
plane
[OP]
to
represent
the
plane
that
is
horizontal
from
side
to
side
but
angled
up
and
away
from
the
body.
2.3.4
Quality
Features.
Quality
features
describe
fine
details of
a
segment.
Among
these
are
SLS 64

Liddell
&
Johnson
the
temporal
qualities
prolonged
[long],
shortened
[short],
and
accelerating
[acc],
and
the
non-temporal
qualities
tense
[tns],
reduced
path
[sm],
and
enlarged
path
[1g].
The
quality
feature
contacting
[contact]
indicates
that
the
hand
makes
contact
with
the
other
hand
or
a
body
location
during
the
course
of
the
movement.
It
describes
brushing movements,
in
which
the
hand
travels
between points
on
two
sides
of
a
location, making
brief
contact
with
that
location
as
it
passes.
It
is
also
useful
in
describing
the
movement
in
which
the
hand
moves
to
a
location,
makes
brief
contact
,and
rebounds
to
a
point near
that
location.
2.3.5
Local
Movements.
The
major
classes
of
segments (H
and
M)
reflect
activity
of
the
hand
taken
as
a
whole. It
is
common
for
signs
simultaneously
to exhibit
move-
ment
at
the
finger,
wrist
or
elbow
joints.
Such
movements
are
overlaid
on
the
actual
segmental
activity,
occurring
together
sometimes
with
H
segments and
sometimes
with
M
segments. Thus,
they
are
secondary,
though
linguistically
significant
activities.
Each
of
the
local
movements
is
characterized
by
rapid,
uncountable
repetition.
All
may
occur
in
H
segments.
At
least
wiggling,
twisting,
nodding,
and hooking
may
occur
in M
segments.
3
Wiggling
[wg]
represents
repeated,
sequentially
alternating retraction
at
the
first
joint
of
all
fingers
extended
at
the
first
joint
(COLOR).
Hooking
[hk]
involves
repeated,
simultaneous retraction
at
the
second
and
third
joints
of
all
fingers
that
are
extended
at
the
first
joint
and
retracted
at
the
second
and
third
joints
('hooked'
hand configurations)
(WORM).
Flat-
3
Earlier
work
treated these
as
features
of
hand
configuration
(Liddell,
1984a).
There
is
evidence
for
their independence
from
hand
configuration,
however,
in
the
fact
that
certain
of
the
local movements
function
as
the
sole
manual
markers
of
inflectional morphemes attached
to
signs
which
have
plain
(i.e.,
non-moving)
hand
configurations
in
their
uninflected
forms.
SLS
64

Phonological
Base
tening
[fl]
is
repeated,
simultaneous
retraction
at
the
first
joint
of
all
fingers
that
are
extended
at
the
second
and
third
joints
and
retracted
at
the
first
joint
('flat'
hand
configurations)
(STICKY).
4
Twisting
[tw]
describes
repeated,
alternating rotations
of
the
wrist
(WHERE).
Nodding
[nod]
is a
repeated
retraction
and
extension of
the
wrist
joint
(YES).
5
Releasing
[rel]
involves
rapid,
repeated
opening
of
fingers
that
have
thumb restraint
(SHIRK-RESPONSIBILITY).
Rubbing
[rub]
is
repeated,
back
and forth
rubbing
of
the
thumb
and
the
finger
pads
(DIRT).
Circling
is
a
repeated,
uncountable
local
circling
about
a
central
point
simultaneously
with
either
a
H or
M.
It
requires
the
specification
of
a
plane.
2.4
Detailed
Description of
Articulatory
Bundles.
Each
articulatory
bundle
is
composed
of
eight
entries,
each
representing
a
complex
of
features.
The entries
cluster
into
four
possibly
autonomous
groupings, described
above
as
hand
configuration
(He),
point
of
contact
(POC), facing
(FA),
and
orientation
(OR).
They
are organized
as
shown
in
Figure
15.
2.4.1
Hand
Configuration.
We
have
found
more than
150
HCs
in
ASL
lexical
signs.
Many
more
occur
in
the
surface
forms
of running
sign.
A
system
of
thirteen
mostly
binary
features
will
distinguish
all
HCs
we
know
to
exist
in
sign
languages.
The
taxonomic
symbols
we
use
as
HC
entries
in
our notations
are
capable
4
It
may
be
that
a
single
feature
such
as
"contracting"
unifies
both
hooking
and
flattening.
5
For
certain
hand
configurations
and
under certain
discourse conditions
it is
possible to achieve
twisting
and
nodding with
the
elbow
joint
rather
than the
wrist
joint.
For
example, the sign
WHERE
is
typically
performed
by
twisting
the
wrist
but
by
changing
the
hand
configuration
to
one
with
a
straight,
rigid wrist
the
twisting
can
be
transferred
to
the
elbow. Similarly,
YES
which
normally
nods
at
the wrist
may nod
at
the elbow
in
its
emphatic
form.
SILS
64

Liddell
&
Johnson
of
describing
all
the
HCs
of
ASL
and
many
more.
They
translate
to
features
in
a
very
straightforward
way.
HC:
POC: Part of Hand
Proximity
Spatial
Relation
Location
FA:
Part
of Hand
Location
OR:
Part
of Hand
Plane
Figure
15.
Organization
of
articulatory
bundle
The
HC
entry
is
organized
according
to
the
following
schema.
(/)[Hndshp]
[2nd
Fing](Lax)[Thumb
Rotation] [2nd
Thumb(Contact)
Configuration Rotation
and
of
Fingers
Configuration
of
Thumb
Forearm
Involvement
Figure
16.
Organization
of
handshape features
While most
HC
use
only
the
hand,
others
use
the entire
hand
and
forearm
as
a
unit
(ALL-DAY).
Following
Stokoe(1960),
the
symbol
/
indi-
cates
the
presence
of
such
forearm
involvement
in
the
HC.
If
/
is
absent,
the
HC
is
assumed
to
use
only
the
hand
itself.
The
HC
description
we
have
developed
differs
from
most
other
approaches
in
that
it
notes
finger
configuration
and
thumb
configuration
separately. The
portion
of
the
HC
notation
concerned
with
finger
configuration
contains
slots
for
three
symbols.
The
first
is
handshape,
which
indicates
the state
of
extension
and
retraction
of
the four
fingers.
The
symbols
presented
in
Table
5
represent
those
combinations
of
open
and
closed fingers
we
know
to
occur
in
ASL
signing.
SLS
64

Phonological
Base
SLS
64
symbol
configuration
A
Four
fingers
closed
(pads
contact
palm)
S
Four
fingers closed
(tips
contact
palm)
1 ALL
but index
closed
I ALL
but
middle
closed
ALL
but
pinky closed
Y ALL
but
pinky
closed;
pinky
spread
ALL
but
pinky
and
index
closed;
unspread
ALl
but
pinky
and index
closed;
pinky
and index
spread
H All
but index
and
middle
closed;
unspread
V ALL
but index
and
middle
closed;
spread
K
Ring
and
pinky
closed;
index
open;
middle
partly
open
D Index
open;
all
others partly
open
R
Ring
and
pinky
closed;
index
and
middle crossed
r
Ring
and
pinky
closed;
middle
open;
index
partly
open
and
crossed
under
middle
W ALl
but
pinky
open
and
unspread
6 ALL
but
pinky
open
and
spread
7 ALL
but
ring
open
and
spread
8
ALl
but
middle
open
and spread
F ALl but index
open
and
unspread
9 ALL but index
open
and
spread
B ALL four
fingers
open
and
unspread
4 ALL four
fingers open
and
spread
T All
fingers
closed;
thumb under
index
N All
fingers
closed;
thumb under
middle
M ALL
fingers
closed;
thumb
under
ring
Table
5.
Symbols
for
taxonomic
description
of
major
finger
combinations.
Each
of
the four
fingers
is
independently
capable
of
being
in
one
of
four
basic
configurations:
open
(proximal
joint
(PJ)
and
distal
joint
(DJ)
extended);
closed
(PJ
and
DJ
flexed);
hooked
(PJ
extended,
DJ
flexed);flat-
tened
(PJ
flexed,
DJ
extended).
The taxonomic
symbols
presented
above
function
primarily
to
indicate
which
fingers
are
open
and
which
are
closed.
The
slot labelled
[2nd Fing]
in
the
schema
contains
diacritics
for
the
hooking
and
flattening
of
those
fingers
ordinarily
extended
in
a
given
hand-
shape.
Hooked
is
indicated
by
["];
flattened
is
indicated
by
[^].
Thus,
the
symbol
1"
indicates
that
the
index
is
extended
at
the
proximal
joint
and
flexed
at
the
distal
joints
and
the
symbol
B^
indicates
that
all
four
fingers
are
flexed
at
the
proximal
joints
and
extended
at
the
distal
joints.

Liddell
&
Johnson
The
diacritic
for
lax
[
-]
indicates
an
additional
modification to
the
finger
configuration.
It
relaxes
(slightly
reverses)
the
prominent
muscle
action at
both
the
proximal
and
distal
joints.
If
the
joint is
extended,
lax
will
flex
it
slightly,
although
not
enough
to
be
fully
flexed. Similarly,
if
the joint
is
flexed,
lax will
extend
it
slightly,
although
not
enough
to
be perceived
as
fully
extended.
Thus,
the
effect
of
laxing
is
that
the
finger
remains
as
specified
but not
rigidly
so.
Lax
tends
to
affect
all
four
fingers
but
has
no
effect
on
the
configuration
of
the
thumb.
All
details
of
thumb
configuration
are
specified
in
the
final
cluster
of
symbols.
The
primary
value
for
the
thumb
is
thumb
rotation.
The
proximal
joint
of
the thumb (near the
wrist)
is
capable
of
rotating
about
ninety
degrees
on
its
axis.
When
the
thumb
is
relaxed and roughly
adjacent
to
the
plane
created
by
the
palm
of
the
hand,
it
is
in
its
unopposed
[u]
rotation.
When
the
thumb
is
unopposed,
its
friction pad
faces
across
the
palm,
and
is
capable
of
contacting
the
radial
side
of
the
middle
joint
of
any
(flattened)
finger or
the
radial
side
of
the
palm.
Typically,
if
the thumb
is
touching the
palm,
it
is
in
unopposed
position.
The
thumb
may
also
be
rotated
so
that
its
friction pad
faces
the
palmar
surface.
This
is
its
opposed
[o]
rotation,
in
which
the
tip
of
the
thumb
may
easily
contact
the
tip
of
any
of
the
fingers.
The opposed
thumb
typically
cannot
touch
the
palm of
the
hand
except
at
the
base of
the
little
finger.
It
often
contacts
the
fingers
at
the
tip,
pad,
or
nail,
and
if
the
fingers
are
closed
may
contact
the
back
of
the
penultimate
finger
bones.
Both
opposed
and
unopposed
thumbs
must
also
be
specified
for
one
of four
values of
secondary extension
and
flexion,
indicated
in
the
[2nd
Thumb]
slot.
The
proximal
joint
of
the
thumb
is
near
the
wrist
and
along
with
the
two
more
distal
joints operates
to
define
the
same
four
values
of
extension and
flexion
available
to
the
fingers.
Because
the
thumb
features
are
descriptive
rather
than
taxonomic,
however,
open
and
closed
must
be
indicated. An
open
thumb
is
one
in
which
the
proximal
and
distal
joints
are
both
extended.
Thus
the
symbol
Bu
will
indicate
a
handshape
with
all
fingers
extended
and
unspread
and
a
thumb
that
is
on
the plane
created
by
the
palm
and
extended
at
about
ninety
degrees outward
from
the
radial
SLS
64
226

Phonological
Base
side
of
the
hand. The
symbol Bo
will
designate
the
same
finger
configura-
tion
with
the thumb
extended
at
a
ninety
degree
angle
from
the
palmar
surface.
Leaving
the
PJ
extended
and
flexing
the
DJ
provides
the
hooked
["]
thumb
configuration.
In
flat
[^]
thumb
configurations
the
PJ
is
flexed
and
the
DJ
is
extended.
In
the
[^]
configuration the degree
of
flexion
of
the
middle
joint
is
typically
adjusted
to
bring
the
thumb
pad into contact
with
either
a
finger
pad
(for
[o^]
thumbs)
or
the
middle
joint
of
the
first
finger
flexed
at
the
PJ
(for
[u^]
thumbs).
When
the
[u^]
thumb
is
not
in
contact
with
a
finger
it
is
in
pad
contact
with
the
radial
side
of
the
palm.
The
closed
[-]
configuration
flexes
both
the
PJ
and
the
DJ.
The
symbol
Bu-
indicates
the
B
fingers
with
the
thumb
flexed
and
in
contact
with
the
palm.
Ho-
indi-
cates
a
hand configuration
in
which
the
index
and
middle
fingers are
extended
and
the
thumb
is
closed
over
the
ring
and
little
fingers.
In
many
hand
configurations
the
thumb
contacts one or more of
the
fingers.
The
specifications
for
this
are
the
final
entry
in
the
hand
configura-
tion schema.
There
are
four
kinds
of
contact:
tip
contact
[c];
thumb
pad
contact
[p],
in
which
the
thumb
pad contacts
either
the
finger
pad
or
the
radial
side
of
the
finger;
finger
restrained
contact
[f],
in
which
the
thumb
pad
contacts
the
finger
nail;
and
thumb
restraint
[t],
in
which
the
finger
pad
contacts
the thumb
nail.
These
symbols
combine
to describe
every
hand
configuration
we
know
to
exist
in
ASL.
A
selection
of
them
is
presented
in
tabular
form
in
Appendix
A.
2.4.2
Point
of
Contact.
The
Point
of
Contact
(POC)
cluster contains
slots
for
four
symbols.
These
are:
location,
analogous
in
function
to
place
of
articulation
in
that
it
identifies
a
place on
the
passive
articulator;
handpart,
the
part
of
the
hand
that
is
located
there;
proximity,
how
near
the
handpart
is
to
the
location;
and
the
spatial
relationship
between
the handpart
and
the
location.
Three
different
kinds
of location
specification
may
be
entered
in
the
location
slot.
Some
signs
are
made
with
reference
to
a
location
on
the
body,
SLS 64

Liddell
&
Johnson
some
are
made
in
the
signing
space
surrounding
the
front
of
the
head
and
torso,
and some are
made
at
a
specific
place on
the
weak
hand.
Body
Locations
are those
places
where
lexically
distinctive
signs
may
be
made
on
the
head,
neck,
torso,
upper
legs,
or arms
(exclusive
of
the
hands).
We
have
found
that
the
accurate description
of
ASL
requires
many
more
phonetically
distinctive
body locations
than
proposed
in
earlier treat-
ments
of
sign
notation.
The
entries
describing
body
location
are
composed
according
to
the
following
schema:
(%) (i)
location
(t
or
b)
The
slot
labelled
location
is
filled
by
one
of
the
eighteen
major
body
locations
shown
in
Table
6.
BH
back
of head CN
chin
TH top
of
head
NK neck
FH
forehead
SH
shouLder
SF
side
of
forehead
ST
sternum
NS
nose
CH chest
CK
cheek
TR
trunk
ER ear
UA upper
arm
MO
mouth
FA
forearm
LP Lip AB
abdomen
JW jaw LG Leg
Table
6.
The
eighteen
major
body
locations
Diacritic
symbols
may
be
added
to
each
of
the
major
body
location
descrip,
tions
in
order
to
specify
other
locations
near
them. The
diacritic
[%]
indi-
cates
that
the
location
specified
is
on
the
side
of
the
body
contralateral
to
the
signing
hand.
If
this slot
is
empty
the
location
is
assumed
to
be
ipsi-
lateral.
Most
of
the
major
locations
specified
above
are
surrounded
by
a
set
of corresponding
locations
that
may
be described
by
adding
two
diacritics to
the
basic
location
symbol.
The
first
is
ipsilateral
[i],
indicating
that
the
hand
is
at
a
location
slightly
toward
the
outside
of
the
body
from
the
major
loca-
tion.
The
second
indicates
a
location
in
the
top
[t]
portion
or
bottom
[b]
portion
of
the
major
location.
Combining
these
entries
provides
the
loca-
SLS
64

Phonological
Base
tions
represented
in Figs.
17, 18,
and
19.
Appendix
B
presents
examples
of
lexical
signs
made
at each of
the
locations
we
know
to
be
distinctive
in
ASL.
Figure
17.
Articulatory locations
on the
head and
neck
Figure
18.
Articulatory locations
on
the
torso
SLS 64

Liddell
&
Johnson
Figure
19.
Articulatory
locations
on
the
arms
Signs
may
also be
produced
at
locations
in
the
signing
space
surrounding
the front
of
the
body
and
head.
Such
spatial
locations are
described
by
a
combination
of
a
diacritic
indicating
a
distance forward from
the
body
on
a
perpendicular
line,
a
symbol
indicating
the
extent
of
ipsilateral
offset
from
the
midline, and
the
symbol
for
a
major
central body
location:
Proximity
-
Ipsilateral
Offset
-
Central
Location
We
currently
distinguish
four
degrees of forward
distance for
spatial
locations:
proximal
[p],
indicating
a
location
within
a
few
inches
of
the
body
location;
medial
[m],
a
position
roughly
an elbow's
length from
the
body
location;
distal
[d],
a
comfortable
arm's
length from
the
body;
and
extended
[e],
a
full
arms
length
from
the
body
location.
The
side-to-side dimension
appears
to
require
two
degrees
of
ipsilateral
offset.
The
first
of
these
is
roughly
in
line
with
the breast
and
the
SLS
64
230

Phonological
Base
second
is
roughly
in
line with
the
outside
edge
of
the
shoulder.
In
order
to
avoid
confusion
with
the
set
of
finer
distinctions among
ipsilateral
offset
for
the
body locations,
we
refer
to
the
degrees
of
ipsilateral
offset
for spatial
signs
with
the
numbers
[0]
(no
offset),
[1],
and
[2],
respectively.
The
last
symbol
indicates
the
height of
the
spatial location.
It is
chosen
from
among
the
major
body
location
symbols
that
refer
to
points
along
the
midline
of
the
body
(TH,
FH,
NS,
MO,
CN,
NK,
ST,
CH,
TR,
AB).
Thus,
each
spatial location
is
represented
by
a
complex
of
three
symbols.
For
example,
the
symbol
m-0-TR
describes
a
location
about
an
elbow's
length
directly
in
front
of
the
solarplexis.
The
symbol
m-1-TR
indi-
cates
a
location at
the
same
height and
distance
forward,
but
on
the breast-
line.
Similarly,
the
symbol
d-2-FH
describes
a
location
about
an
arm's
length
forward
and
a
shoulder's
width
to
the
ipsilateral
side
of
the
center
of
the
forehead.
Appendix
C
presents
selected
signs
produced
at different
spatial locations.
Most
signs
appear
to
locate
on
points
like
those
described
above.
However,
one important
class
of
signs
makes
use
of locations
created
by
vectors
radiating
from
midline locations.
We
have
found
use
for
seven
such
vectors.
These
vectors
([L3]
[L2]
[L1]
[0]
[R1] [R2]
[R3])
and
the
locations
they
create
around their
intersection
with
the
lines
representing
degrees
of
distance
from
the
body
are
presented
in
Fig.
20.
One
such
semi-circular
system
of
locations
may
exist
at
each
contrastive height along
the
midline.
The
vector specification
substitutes
in
the
spatial
location schema for
the
ipsilateral
offset
number.
Thus,
m-R1-TR
specifies
a
location
at
TR height,
about
an
elbow's
length
out
from
the center
line
on
an
approximately
thirty
degree
right
vector.
Although the
addition
of
a
second
set
of
location
specifications
may
appear
to
be
excessive,
the
behavior
of
predicates
inflected for subject
and
object
agreement
and
the
behavior
of
locative
predicates
require
it.
We
will
return
to this
issue
in
more detail
below.
SLS
64

Liddell
&
Johnson
* /
eLITR
/
/dR.TR
m rL1TR
e
/
dLZTR
o
R7/
R
,,,LZTR
RT
Figure
20.
Spatial
vectors
used
by
agreement
verbs
For
many
signs,
the
location
of
the
strong hand
is
a
point
on
the
weak
hand
(FIRED).
The
schema describing weak
hand
locations
is
composed
of
two
symbols:
one
indicating
a
major
part
of
the
hand
(hand,
fingers,
forearm, thumb,
etc.),
and
the
other
indicating
a
zone
in
that
major hand
part
(inside, back,
radial
edge,
etc.).
The specifications
for
locations on
the
weak
hand
and examples
we
have
found
in
ASL
appear
in
Appendix
D.
The
handpart
slot
of
the
POC complex
will
contain
a
handpart
specifi-
cation
constructed
in
the
same
way
as
those
described
above.
Whereas the
handpart
specifications exemplified
in
Appendix
D
specify
weak
hand
loca-
tions,
the
handpart
slot
proper
indicates
which
part
of
the
strong
hand
makes
reference
to
or contacts
the
location of
the
POC.
An
inventory
of
strong hand
handparts
we
know
to
occur
in
ASL
is
presented
in
Appendix
E.
Combining
handpart
and location
in
POC,
we
would
find
that
the
first
segment
of
the
sign
GOOD,
for
example,
contacts
the
LP
location
with
the
SLS
64
232

Phonological
Base
fingerpads
of
the
strong
hand. The
POC
of
this
segment
will
contain
PDFI
in
the
handpart
slot
and
LP
in
the
location slot.
In
the
final
segment
of
the
sign
STOP,
the handpart
is
UL and
the
location
is
PA.
The
proximity
slot
of
the
POC
cluster
specifies
whether
the
handpart
is
in
contact
[c]
with
the
location
or,
if
not
in
contact,
then
its
distance
from
the
location.
It
appears
that three
distance
specifications (proximal
[p],
medial
[m],
and
distal
[d])
are
sufficient.
The
spatial
relationship
slot
of
the
POC
cluster
describes
the
direc-
tion
at
which
the
handpart
is
offset
from
the
location.
In
brushing
signs
the
hand
moves
between
points
on
two
sides
of
a
location,
making
brief
contact
as
it
passes
the
location.
For
example,
in
the
sign
FALSE
the
handpart
is
the RAFI
of
a
1o-
(index
extended)
hand configuration.
The
location
is
NS,
the
tip
of
the
nose.
The hand
begins
at
a
point
proximal
and
to the
ipsilateral
side
of
the
nose and
moves
to
point
proximal
and
to the
contralateral
side of
the
nose,
briefly
contacting
it
as
it
passes
(Fig.
21).
major
class
I M 1 H
contour
str I I
plane
| I
quality
Icontact|
I I
local
miovemient
___ ___I\
hand
configuration
| lo- I lo-
handpart
RAFI RAFI
point
of
proximity
I p I I
contact-
spatial
relation
I ipsi I Icontra
I
Location
NS NS
facing
--------
handpart
UL UL \
Location
VP VP
orientation
-----
handpart
I BA I I PA
I
plane
|_HP
l |
_HP
l
Figure
21.
FALSE
We
use
two
sets
of
spatial
relationship
symbols.
One
set
refers
to
locations
on
the
body
or
in
space
and
the
other
set
refers to
locations
on
the
weak
hand. Those
for
body
and
spatial
locations
are
the absolute
directions
over,
under,
behind
(toward
body
from
spatial
location),
ahead, contra,
and
ipsi.
Because
the
weak
hand
can
move,
the
spatial
relations
specified
with
respect
to
weak
hand
locations are
relative
to
parts
of
the
hand. The
set
SLS
64

Liddell
&
Johnson
includes:
tipward
[toti],
baseward
[toba],
toward
ulnar
side
[toul],
toward
radial
side
[tora],
palmward
[topa],and
backward
[tobk].
An
articulatory
bundle
specified
c
in
the
proximity
slot
may
be
left
unspecified
in
the
spatial
relation
slot.
2.4.3
Describing
Hand
Orientation.
The
POC
entries
in
the notation
simply
place
a
part
of
the hand
at
a
location.
At
any
location
it
is
possible
for
the
hand
to
assume
countless
orientations.
The
orientation
of
the
hand
is important
in
ASL signs,
for
both
lexical
contrast
and morphological
functioning.
It
appears
that
signs
make
use of
two
dimensions functioning
together
to
orient
the
hand.
The
first
of these
is
facing,
which
'points'
a
part
of
the
hand
at
a
location.
The
second
is
orientation
proper
which
usually
indicates
which
part
of
the
hand
is
pointing toward
the
ground.
The
facing
cluster
is
composed
of
two
entries:
one
for
a
handpart
and
one for
a
location.
The
orientation
cluster
is
also
composed
of
two
entries: one
for
a
handpart
(other
than
that
used
in
facing)
and
one for
a
plane
(usually
HP).
The
sign
STARE
exemplifies
the
interaction
of
facing
and
orientation.
In
citation
form
it
is
produced
as
a
hold
with
the
hand located
near
and
in
front
of
the shoulder,
with
a
V^o-
hand
configuration.
If
the
third person
object
is
associated
with
the
vector
R1,
the
tips
of
the
fingers
point
directly
forward toward
RI
and
the
base
of
the
hand
points toward
the
ground.
If
the
object
is
associated
with
the
vector
L2,
the
hand
remains
in
front
of
the
shoulder,
and
the
base
continues
to
point
to
the
ground,
but
the
tips
point
to
the
object
agreement
location,
in
this
case
mL2SH.
6
Numerous
object
agreement
inflections
may
be
achieved
by
altering
the
facing
complex
of
STARE,
independently
from
POC
and
orientation.
6
This
sign
also inflects
for subject
agreement.
In
fact,
the
example
shown
in
Fig.
22
is
the
appropriate
one
for
a
first
person
subject,
but
we will not
deal
with
this
issue
here.
SLS
64

Phonological
Base
STARE-R1
STARE-L2
Maj
CL H H
I I I1 1
I I I I
I I I I
I__
_I
I_
__I
HC
v
0- I v
0-
I
HdPt BK BK
Prox
p P
Sp.Ret
Loc sS I s
HdPt
| TIFI TIFI
Loc
mR1SH
mL2SH
HdPt
BA BA
Plane
HP HP
Figure
22.
STARE:
Two
different third
person
objects
3.
Morpheme
Structure Constraints
Upon
recording
a
corpus of
connected
signs
using
the
system
described
above,
it
becomes
clear
that
certain
phonetic
details
of
the
segmental
strings
are
predictable. For
example,
some details
of
phonetic
representations
recur
as
consistent
patterns
in
the
lexicon.
These
may
be
stated
as
morpheme
structure constraints
(MSC)
on
the combinations
of
features
and
segments
permissible
in
novel
lexical
forms.
Battison
(1974,1978)
identifies
several
MSC's
in
ASL,
based
on
the
notations
present
in
Stokoe
et
al. (1965).
As
a
result,
they
are
stated
largely
in
terms
of
a
simultaneous
model of
sign
structure. Nonetheless,
he
identifies
both
simultaneous
and
sequential
conditions
on
the
structure
of
ASL signs.
For
example,
he observes
that
the
hand configuration
R
may
contact
locations
in
only
a
relatively
limited
number
of
ways
(1978:38).
This
observation
can
be
restated
explicitly
as
a
segmental
MSC:
If
the
hand
configuration of
a
segment
is
specified
as
Ro-,
then the
hand
part
specifica-
tion
in
POC
will
be
one
of
the
following:
TIFI
(DONUT),
PDFI
SLS 64

Liddell
&
Johnson
(RESTAURANT),
BAFI
(CIGAR),
BA
(ROCKET).
7
Segmental
MSC's
such
as
this
will
constrain
the
inventory
of
segments
that
may
be
utilized
in
forming
novel
morphemes.
Similarly,
Battison
noted
that
in
signs
in
which
the
hand configuration
changes,
only
a
limited
number
of
sequences
occur.
One
such
sequential
MSC
states
that
if
two
segments
of
a sign
contain
different
hand
configura-
tion
specifications
and
the
final
hand configuration
is
lo-,
then
the
first
hand
configuration
will
be
1"o-f
(UNDERSTAND).
Similar
sequential
constraints
appear
to
pertain
to
the
following
final/initial
pairs
of hand
configurations:
Ho-
/
Ho"-f
(BEAT),
Vo-
/
Vo"-f
(TWELVE).
Such
constraints
describe
the preferential
structure
of
lexical
items
but
do
not
operate
as
phonological
processes
across
word
boundaries.
For
example,
in
the
clause
EXTREMELY-FOND-OF
##
NAME
'I
am
extremely
fond
of
that
name',
the
Ho- hand configuration
of
the
final
sign
NAME
does
not
predict
a
H"o-f
hand
configuration for
the
preceding
sign.
EXTREMELY-
FOND-OF
retains
its
So-
hand configuration,
resulting
in
the
sequence
So-
Ho-.
The sequence
H"o-f
Ho-
would be
ungrammatical
for
this
clause.
Many
other
constraints
such
as
these
appear
to
exist
in
the
lexicon,
and
will
ultimately describe
the
extensive
harmonic
sequencing observable
in
ASL
signs.
Battison
also
identifies
another,
more
unusual sort of
MSC,
which
specifies
co-occurrence
relationships
between
the
two
hands
(1974).
Spoken
languages
have
little
need
for
specifying
the
possibilities
of
co-
occurrence among
the
independent
articulators,
although
constraints
on
the
feature
[round]
and
constraints
describing
coarticulated
implosives
are
probably
similar
in
function.
In
ASL
it
is
possible
to
have
fully
specified
strong
and
weak
hands
performing
identical
activities
(LARGE)
or
mirror
7
Recently
introduced
signs
for
representing
English
words
whose
spellings
begin
with
ruse
three
other
hand
parts
in
POC:
PA (RELAX), UL
(RIGHT),
RAFI
(REALLY),
but the
use
of
such
introduced
signs
is highly
constrained.
SLS
64

Phonological
Base
image
activities
(MAYBE),
or
completely
different
activities
(FIRED).
Moreover,
there
are
minimal
contrasts
among
one-handed
and
two-handed
signs
(LIKE;
INTERESTING),
so
the
weak
hand
is
not
completely
predict-
able,
and
must
be specified.
Battison's Dominance Condition
specifies
rather
rigid
limitations
on differences
between
the
hands.
He
points
out
that
if
the
two
hands
have
different
hand
configurations
then the
hand
configuration of
the
weak
hand
must
be
chosen from
a
very
limited set
of
easily
discriminable
hand
configurations,
while
the hand
configuration
of
the
strong
hand
is
much
less
constrained. The
refinement
of
MSC's
of
this
type
promises
to
be
a
rich
area
of
research
in
the
segmental
phonology
of
ASL.
4.
Phonological
Processes
The
phonological
strings
contain
still
another
sort
of
predictable
detail,
traceable
to
phonological
processes, producing
alternations
among
surface
forms. These
processes
are
typically
described
by
a
complex
of
phonological
rules,
each of
which
may
alter
some
detail
of
the
representa-
tion
of
a
form
or
add
non-lexical
phonological
information
to
a
string.
The
combined
action
of these
processes
ultimately
derives
the
surface
represen-
tation
of
the
string.
4.1
Movement
Epenthesis.
Phonological
processes
proper
influence
the phonetic
shape
of
phonological
strings.
Many
of
the
phonological
processes
known
to
occur
in
spoken
languages
appear
also
in
ASL.
The
most
easily
described
is a
process
which
inserts
a
movement
between
concatenated
segments,
the
second
of
which
begins
with an
initial
articulatory
bundle different
from
the
final
articulatory bundle
of
the
preceding segment.
For
the
most
part,
this
process
applies
at
the
boundary
between
signs
and
enjoys
the
relatively
straightforward
function
of
moving
the
hand
from
the
articulatory
posture
SLS
64

Liddell
&
Johnson
MOTHER
##
MULL-OVER
'Mother
is
muLLing
[it]
over'
underLying
form
surface
form
H | H | H | M | H |
P
| I I II II
SP
IIi
I I II
H
i
I_
I -I
ZI
I
4u
| 1
B^
o-c|
4u
IB^
o-cl
|
TITH
TIFI
TITH
TIFI
l IIci
Ic
I Ic
l I l I I I I
CN j iFH |CN | iFH
RA I A RA PA
SP SP SP SP
BA BR BA j BA
EL HP EL
_HP
Figure
23a. Effect
of
movement epenthesis
Figure
23c.
MULL-OVER
SLS
64
Figure
23b.
MOTHER

Phonological
Base
23d.
MOTHER MULL-OVER
Figure
23
The effect
of
M-epenthesis
in
the
string
MOTHER MULL-OVER
that
ends
one
sign
to
the
articulatory
posture
that
begins
the
next.
In
the
case
of
MOTHER
MULL-OVER
the
Movement
Epenthesis
rule
inserts
an
M
segment between
the
last
segment
of
MOTHER
and
the
first
segment
of
MULL-OVER.
Although
it
may
seem
to
be
unnecessary
to
propose
a
rule describing
a
process
so
predictable,
pervasive, and
physiologically
motivated,
the
M
segment
introduced
into
strings
by
the
M
epenthesis
rule functions
as
a
critical
part
of
the
environment
that
feeds
another
phonological
process.
4.2
Hold
Deletion
That
process
is
Hold
Deletion,
which, with
certain
exceptions,
eliminates
hold
segments
occurring
between
movement
segments.
The
surface form
of
the
phrase
GOOD
##
IDEA
'good
idea'
demonstrates
the
application of
the
Hold
Deletion
Rule.
SLS 64
239

| H |
GOOD
M|
str
I
I I 1 1 I
BAu^
B^u^
PDFI BKFI
C
[
C
I
| LP |
Im-0-CH
I
| PA I I BA I
I CH I CH I
I BA I I
HP II HP I
Liddell
&
Johnson
## IDEA
H | M H
str
1
1 1 1
|
Io-
| |
Io-
PDFI
| PDFI
IC l I P I
I
Iahead
iFH
iFH
I
PA PA
SP SP
BA BA
HP jI HP
Figure 24.
Underlying form
of
GOOD
IDEA
Because
the
sign
GOOD
ends
with
a
segment
articulated
in
a
different
way
from
the
initial
segment
of IDEA,
the
M
Epenthesis Rule
will
insert
a
segmental
bundle,
specified
as
M,
between
the
two signs.
This has
the
effect
of
moving
the
hand
from
the
area
immediately
in
front
of
the
chest
to
a
location
in
contact
with
the
side
of
the
forehead
and
simultaneously
changing
the
other
articulatory
specifications
from
those
describing an
open
hand
oriented
with
its
back to
the
HP
to
those
of
a
hand
with
only
the
little finger
extended
and
oriented
with
the
tip
of
the
little
finger
upward.
Whereas
the
isolated
signs
GOOD
and
IDEA
end and
begin
with
substantial
holds,
when
juxtaposed
in
this
phrase
the
final H
of
GOOD
and
the
initial
H of
IDEA
are
deleted.
The
critical
environment
for
the
applica-
240
SLS
64
H|

Phonological
Base
H
M H M H M H
str str
I II
lI
Il II 11 11 I
I 11 II I1 II II
||I
| I I 1 1 I
li
_
B^u^
u^
I 10- 10-
PDFI BKFI PDFI PDFI
Ic IC I IC I IP
I I I I I I
I
ahead
LP
m-0-CH
iFH
iFH
PA BA | PA PA
CH CH I SP SP
BA j BK j BA BA
I HP I HP I I HP HP
Figure
25. GOOD
IDEA
(Movement
Epenthesis
applied)
H M M |M | H
I |[ str I I I
I
str II
I I
l I I 1 I l l
I II I I I I II I
B u
B^u^
o-
I
o-
PDF I BKF I PDFI | PDFI
c l I eIc Ic I p1
I I I I I I I ahead
LP
|m-0-CH
I
iFH
iFH
PA
I
I BA I I PA PA
CH | CH I I SP SP
BA | BK I A
I
BA
HP 1 HP I I HP IHP
Figure
26a.
Final
form
of
GOOD
IDEA
(Hold
Deletion
applied)
SLS
64

Liddell
&
Johnson
Figure
26b.
GOOD
Figure
26c.
IDEA
26d.
GOOD
IDEA
Figure 26.
The
effect
of
M-epenthesis
in
the
string
GOOD
IDEA
SLS
64

Phonological
Base
tion
of
this
rule
seems
to
be
the
M
segments
that surround
each
H
segment.
8
The
surface
form
of
the
clause
IDEA##GOOD
'The
idea
is
good'
is
also
affected
by
the
H
Deletion
rule,
which
again
causes only
the
inter-M
holds
to
be
deleted.
IDEA##GOOD > (M
Epenthesis)
> (H
DeLetion)
Ii
I~LIllHI
Ifltlimlitil~lil
hilmliililil
LI/LII
L^ i
L
I
L
\ I L\ I
LI LI
L
NI
1-1
M 1
HI11
MI
I H 1- 1
11
1- 1 1 Ml 161 161
1 1M
1-1 1-1
Figure
27.
M
epenthesis
and
H
deletion
Certain
conditions prohibit
application
of
the
H
Deletion
Rule.
Holds
that
are
lengthened,
either
by
the
presence
of
local
movement
or
by
morphological
processes
such
as
the
one
which
produces
a
lengthened
H
at
the
beginning
of
emphatic
forms,
tend
not to
delete.
Moreover,
it
appears
that
the
application of
H
Deletion
is
variable
by
context.
Although
the
extent
and
exact
nature
of
the
variation
is
not
yet
clear,
it
appears
that
H
segments
that
do
not
contact
the
body
or
the
other
hand
are generally
deleted
in
inter-M
contexts
(as
long
as
they
are not
lengthened),
whereas
8
A
treatment
whereby
lexical forms
of
such
signs
contain
terminal
M
segments
and
H
segments
are
inserted
finally would also
have
to
propose
that
the initial
H
segments
were
also
inserted
by
phonological
process.
This
is
not
an
appealing
solution,
however, since there
exist
signs
with
initial
M
segments
that
are
not
preceded
by
H
segments,
even
in
isolation
(WHEN).
We
know of
no
principled
way
to to
predict
which
signs
would
add
an
H
and
which would
not.
Moreover,
a
number
of signs
consist
of
only
a
hold
in
isolation
but
are
deleted
between
M's. The
underlying
M
solution
would
clearly
not
work
for
such
signs since
they
have
no
M.
The
alternative proposal
would amount to
a
claim that
they
have no
segmental
structure
in
their underlying forms
which
appears
to
introduce
unnecessary
complication
to
a
theory
of
lexical
structure
of
ASL.
SLS
64

Liddell
&
Johnson
those
that
do
contact
another
body
part
are
variably
deleted.
The
following
combinations
result
(+
indicates body
contact).
GOOD##IDEA GOOD##IDEA
1IIl
L
I-I 1 1
IJ
I l II
I~~~~~~
/\C/OD/ -/
I /\
A
GOOD##IDEA
* 1iA11
1111
l 1
111
jal IbI |c| |dl
Figure
28.
Possible
and
disallowed
application
of
H
deletion
4.3
Metathesis.
A
number
of
signs
exchange
an
initial
sequence
of
segments
with
a
sequence
of
final
segments
in
certain
contexts
that appear
to
be
purely
phonological.
The
sign
DEAF
is
typical
of
such
metathesizing
signs.
I M l H
lM i M I H I
| str
str
str I
I II II1-
1 1 1
li
I10
-
I II II
II
II I
[2
lo-
|_lo-
I
RAFI RAFI
RAN
1 RAFt
CK
1 CK JW JW
RA RA RA RA
SP SP SP SP
BA 11 BA BA 11 BA
_HP
_ HP j
_HP
l
_HP
l
Figure
29a.
DEAF
SLS
64
244

Phonological
Base
Figure
29b.
DEAF
In
this
form
of
the
sign
the
index
finger
first
moves
to
contact
the
cheek and
then
moves
to
contact
the
jaw.
This
form
of
the
sign
typically
occurs
immediately
following
signs
produced
in
the
higher
facial
areas.
Thus,
it
would be
likely
to
occur
in
the
clause
FATHER
##
DEAF
'Father
is
deaf,
since
FATHER
is
produced
with
contact
at
iFH.
However,
if
DEAF
is
immediately
preceded
by
a
sign
in
the
lower
facial
regions
(and
perhaps
other
lower
areas), the
initial
two
segments
are
exchanged
with
the
final
two
segments.
In
the
clause
MOTHER
##
DEAF
'Mother
is
deaf,
the
sign
MOTHER
produced
at
the
chin
causes
DEAF
to
be
produced
as
follows.
SLS 64

Liddell
&
Johnson
M H M
I
str
1
str
1IIIIII
lo-
I
RAFI RAFI
I P
ll
e lI
I II J I I
RA RA
SP | SP
BA BA
l_HP_|
_HP
1 L
M I I
str
lo-
o
RAFI RAFI
PI l
CK CK
RA RA
SP SP
BA BA
HP_I|_HP_1
Figure
30a.
DEAF
(after
metathesis)
Figure
30b.
DEAF
(after
metathesis)
The
sign
WE
further
illuminates
the
metathesis
process.
There
are
two
forms
of
WE;
one
has
a
segmental
structure
like
that
of
DEAF,
the
SLS
64

Phonological
Base
other
has
an
HMH
sequence,
with an
arc
M.
WE
1
metathesizes
but
WE
2
does
not
(Fig.
31).9
l
I I
li i
alibi
Ic|ldi
L
llI
1_11 1
Figure
31.
ISlIllIIl
II 1-1
b. WE
2
a.
WE
1
Figure 31c.
WE
1
The
signs
CONGRESS,
FLOWER,
RESTAURANT,
DEAF,
HONEYMOON,
NAVY,
TWINS,
BACHELOR,
PARENTS,
HOME,
and
9
The
feature bundles
in
these two signs
share
many features.
That
is,
feature
bundle
'b'
is
closely
related
to
feature
bundle
'B'.
Likewise, feature
bundle
'd'
is
very similiar
to
feature
bundle
'D'-
SLS 64

Liddell
&
Johnson
HEAD
have
all
been
observed
to
undergo
metathesis.
All
these
signs
have
31d.
WE
2
Figure
31.
Alternate form
of
WE
the
same
basic
segmental
structure
as
DEAF,
i.e.,
a
movement to
a
hold
at
one
location
followed
by
a
movement
to
a
hold
at
another
location.
Because no
sign
with
another
segmental
structure
has
been
observed
to
metathesize,
application
of
the
phonological
rule
appears
to
require
this
underlying
segmental
structure.
However
not
all
signs
with
this
underlying
segmental
structure
may
metathesize.
BODY,
KING,
CHRIST,
INDIAN,
BLOUSE, THANKSGIVING,
CHILDREN,
and
THING
all
have
the
appropriate
segmental
structure
but
may
not
metathesize.
Most
of
these
share
the
characteristic
that
their
two
contacts are
in
markedly
different
locations
on
the
body.
The
last
two
do
not make contact
with
the
body.
These tentatively
appear
to
be
additional
phonological
constraints
on
the
application
of
the
rule.
These
observations
carry
two
important
implications
for
the
general
theory
of
the
structure
of
signs we
are proposing
here. The first
is
that
we
have
some
justification
for
treating
signs
with
this
segmental
structure
as
SLS
64

Phonological
Base
having
two
lexical
parts.
Specifically,
we
propose that the
underlying
form
of
such
signs
contains
two
unconnected
MH
sequences,
which
are
subject
to
metathesis
and
which
(whether
or not
metathesis
has
applied)
are
connected
by
the
M
Epenthesis Rule,
as
represented
in
Fig.
32.10
Lexical
Form >
Metathesis
> M
Epenthesis
MI11 I I lxi t 16li 1611HI IMl lxi
MiIMIH
|_1I
Li
I
II
LIL
L
IA
_1I
WIbi
clidi IcIldi alibi Iclidi
alibi
Figure
32.
Relationship
between metathesis
and
epenthesis
Signs
such
as
WE
2
have
a
unitary
lexical
form
HMH,
which may
not
be
permuted
by
metathesis
and
in
which
the
segmental
information
in
the
M
must
be
specified
as
an
arc.
The
second
important
implication of
these
observations
suggests
that
a
complete
feature
analysis
of locations
will
provide
insights
into
the
nature
of
phonological
processes. First
it
is
probable
that
some
feature
or
set
of
features
unites
the
sets
of locations
between
which
metathesis
may occur
and
distinguishes those
which
are
saliently
distant
enough
to
prohibit
metathesis. Moreover,
the
conditioning
of
the Metathesis
Rule
by
prior
signs
will
depend
on
a
feature
analysis
that
recognizes
that
certain
locations
are more
to
the
left
or
right
or
below
or
above
certain
other
locations.
Only
features
that
carry
this
sort
of
information
may
condition
the appropriate
application of
the
Metathesis
Rule.
Such
featural
information
will
account
for
the
fact
that
signs
made
on
the stomach,
the
chest,
or
the
chin
may
all
provide
the
condition
that
selects
initial
occurrence
of
the
lowermost
sequence
of
DEAF.
10
Hold
Deletion
may
optionally
apply
to
the
first
hold
of this
string,
yielding
an MMMH
surface form.
In
addition, although
the
derivation
is
presented
in
ordered form,
M
Epenthesis
and
Metathesis
appear
to
be
unordered with respect to
each
other.
SILS
64

Liddell
&
Johnson
4.4
Gemination.
Although
such
occurrences
are
rather
rare
in
ASL,
it
sometimes
happens
that
the
terminal
segment
of
one
sign
is
identical to
the
initial
segment of
the
following
sign.
In
the sentence,
SPAGHETTI
3
A, MOTHER
REPULSED-BY
3A
'mother
really
hates
spaghetti'
the
final
segment
of
MOTHER
(the
form
of
MOTHER
without
local
movement)
and
the
initial
segment of
REPULSED-BY
are identical
holds.
The result
is
a
single
long
hold.
An
epenthetic
movement
away
from the
chin
or
a
hold
of
normal
length
is
ungrammatical.
4.5
Assimilation
There
are
numerous
instances
of
assimilation
in
ASL.
For
example,
the
hand configuration
of
the
sign
ME
typically
assimilates
to
that
of
a
contiguous
predicate
in
the
same
clause.
Thus,
whereas
the
underlying
form
of
ME
contains
a
lo-
hand configuration,
in
the
string
MOTHER
3ASTARE-AT1. ME
GULP.
'Mother
was
staring
at
me
and
I
was
nervous
about
what
was
to
come'
ME
assumes
the
9o-c
hand
configuration
of
GULP.
The
extent
to
which
signs
other
that
ME
assimilate
to
the
hand
configuration
of
another
sign,
although
not
yet
thoroughly
investigated,
appears
to
be
considerably
more
limited.
SLS
64
250

Phonological
Base
33a
ME
33b.
GULP
33c.
ME
GULP
Figure
33.
Phonological assimilation
of
handshape
features
in
the string
ME
GULP
SLS
64

Liddell
&
Johnson
Assimilation
of
the
hand
configuration
of
the
weak
hand to
that
of
the strong hand
in
two
handed
signs
is
quite common.
For
most
signers
it
appears
to
be
variable,
probably
controlled
by
formality and fast-signing
constraints.
Thus,
it
is
common
that
in
signs
in
which
the
strong
and
weak
hand
configurations are
different
in
formal
signing,
the
weak
hand
configuration
will
be
fully
assimilated
to
the
strong
hand
configuration
in
casual
or
fast
signing.
We
have
observed
numerous
other
examples of
assimilation
in
ASL.
Among
these
are
the
assimilation
of
orientation
and
facing
features
of
the
weak
hand
to
those
of
the
strong hand;
assimilation of
features
specifying
location
in
POC
of
an
initial
segment
of
one
sign
to
the
location
features
of
the
final
segment
of
the
preceding
sign;
assimilation
of
location
features
of
the
final
segment
of
a
sign
to
the
location of
the
initial
segment
of
a
following
sign;
two-handed
signs
becoming
one-handed
as
a
result
of
assimilation to
a
one-handed
sign
in
the
same
string;
one-handed
signs
assimilating to
two-handed
signs.
These
processes
await
more
detailed
description.
4.6
Reduction.
Frishberg
(1975)
notes
a
number
of
historical
trends
in
ASL
which
she
identifies
as
'displacement'.
Each
of
these
involves
the diachronic
relocation
of
certain
signs
to
areas
either
less
central to
the
face (and
thereby
less
likely
to
obscure
important
facial signals)
or
to
areas
more
central to
the
lower
head
and
upper
body
regions
of
the
signing
space
(and
thereby more
readily
perceptible).
Although
such
forms
appear
to
be
lexicalized at
their
new
locations,
the
phonological
processes
that
originally
must
have
moved
them
are
still
active
in
contemporary
ASL.
The
rules
which
account
for
them
appear
to
be
variably'selected
by
casual
signing,
and,
like
vowel
reduction
rules
in
spoken
languages,
have
the
effect
of neutralizing
contrasts
of
location.
Thus,
many
signs which
are
produced
with
contact
at
the
SFH
location
in
SLS
64
252

Phonological
Base
formal
signing
may
be
produced
in
casual
signing
at
the
CK
location.
Similarly,
signs
produced
at
the
CK
location
(including
those
moved from
the
SFH
location)
may
be
produced
at
the
JW
location. These
same
signs
also
appear
at
times
without contact
in
the
area
immediately
in
front
of
the
iNK
location.
The
first
segment
of
the
sign
KNOW-THAT
is
produced
formally
at
the
SFH
location
but
may
occur
in
casual signing
at
any
of
the
other
locations described
above.
In
a
somewhat
similar
manner,
signs
produced
at
a
location
proximal
to,
but
not
in
contact
with
FH
or
NS
in
citation
form
(KNOW-NOTHING,
DOUBT)
and
signs
produced
with
contact
at
the
mouth
(GLASS)
may
be
produced
at
the
CH
location.
Signs
that
do
have
underlying
contact
at
the
FH
or
NS
locations
are not
subject to
the
effects
of
this
rule
(FATHER,
BLIND).
Similar
rules
exist to
reduce
peripheral
locations on
the
torso
to
more centralized
locations.
It
appears
also
that
there
are
rules
that
reduce
the
distance
between
the
locations
of
two-location
signs
in
casual signing.
The
MMMH
sequence
of
the
type
isolated
by
the
metathesis
rule
(CONGRESS,
HOME)
is
commonly
reduced
by
such
a
rule,
and
it
appears
that
many
other
segment
sequences
also
undergo
a
similar
reduction
process
(GOOD,
GIVE,
etc.).
Similarly,
the
size
of
the
first
(round)
movement
in
MMH
sequences
such
as
YEAR,
WHEN,
POLITICS,
and QUESTION
is
often
reduced
in
casual
signing.
4.7
Perseveration
and
Anticipation.
Typically,
signed
strings
contain both
one-handed
and
two-handed
signs.
When
a
one-handed
sign
follows
a
two-handed
sign,
although
the
weak
hand
is
not
required,
in
casual
and
fast
signing
it
commonly
either
perseverates
features
of
the
former
sign
or
anticipates
features of
the
following
sign,
or
both,
rather
than
returning
to
a
resting position.
Although
these
processes
and
other
very
late
phonological processes
such
as
reduction
have
the
relatively
trivial
phonological function of
speeding
and
smoothing
the phonetic
string,
they apply
very
broadly. Thus,
because
SLS 64

Liddell
&
Johnson
they
apply
to most forms
produced
in
comfortable
signing,
these
processes
commonly have
a
substantial
impact
on
the
underlying form
of
lexicalized
compounds
and
other
lexical
entries
that
result
from
the
lexicalization
of
productively
produced
forms.
5.
Morphological
Processes
Another
sort
of
predictable detail
originates
in
the
morphology,
where
morphological
processes create
words.
Across
languages,
words
are
formed
by
attaching
lexical
forms
to
one
another
and
by
moving,
reproducing, deleting
from, adding
to,
and
altering
the
phonological
information
carried
by
lexical
forms.
Although
both
morphological
proc-
esses
and phonological processes
may
add,
delete,
alter,
or
move
phonological details,
they
differ
in
that
phonological processes do
not
account
for meaning
changes
whereas morphological processes do.
Below
we
will
describe
a
small
selection
of
ASL
morphological proc-
esses
that
illustrate
the
diverse
phonological
effects
which
result
from
their
application.
We
have
divided
these
processes
into
two
broad
categories.
In
the
first,
meaningful
feature
bundles
(morphemes)
are
inserted
into
one
or
more
segments of
a
root
with
incomplete articulatory
feature
bundles.
This
insertion results
in a
phonologically
fully
specified
stem.
In
the
second
major
category,
the
morphological
processes
operate
on
a
completely
formed
stem
either
by
removing some of
its
phonological
features
and
inserting
them
in
a
segmental frame,
by
modifying
them
through
reduplica-
tion,
or, rarely,
by
attaching
an
affix.
5.1
Processes
that
Insert
Features
in
Roots,
For
many
ASL signs,
we
posit
lexical
forms
of
roots
with
empty
spaces
(or
'cells')
in
their
underlying
feature
specifications.
A
number
of
ASL
SLS
64
254

Phonological
Base
34a
FIRST-PLACE
34b.
SECOND-PLACE
34c.
THIRD-PLACE
Figure
34.
Substitutability
of numeral
morphemes into
a
phonologically
incomplete
root
morpheme
morphological
processes
'fill
out'
such
incompletely
specified
roots
with
morphemes
which
consist
of
the
small
bits
of phonological
information
SLS 64

Liddell
&
Johnson
used
to
fill
the
empty
cells in
the
root.
The
three
signs
in
Fig.
34
are
repre-
sentative
of
a
large
class
of
such
signs,
built
from
roots
specified
for
all
their
features
except
hand
configuration.
These
three
signs
are identical
except
for
their
hand
configuration.
FIRST-PLACE
is
produced
with
a
lo-
hand configuration,
SECOND-
PLACE
has
a
Vo-
hand
configuration,
and
THIRD-PLACE
has
a
Vu
hand
configuration.
Signs
meaning
FOURTH-PLACE
through
NINTH-PLACE
can
be
formed
by
using
other
hand
configurations.
In
numerous
other
signs
the
same
hand
configurations
convey
equivalent
meanings
of
numerosity.
We
contend
that
these
signs
(and
others
with
numeral
hand
configurations)
contain
at least
two
morphemes:
the
root
morpheme,
a
numeral
classifier
which
means
"place
in
a
competition", and
the
numeral
morpheme.
The
two
morphemes
in
SECOND-PLACE
and
their
phonological
relationship
to
one
another
are sketched
below.
I
I
I
Mj
I
H
I
I
I
I
Str
UL UL
P Ic
ahead
Sp2T p2ST
BA BA
I SP SE
jUL
UL
HP HP
PLACE-IN-COMPETITION
NUMERAL-2
(incomplete S-morph) (p-morph)
Figure
35.
The
two bound
morphemes
required
for
"first
place",
second
place",
etc.
The
root,
PLACE-IN-COMPETITION,
is
composed
of
three
segments
and
two
incompletely
specified
articulatory
feature
bundles.
A
numeral
morpheme
is
required
to
complete
the
phonological
representation
of
the
SLS
64

Phonological
Base
stem
SECOND-PLACE.
We
refer
to
roots
such
as
PLACE-IN-COMPETITION
as
"Incomplete
S-morphs",
since
their
phonological
representation is
segmental,
but incomplete (Johnson
and
Liddell
1984)
The
numeral
morpheme
is
referred
to
as
a
"P-morph"
since
it
only
provides
paradigmatic
contrast
(i.e., it
contains
no
segmental
information).
It
can
be inserted
into
a
root
consisting
of one or
more
segments
and
its
features
simply
spread
according
to
autosegmental
principles.
We
have
identified
more
than
thirty
different
incomplete
S-morphs
which,
like
PLACE-IN-COMPETITION,
require
the
insertion
of
a
numeral
morpheme.
11
A
second major
category
of
incomplete
S-morph contains
verb
roots
with
unspecified location
information.
The completed
form
of
the
verb
stem
of
such
signs
contains location
(vector)
specifications received
through
the insertion
of
subject
and/or
object
agreement
morphemes.
Two such
verbs,
ASK
and
TELL
are
illustrated
below.
36a.
TELL
36b.
ASK
Figure
36.
An
object
agreement
verb
(TELL)
and
a
subject-object
agreement verb
(ASK)
11
Many
of
these
are
analyzed
in
detail
in
Liddell,
Ramsey, Powell,
and
Corina
(1984).
SLS
64

Liddell
&
Johnson
The
initial
location for
TELL
is
the
chin.
Its
final
location,
however,
is
determined
by
the
insertion
of
an
object
agreement
morpheme.
In
the
illustration
above,
TELL
agrees
in
location
with
the
3rd
person
object
already indexed on
the
signer's
left.
12
I HI [
s
I j
str
|2
lo
lo-
IPDF
TIFI
I
e
I
I
P
I
I I
_behind|
|I CN I
IsrI
IL1
I
_BK
BA |
I1L-
ICN 1 I
sP
I
I BA j BK |
LHPI I
_HP_I
TELL
Agreement
Morpheme
Figure
37.
The
shape
of
TELL
with
an
object
agreement morpheme
ASK
is
structured
so
as
to
allow
both
object
agreement
and subject agree-
ment
morphemes
to
be
inserted.
The
subject
agreement
morpheme
pictured
on
the
left in
Fig.
38
is
determined
by
the
person
and
location
of
the
subject
nominal,
and
is
inserted
into
specific
places
in
the
two
feature
bundles.
The
object
agree-
12
In
ASL
discourse
any
nominal
may
be
assigned
a
grammatical
association
with
a
spatial
location
or
vector.
The
process
of
assigning this
association
has
been
called
'indexing'
and the
location
or vector
associated
with
the
nominal
has been
called
its
'index'.
While
ASL
pronouns
may make reference
to
a
nominal
by
pointing
at
its
index,
verbs
such
as
TELL
and
ASK
agree with
their
subject
and
object
nominals
through
the insertion
of
agreement morphemes.
The
agreement morphemes
are
p-morphs,
the
phonological
form
of
which
is a
specification
determined
by
the
location
of
the index
of
a
nominal.
SLS
64

Phonological
Base
ment
morpheme
is
determined
by
the
person
and
location
of
the
object
and
is
similarly
inserted
into
both
articulatory
bundles.
Thus,
the
completed
verb
stem
3a-ASK-3b
is
composed
of
three
morphemes:
one
root
and
two
agreement
morphemes.
In
the
examples
of
feature
insertion
discussed
so
far,
the
root
contains
only
a
small
number
of empty
cells.
Many
other
signs
are
built
from
roots
that
are
specified
only
for
segment
type,
and contain
empty
cells
for
all
other
segment
features
and
all
articulatory
features.
This
class
of
signs
has
been
referred
to
as
"classifier
predicates"
by
Liddell
(1977),
and
"verbs
of
motion
and
location"
by
Supalla
(1978),
who first
proposed
the
idea
of
movement
roots
in
the
analysis
of
these
signs.
Morphological
processes
insert
a
number
of
morphemes
in
appropriate
cells
to
derive
a
polysynthetic
predicate
stem.
I
H
I
I I I
H
I
strj
I
I I
str_
I I
_ _
I
lo-
II
0-
BK j PA
mR21
_|
|ST|
|l_ STJ mL21
I -I
I
l- I -
1I IIBK
BA j BA
HP HP
Subj
ASK
Obi
Agreement Agreement
Figure
38.
The shape
of
ASK
with
subject
and
object
agreement morphemes
The
type
of
information
which can
be
inserted
into
such
movement
roots
has
been
investigated
in
depth
by
Supalla
(1978).
We
will
not
SLS 64
259

Liddell
&
Johnson
SLS
64
I ...
trajectory morpheme
...
manner
morpheme
...
cLassifier
morpheme
I I I ...
primary
facing
mphm.
...
proxim.
to Loc.
mphm.
I I I I ...
sp.
reL.
morpheme
...
primary
Loc.
mphm.
I I I I_ ...
sec.
facing
morpheme
... sec. Location
morpheme
SI
I
_
I ...
verticatity morpheme
I
_
I I ...
pLane
of
reference
morpheme
CLassifier
P-morphs
Predicate
BE-AT
Figure
39.
The
composition
of
a
classifier
predicate
provide
additional
analysis
here, but
simply
observe
that
this
category
of
predicate
is
highly
productive
in
ASL
and
is
responsible
for
a
significant
number
of
the
signs
observed
in
ASL
discourse.
5.2
Processes
that
Operate
on
Fully
Specified
Stems.
The
processes
we
describe
below
all
operate
on
fully
specified stems.
Such
stems
can
either
come directly
from
the
lexicon
as
completely
specified
s-morphs,
or
become
fully
specified
through
processes
like
those
described
above.
5.2.1
Frames.
Many
ASL
inflections
have
an
unusual characteristic.
Regardless
of
the
syllable
structure of
the
uninflected stem
(the
input
to
the
process),
the
syllable
structure
of
the
inflected form
(the
output)
is
completely
uniform.
For
example,
Liddell
(1984b)
describes
the
verb
inflection
for
unrealized-
inceptive
aspect.
The
input
to
the
inflection
could
be
a
verb
with
a
single

Phonological
Base
segment,
two
segments,
or
even
three
segments.
The
inflected
verbs,
however, uniformly
have
the
shape
MH.
40a.
TELL
40b.
TELL
(unrealized-inceptive)
Figure 40. TELL and its
unrealized-inceptive
form
In
this
analysis
the
inflected
verb
is
not
strictly
a
modification
of
the
verb
stem,
but
rather
results from
feeding
a
small
piece of
articulatory
information
from
the
verb
stem
into
a
segmental
structure refered
to
as
an
'inflectional
frame'.
The
diagram
below
shows
the shape
of
the
uninflected
verb
stem
TELL,
and
its
form
when
inflected for
the
unrealized-inceptive
(U-I)
aspect.
For
verb stems
in
the
same
verb
class
as
TELL,
the
initial
feature
bundle
of
the
stem
is
identical
to
the
final
feature
bundle
of
the
U-I
form
of
the
verb.
Further,
all
of
their
U-I
forms
have
the
form MH,
and
all
have
the
same
location
features
in
the
initial
feature
bundle.
The
inflectional
frame
is
the
phonological
structure
provided
by
the
inflection itself. This
frame
is
not prefixed
or
suffixed
onto
the
stem,
but
rather,
serves
as
the
phonological
framework used
to construct
the
inflected
sign.
SLS
64

H M H
str
I II
___ __
lo-
|
lo-
aFI
|
TIFI
Ic
|pl
I I I
behind
|CN
_mL1ST
BK BA
| mL1CN SP
BA BK
HP HP
TELL
Verb
Stem
SLS
64
Liddell
&
Johnson
IM
H
arc
1 0 1 1
_Jl-
_
lo-
|
TIF
[ PDFI
| c
I
c
_( mOTRt _ CN
PA BK
| SP _ mL1CN
UL BA
HP HP
TELL
Unrealized-Inceptive
Figure
41.
The
stem
TELL
and
its
unrealized
-inceptive
form
I Nre I
HI
I M I
I I
I I
I I
| mOTRt
I I
Figure
42
Unrealized-Inceptive
Frame
The frame
has
a
partially
specified
initial
feature
bundle,
but
no
final
bundle
of
features.
For
verbs
like
TELL,
which
begin
in
contact
with
the
body,
the
initial
bundle
of
articulatory
features
is
removed
from
the
stem
and
inserted
into
final
position
in
the
frame.
The
remainder
of
the

Phonological
Base
phonological information
from
the
verb
stem does
not
appear
in
the
inflected
form.
1 3
The resulting
sign
begins
at
the
location
specified
by
the
inflectional
frame
and
moves
to
what
was
the
original
location
specified
in
the
stem.
Many
details
have
been
left
out of
our
description
of
this
inflection.
In
fact,
three
such
frames
(i.e.,
three
allomorphs)
are
needed
to
account
for
the
U-I
data.
A
fuller
account
can
be
found
in
Liddell
(1984b).
There
are
a
number
of
other
ASL
inflections
which
will
naturally
lend themselves
to
an
analysis
utilizing
inflectional frames.
5.2.2
Reduplication.
Reduplication
is
common
in
ASL.
Habitual
aspect and
iterative
aspect
are
each
marked
in
ASL
by
a
different
type
of
reduplication
rule.
14
Figure
43
illustrates
the
form
of
the
verb
LOOK
along
with
its
habitual
and
iterative
forms.
For
purposes
of
our
discussion,
we
will
use
the
verb
stem
ASK,
described
earlier,
and
its
habitual
and iterative
forms.
The shape
of
the
movement
of
these
forms
is
the
same
as
that
seen
in
Figure
43.
The
verb
stem
ASK
is
an
incomplete
S-morph.
It
has
phonological
cells
which
are
filled with
subject
and
object
agreement morphemes.
13
We
hesitate
to
talk
about
"deletion"
here
since this
constructive process
may
take
place
within the
lexicon.
If
so,
then the
process
merely
copies
(reads,
selects)
specific
information from
the
lexical entry
of
the
stem
and
there
is
nothing
to
delete.
14
The
data
on
these
aspects
are
from
Klima and
Bellugi
(1979),
who
first
described
them.
SLS
64

Liddell
&
Johnson
43a
LOOK
43b.
LOOK
(habitual)
43c.
LOOK
(iterative)
Figure
43
The
habitual
and
iterative
forms
of
LOOK
SLS
64

Phonological
Base
str
IIllNl
I
lo-
I
_
_10_
IBK
PA
P P I
_mR2
STI
_|_mL21ST|
| PA | B I
I
mL2
STI
_I
R21STI
I BA I BA |
| HP I HP |
Figure
44.
ASK
After
the
subject and
object
agreement morphemes
are
inserted,
the
phonological
structure
of
the
stem
is
complete.
Habitual
aspect
is
then
marked
for
the
verb
ASK
through
the
application
of
a
reduplication
rule
like
the
following:
Habitual
Aspect
Rule: 1 2
3
-1 2
3
1 2 3 1 2 3 1 2 3
(for HMH
signs)
srt srt srt
srt
The
rule produces
four
copies
of
the
verb
stem
and
shortens
each
of
the
movements
(srt).
15
The
application
of
this
rule creates
the
environment
for
the
M-epenthesis
rule
described
under
phonological
processes
above.
15
The
actual
number
of
repetitions
can
vary.
For
example,
it
could
easily
be
produced
with
three
rather
than
four
repetitions.
265
SLS
64

Liddell
&
Johnson
H M H M H M H M H M H M H M
1 1 II
li
i
l II II II
II
lI
il II II I 1 II Ii
II
1 2 3 1 2 3 1 2 3 1 2 3
Figure
45
Habitual
form
after
application
of
M
epenthesis
The
circled M's
are
inserted
between
the
final
H of
one
repetition
and
the
initial
H of
the
next
by
the
M-epenthesis
rule. Because
none of those H's
are
attached
to
articulatory
bundles
specified
for
body
contact,
the
H-
deletion rule
applies. It
deletes
every
H
except
for
the
first
and
the
last,
producing
the
following
structure.
H M M M M M M M H
I l
lI
II 1 1
II
II i 1
II
II I_
L
III I
L
I
L
I I
1 2 3 1 2 3 1 2 3 1 2 3
Figure
46.
Surface form
of
ASK
(habitual
aspect)
The
epenthetic
M's
and
the
feature
bundles
attach
as
shown
above. This
produces
what,
for
ASL,
is
a
relatively long
word
consisting
of
nine
segments.
A
different
and
slightly
more
complicated
reduplication
rule
could
have
applied, producing
the
iterative
aspect.
Iterative
ruLe:1
2 3 -1 2 3 M 1 2 3 M 1 2 3
Long arc Long
arc Long
The
application
of
this
rule
to
ASK
will
produce
the
following
structure.
SLS
64
266

Phonological
Base
H M H M H M H M H M H
strlonac str Long
arc
str Long
1 1 1 1 I I I
I
I I l I
I I I I I 1 I I I I l I
1 2 3 1 2
3
1 2 3
Figure 47. Result
of
application
of
Iterative
Rule
to
ASK
In
this case
the
M-epenthesis rule
will
not
apply
because
the
reduplication
rule
itself
has
already
inserted
a
particular
type
of
M
(with
the
feature
'arc')
between
each
repetition
of
the
stem.
The
rule
has
also
marked
some of
the
H's
with
the
feature
[long],
which
prohibits
application
of
the H-deletion
rule. The
H-deletion
rule
may
apply
to
unlengthened
H's,
however.
Its
application produces
the
structure
in
Fig.
48.
H M H M M H M M H
str
tong
arc
str
Long
arc str Long
I I I I I
I
I | | I I I I
I I I I I I l I I I I
I I II
L
I
L
I I
1 2 3 1 2 3 1 2 3
Figure
48.
ASK(iterative
form) after
H
deletion applies
The
application
of
the
Iterative
Rule
has
also
produced
a
rather
long
ASL
sign,
though
its
structure
is
significantly
different
from
that
produced
by
the
Habitual
Aspect
Rule.
We
will
now
summarize
the
morphological
and
phonological
proc-
esses
which
have
interacted
to
form
these
two
forms of
ASK.
Each
began
as
a
phonologically
incomplete
stem.
The
stem
was
made complete through
a
morphological
rule
which
inserts
agreement morphemes
into
the
stem.
The
completed
stem
then
underwent one
of
the reduplication
rules,
which
produced
an
aspectual
inflection.
The
application
of
either
of
the
reduplicative
rules
creates
the
environment
for
the
application of one
or
SLS
64

Liddell
&
Johnson
more
phonological
rules.
The
phonological rules
then
apply
to
produce
the
correct
surface form.
It
has
been
common
practice
in
the
past
to refer to
signs
which
have
undergone
a
reduplication
process
as
being
marked
by
the
phonological
feature
[+
redup]
(Fischer and Gough,
1978;
Supalla
and
Newport,
1978;
Klima
and
Bellugi,
1979;
Padden
and
Perlmutter,
1984).
It
should
be
clear
from
the
two
reduplication
rules
we
have
examined
that
such
an
approach
is
not adequate.
The
two
reduplicated
forms
do
not
differ
from
their
stems
by
the
single
phonological
feature
[+ /-
redup].
They
have
undergone
a
reduplicative
process
which
copies
phonological
segments, adds
phonological
features,
and
triggers
the
application
of
phonological rules.
5.2.3
Affixation.
Across
spoken
languages,
one
of
the
most common
phonological
means for
marking
the
application
of
a
morphological
process
is
the
affixa-
tion
of one or
more
segments
to
a
stem.
This
also occurs
in
ASL,
but
it
is
uncommon.
The
one
clear
case
is
a
nominalizing
suffix
having
the
structure
MH.
When
suffixed
to
the
verb
TEACH,
it
produces
a
word
meaning
"teacher,"
and, when
suffixed
to
the
noun
LAW,
it
produces
"lawyer."
This
is
the
only ASL
morpheme
we
know
of
which
clearly
has
the
status
of
an
affix.
Most
ASL
morphological
activity
involves
filling
in cells in
phonologi-
cally
incomplete
segments, or
operations
on phonologically
complete
stems,
which
either
modify
them through
the
use
of frames, or
through
some
type
of
reduplicative
process.
6.
Conclusion
Early
in
this
paper
we
suggested
that,
although
the
terminology
of
modern
phonology
would
appear
to eliminate
signed
languages
from
phonological
analysis,
the
concepts
that
underlie
the
terminology are
suffi-
ciently
broad
to
permit
its
application
to
the
levels
of
organization
of
sign
language.
Our
discussion
of
the
phonetic,
phonological,
and
morphological
SLS
64

Phonological
Base
structures
of
ASL has
been
aimed
at
demonstrating
the
often surprising
degree
to
which
both
the
levels
of
organization
and
the
processes and struc-
tures
of
ASL
parallel
those
found
in
spoken
languages.
Thus,
it
should
now
be
possible
to
refer
to
the phonetic
structure,
the morpheme structure
conditions,
or
the
phonological
processes of
sign
languages
and
be
confi-
dent
that
what
is
being described
is
analogous
to
similar
phenomena
in
spoken
languages.
This
potential
for
comparison permits
an
expansion
of
our
knowledge
about
language
universals,
and should
encourage
the
description
of
some
of
the
dozens of
independent
sign
languages
of
the
world.
More importantly,
the unique
lexical
structures
and
morphological
processes
we
have
identified
and described
add
to
our
knowledge
of
the
variety
of
forms of
human
language.
Note:
Many
individuals
have
commented
on earlier
versions of
the
work
presented
here
and
we
have
benefited
from
their
insights.
In
particular,
we
would
like
to thank
Ursula
Bellugi,
Edward
Klima,
Ceil
Lucas,
Carol
Padden,
David
Perlmutter,
William
Stokoe,
the
participants
at
the
1984
Clear
Lake
Confer-
ence
on
Sign
Language
Linguistics,
and
our
students over
the past
three years.
In
addition,
we
would
like
to thank
the
Gallaudet
Research
Institute
for
supporting
the
production
of
the
illustrations
used
in
this paper.
We
consider
the
authorship
of
this
paper
to
be
equal. This
is
an
unrevised
version
of the
original
unpublished manuscript
of
the
same
title
cited
as
Liddell
and
Johnson
(1984)
in
many
recent
papers
on
ASL
phonology.
It is
published
here
to provide access to
that
paper
as
a
stage
in
the
devel-
opment of
ASL
phonological theory.
REFERENCES
Battison,
R.
1974.
Phonological
deletion
in
American
Sign
Language.
Sign
Language
Studies
5.1-19.
1978.
Lexical
Borrowing
in
American
Sign
Language.
Silver
Spring,
MID:
Linstok
Press.
SLS
64

Liddell
&
Johnson
Chomsky,
N. & M.
Halle.
1968.
The
Sound
Pattern
of
English.
New
York,
NY:
Harper
and
Row.
Fischer,
S. & B.
Gough.
1978.
Verbs
in
American
Sign
Language. Sign
Language
Studies
18.17-48.
Frishberg,
N.
1975.
Arbitrariness
and
iconicity:
historical
change
in
American
Sign
Language.
Lg.
51.676-710.
Goldsmith,
J.
1976.
Autosegmental
phonology.
MIT
dissertation.
Johnson,
R. & S.
Liddell.
1984.
Structural
diversity
in
the
American
Sign
Language lexicon.
In
Papers
from the
Parasession
on
Lexical
Semantics,
Testen, Mishra,
&
Drogo
eds.
Chicago,
IL:
Chicago Linguistic
Society.
173-186.
Klima,
E. & U.
Bellugi.
1979.
The
Signs
of
Language.
Cambridge,
MA:
Harvard
University
Press.
Liddell,
S.
1977.
An
investigation
into
the
syntactic
structure of
American
Sign
Language.
UCSD
dissertation.
1984a.
THINK
and
BELIEVE:
Sequentiality
in
American
Sign
Language signs.
Lg.
60.372-399.
270
SLS
64

Phonological
Base
1984b.
Unrealized
inceptive
aspect
in
American
Sign
Language:
feature insertion
in
syllabic
frames.
In
Papers
from
the
Twentieth
Regional
Meeting
of
the
Chicago
Linguistic
Society,
Drogo, Mishra,
&
Testen
eds.
Chicago,
IL:
Chicago
Linguistic
Society.
257-270.
---------
,
C.
Ramsey,
F.
Powell,
& D.
Corina.
1984.
Numeral
incorporation
and numeral
classifiers
in
American
Sign
Language.
Gallaudet
University,
ms.
McCarthy,
J.
1979.
Formal
problems
in
Semitic
phonology
and
morphology.
MIT
disserta-
tion.
Padden,
C. & D.
Perlmutter.
1984.
American
Sign
Language
and
the
architecture
of
grammatical
theory,
ms.
Stokoe,
W.,
Jr.
1960.
Sign
Language
Structure:
An
Outline
of
the
Visual
Communication
Systems
of
the
American Deaf.
(Studies
in
linguistics:
occasional
papers
8)
Buffalo,
NY:
University
of
Buffalo.
[Rev.
edn., Silver
Spring,
MD:
Linstok
Press,
1978].
---------
,
D.
Casterline;
& C.
Croneberg.
1965.
A
Dictionary
of
American
Sign
Language
on
Linguistic
Principles.
Wash-
ington,
DC:
Gallaudet College
Press. [Rev. edn., Silver
Spring,
MD:
Linstok
Press,
1976.]
Studdert-Kennedy,
M. & H.
Lane.
1980.
Clues from the
differences
between
signed
and
spoken
language.
In
Signed
and
Spoken
Language:
Biological
Constraints
on
Linguistic
Form,
SLS
64

272
Liddell
&
Johnson
SLS 64
Bellugi
&
Studdert-Kennedy
eds.
Weinheim:
Verlag Chemie.
29-39.
Supalla, T.
1978
Morphology
of verbs
of
motion
and
location
in
American
Sign
Language.
In
Papers
from
the
Second
National
Symposium
on
Sign
Language
Research
and
Teaching. Caccamise
&
Hicks eds.
Silver
Spring,
MD:
NAD.
27-45.
---------
&
Elissa
Newport.
1978.
How
many seats
in a
chair? The
derivation
of nouns
and
verbs
in
Amer-
ican
Sign
Language.
In
Understanding Language
Through
Sign
Language
Research,
Siple
ed.
New
York:
Academic
Press.
91-132.

273
Phonological
Base
SLS
64
Appendix
A.
Selected
signs
produced
with
an
extended
index
finger.
Handshake
Sign
LICENSE
LAZY
CORRECT
21
SUNGLASSES
UNDERSTAND
DON'T-HAVE-TO
RUN
WRITE
BLACK
BIRD
LATER
QUICKLY
BIG

Liddell
&
Johnson
Appendix
B.
Inventory
of
ASL
Body
Locations
and
Representative
Signs,
ipsiLateraL
ipsitateraL
center
tine
center
line
peripheral
peripheral
BH |
iBH
PONYTAIL
j
PIGTAILS
TH I
BALD
FHt P~tI I I
HEIGHT jI________
______
FH iFH SF
iSF
FATHER
THINK
DEER
HAIR
FHb
INDIA
NSt iNSt
CKt
ERt
STRICT
CRY
CHINESE
IHEARINGAID
I
NS INS CK ER
CLOWN BORED SEE HEAR
NSb
iNSb
CKb
ERb
KID
THIEF
HOME
(Last
seq) I
CALIFORNIA
MO iMO
SUCKER HOOKED
LP
iP
JW
HEARING DROOL HOME (first seq) I
CII iCR
EAT
RESTAURANT
CNb
NK INK | SH
iSH
CURIOUS
CONVINCE
BURDEN
REMIND-ME
ST | SHb
COUGH I
VACATION
CH iCH UAt iUAt
FINE
CONGRESS
POWER
SCOTCH
TR iTR UA iUA
UPSET
MONKEY
COKE
GARBAGE-CAN
AB AB UAb iUAb
PREGNANT RUSSIA
DRUGS
POOR
LGt | FAt iFAt
NAVY TABLE j
PUNISH
LG FA iFA
BIKINI-BOTTOM
MUSIC
IMPROVE
LGb FAb
iFAb
j
DOG
DOCTOR
I DUTY
SLS 64

275
Phonological
Base
SLS
64
Appendix
C.
Selected
Locations
in
Signing
Space
and
Representative
Signs.
I----------I
TH/FH
d
m
CLOUDS
I
--------
p
GIRAFFE
(last
seg)
CN/NK
d
-----------
m
ETCETERA
(1st seg)
I ---
-- I
---------
DEEP-FUTURE
(Last seg)
---
- --- --
--
----- - -
DREAM
HEAVEN
(Last seg) (Last
seg)
------------
I - --
--------
I
INVENT
BECAUSE
(Last seg) (Last
seg)
FUTURE
(Last seg)
------- ----I ----- --- - -
ETCETERA
#GO
(Last seg) (Last seg)
--
I-----------
p
|
FASCINATED
THRILL
j(seg
3, str
hd)
(Last
seg)
I
===========================
ST/CH d
.-----
------- -
m FAMILY MONDAY
(Last seg)
------------
I ----- ---
p
FAMILY
HAMMER
(1st seg) List
seg)
S-==
-=
-I --
======-====
--------------
I
HEALTHY
(1st seg)
= ============
-------------
I
FAR-OUT
(Last seg)
---
----
--- --
AGENT(suffix)
(1st seg)

Liddell
&
Johnson
Appendix
D.
Locations
on
the
Weak
Hand and
Representative Signs.
HAND FINGERS THUMB INDEX MIDDLE RING
LITTLE
I ---
--------
I I - I
I---------------I
INI PA INFI INXF INMF INRF INLF
OWE
REJECT
JR-LV
SOPH-LV
FR-LV
PR-LV
I
-- - - - -- - ---
I
-I I- - - - | - - - I - I- -- -- I
PD PDFI PDTH
PDXF
PDMF PDLF
EXCUSE
FIRST
SECOND
THIRD FIFTH
I- ---I-|------
-
-1----
I - I
---- I
BKI BK BKFI BKTH BKRF
HARD
CHAIR
REMEMBER
ENGAGED
RAI RA RAFI
RAXF
RAMF RARF RALF
ENOUGH
NAME [MONTH-AFTER-MONTHI LAST
|- | -- ---
I-
- - - -- - - -
-|---------
---
I
ULI UL ULFI ULTH
HELP
KICK
GIVE-RIDE
TI TIFI
TITH
TIXF TIMF TIRF TILF
TOPS
SR JR SOPH FROSH PREP
- -- I
-I------ I
I-1--------
I
KN TIKN
GAME
..-....
I-----------I
------
---
BA| BA
FOLLOW
|I-
- I
1-I-------
I --- I -- I
HLI HL
I PAPER
1-I --- I ---
1- I I ----
WBI W8FI
WBTH
WBXF
WBMF
WBRF
PREGNANT BOTHER
START
HALFTIME
3rd-QUAR-BREAK
I -... ---- 1 -
---- -- I I ---
I------
I
Key:
IN inside;
PD
pad;
BK back;
RA radiaL;
UL uLnar; TI tips;
KN
knuckle;
BA base;
HL heel; WB web;
PA paLm;
JR-LV
'leave
during
junior year'
SLS
64

Phonological
Base
Appendix
E.
Strong
Hand,
Contacting
Handparts
and
Representative
Signs.
hand
fingers
forearm
thumb
---------
--
------------ ------------
I
inside
PA INFI INFA
j
INTH
PRESSURE
WARN PROMISE
J
NOT
-------------
-------------
-------
-----------
pad PDFI j PDTH
KNOW REMEMBER
.---------
1-
I
I - -
----------
back BK
BKFI BKTH
GRAD
STUDENT
PROVE SUFFER
I -
---------
I ----
----
I
- - - - - -
--- -- - - -
radial RA RAFI RATH
COUGH
BLACK
GIRL
----------
------------
--------
I--------
-
uLnar UL ULFI |
ULFA
ULTH
STOP NAME
DEFEND
FARMER
---.-- -- ---- -- -------------
tip J TIFI TITH
|
AGAINST
SHAVE
-- -
I-III---------
1
-------
I---------I
knuckle
KN KNFI |
GAME FOOL
-----------------
---------------
------------
I
base
BA BAFI
BAFA
BATH
CANDLE
PANT
TREE
TURTLE
-----------------
---------------
------------
I
heel HL
CHEESE
---
--
------
-------------
I -- - -
--
-----
web WBFI WBTH
TICKET
j
BRACELET
-- - -- - - I - - - - - ------- - -- - -- - -
Scott
K.
Liddell
is a
Professor
in
the
Department
of
Linguistics
and
Interpreting
at
Gallaudet
University.
He has
a Ph.D.
in
linguistics
from
the
University
of
California
at San
Diego.
For
the
past
fifteen
years
his
research
has
focused
on
the
linguistic description
of
American
Sign
Language.
His initial
work
dealt
with
ASL
syntax.
During
the past
six
years
the
focus
of
his
research
has
been
the
development
of
a
segmental
description
of
ASL
signs
and
ASL
morphological
processes.
Robert
E.
Johnson
an
anthropologist
and
linguist
interested
in
the study
of sign languages
and their
place
in
deaf
communities.
He
earned
his
Ph.D.
in
Anthropology
from Washington
State
University
in
1975.
He
is
currently
the
chair
of
the Department
of
Linguistics
and
Interpreting
at
Gallaudet
University.
His
most recent research interests
include
the
development
of
linguistic
notation
systems
for
tran-
scribing
sign
languages,
ASL
verb
morphology,
the
acquisition
of
ASL
by
deaf
preschool
children,
and
the
sign
language
of
deaf
and
hearing
inhabitants
of
a
Yucatec
Maya
village
in
Mexico.
SLS
64