Professional Documents
Culture Documents
|
Q.
W
|
'o
E
E
z:
-z.
225-
200-
175-
150-
125-
100-
75-
50-
25-
0-
o 10th percentile
-o--- 25th percentile
- - - o- - - 50th percentile
- - - A- - - 75th percentile
- - -a- - - goth percentile
..--""
"' ' A- " ' '
f f l - . - - - - "
A
Pt:-:-;-:>-:---------fi--- ----"-,- - - - - ra-.:.:.:.:.:.:. ,
1-20 21-50 51-100 101-150 151-200 >200
Number of words comprehended
Figure lO.ib Variabilily in word production as a funct i on of Comprehension vocabuhiry size.
A
l
t
h
o
u
g
h
t
h
i
s
c
o
m
p
a
r
i
s
o
n
d
o
e
s
s
u
g
g
e
s
t
t
h
a
t
a
c
o
m
m
o
n
g
r
o
w
t
h
f
u
n
c
t
i
o
n
i
s
o
b
s
e
r
v
e
d
m
b
o
t
h
c
r
o
s
s
-
s
e
c
t
i
o
n
a
l
a
n
d
l
o
n
g
i
t
u
d
i
n
a
l
d
e
s
i
g
n
s
.
w
e
a
r
e
s
t
i
l
l
l
o
o
k
i
n
g
a
t
g
r
o
u
p
d
a
t
a
i
n
m
e
n
p
(ji
f
3
o;
S
~
g
' X
c~ ^*-
o:
72.
K
c l
g: c'
!A O'
3_ .
? ~
.
c
<<;
:.
-.
r:
O
?
o
S o "
S M ni
&
'-3
Si & w TO
M 3 3 "
fl: (U E C
5 ~" ^
o ' 5'
" 3 E.
=="- ff
"g ~ 3
^ - " & _
" ! $
~~ ^' k-5
H |f ^
p E' r l
T s" 3 E"
' i s a
TO 3 g
2 "S- 5
8 ;
< ~ ru
S.^ 3
S. 3 3
o' gj
3
K
S-
rf- "O J.
3- re cr
o: ^ *<
< S s.
1 o 3
II f
2 B: w
i 3 -i
K> TO T
|| |
n> Q
n 3" 3
2 >T
1 f^
C n-
1
v 1
m -n
O 7
ai
T _
Q
O
K> ~
O
ro
< 2
|r
2
N' i -
CD .
O
O
&
o
1 ~
S
o
ai
0
o
V
g -
o
j
\
\
c
g
r
a
m
m
a
r
w
i
t
h
i
t
s
e
l
f
!
T
o
c
o
n
l
r
o
l
f
o
r
t
h
i
s
p
o
s
s
i
b
i
l
i
t
y
,
w
e
r
e
c
a
l
c
u
l
a
t
e
d
t
o
t
a
l
v
o
c
a
b
u
l
a
r
y
s
i
z
e
f
o
r
t
h
e
f
u
l
l
M
a
c
A
r
t
h
u
r
s
a
m
p
l
e
,
s
u
b
t
r
a
c
t
i
n
g
o
u
t
g
r
a
m
m
a
t
i
c
a
l
f
u
n
c
t
i
o
n
w
o
r
d
s
f
o
r
e
a
c
h
i
n
d
i
v
i
d
u
a
l
c
h
i
l
d
.
F
i
g
u
r
e
1
0
.
4
i
l
l
u
s
t
r
a
t
e
s
t
h
e
r
e
l
a
t
i
o
n
b
e
t
w
e
e
n
g
r
a
m
m
a
r
a
n
d
v
o
c
a
b
u
l
a
r
y
t
h
a
t
i
s
o
b
s
e
r
v
e
d
w
h
e
n
v
o
c
a
b
u
l
a
r
y
c
o
u
n
t
s
a
r
e
b
a
s
e
d
e
n
t
i
r
e
l
y
o
n
t
h
e
r
e
m
a
i
n
i
n
g
c
o
n
t
e
n
t
w
o
r
d
s
.
T
h
e
n
o
n
l
i
n
e
a
r
f
u
n
c
t
i
o
n
t
h
a
t
r
e
m
a
i
n
s
i
s
,
i
f
a
n
y
t
h
i
n
g
,
e
v
e
n
m
o
r
e
p
o
w
e
r
f
u
l
t
h
a
n
t
h
e
o
r
i
g
i
n
a
l
f
u
n
c
t
i
o
n
w
h
e
r
e
a
l
l
w
o
r
d
s
a
r
e
i
n
c
l
u
d
e
d
i
n
t
h
e
v
o
c
a
b
u
l
a
r
y
t
o
t
a
l
.
A
n
o
t
h
e
r
p
o
s
s
i
b
l
e
o
b
j
e
c
t
i
o
n
t
o
t
h
e
s
e
f
l
n
d
i
n
g
s
r
e
v
o
l
v
e
s
a
r
o
u
n
d
t
h
e
c
r
o
s
s
-
s
e
c
t
i
o
n
a
l
n
a
t
u
r
e
o
f
t
h
e
n
o
r
m
a
t
i
v
e
s
a
m
p
l
e
.
B
e
c
a
u
s
e
t
h
e
f
u
n
c
t
i
o
n
s
i
n
f
i
g
u
r
e
s
1
0
.
1
-
1
0
.
4
a
r
e
c
o
l
-
l
a
p
s
e
d
a
c
r
o
s
s
d
i
f
f
e
r
e
n
t
c
h
i
l
d
r
e
n
a
t
d
i
f
f
e
r
e
n
t
a
g
e
l
e
v
i
s
,
w
e
c
a
n
n
o
t
a
s
s
u
m
e
t
h
a
t
t
h
e
y
r
e
p
-
r
e
s
e
n
t
p
a
t
t
e
r
n
s
o
f
g
r
o
w
t
h
f
o
r
a
n
y
i
n
d
i
v
i
d
u
a
l
c
h
i
l
d
.
I
n
a
m
o
r
e
r
e
c
e
n
t
s
t
u
d
y
(
G
o
o
d
m
a
n
,
1
9
9
5
;
J
a
h
n
-
S
a
m
i
l
o
.
1
9
9
5
;
T
h
a
l
e
t
a
l
.
,
1
9
9
7
)
.
w
e
h
a
v
e
u
s
e
d
t
h
e
M
a
c
A
r
t
h
u
r
C
D
I
t
o
f
o
l
l
o
w
2
8
i
n
d
i
v
i
d
u
a
l
c
h
i
l
d
r
e
n
l
o
n
g
i
t
u
d
i
n
a
l
l
y
,
w
i
t
h
p
a
r
e
n
t
s
f
i
l
l
i
n
g
o
u
t
t
h
e
f
o
r
m
s
o
n
a
m
o
n
t
h
l
y
b
a
s
i
s
f
r
o
m
8
t
o
3
0
m
o
n
t
h
s
o
f
a
g
e
.
F
r
o
m
1
2
-
3
0
m
o
n
t
h
s
,
w
e
a
l
s
o
s
a
w
t
h
e
c
h
i
l
d
r
e
n
m
o
n
t
h
l
y
i
n
t
h
e
l
a
b
o
r
a
t
o
r
y
,
v
i
d
e
o
t
a
p
i
n
g
f
r
e
e
s
p
e
e
c
h
a
n
d
f
r
e
e
p
l
a
y
a
n
d
a
d
m
i
n
i
s
t
e
r
i
n
g
s
t
r
u
c
t
u
r
e
d
m
e
a
s
u
r
e
s
o
f
w
o
r
d
c
o
m
p
r
e
h
e
n
s
i
o
n
.
w
o
r
d
p
r
o
d
u
c
t
i
o
n
,
a
n
d
c
o
m
p
r
e
h
e
n
s
i
o
n
o
f
g
r
a
m
m
a
r
.
T
h
e
r
e
w
e
r
e
s
t
r
i
k
i
n
g
s
i
m
i
l
a
r
i
t
i
e
s
b
e
t
w
e
e
n
t
h
i
s
l
o
n
g
i
t
u
d
i
n
a
l
s
a
m
p
l
e
a
n
d
t
h
e
l
a
r
g
e
c
r
o
s
s
-
s
e
c
t
i
o
n
a
l
s
a
m
p
l
e
f
r
o
m
t
h
e
M
a
c
A
r
t
h
u
r
n
o
r
m
i
n
g
s
t
u
d
y
n
b
o
t
h
t
h
e
m
e
a
n
a
n
d
t
h
e
r
a
n
g
e
o
f
v
a
r
i
a
t
i
o
n
o
b
s
e
r
v
e
d
a
c
r
o
s
s
a
g
e
l
e
v
i
s
n
w
o
r
d
c
o
m
p
r
e
h
e
n
s
i
o
n
,
w
o
r
d
p
r
o
-
d
u
c
t
i
o
n
,
a
n
d
t
h
e
d
e
v
e
l
o
p
m
e
n
t
o
f
g
r
a
m
m
a
r
.
F
i
g
u
r
e
1
0
.
5
c
o
m
p
a
r
e
s
t
h
e
n
o
n
l
i
n
e
a
r
f
u
n
c
-
t
i
o
n
l
i
n
k
i
n
g
s
e
n
t
e
n
c
e
c
o
m
p
l
e
x
i
t
y
a
n
d
v
o
c
a
b
u
l
a
r
y
s
i
z
e
i
n
t
h
e
r
e
s
p
e
c
t
i
v
e
c
r
o
s
s
-
s
e
c
t
i
o
n
a
l
Mean complexity
- i . - t - i - ' - t i o r o i v > i v > r o u u u u u g
1 ! 1 1 1 1 1 1 1 I 1 1 1 1 1 I 1 I
u
O
4-
! I n
t V i
'' \ 3
o:
i 3
1 | 1 1
\ S & s-
\\ - x
'. \ g s
\\ ~ S 4?
\ =P <D
'. \ C 3
\ \ B
a |
\ \ 0 g
''. \ o S
\ o-
\ O:
c
\ a 0)
<>. El -3
\^
'-.. ^\ "
"x ^v c"
' ' ' " N. ,
'--<X |
i
s
l
u
r
k
i
n
g
b
e
n
e
a
t
h
t
h
e
s
u
r
f
a
c
e
.
A
f
t
e
r
a
l
l
,
t
h
e
v
o
c
a
b
u
l
a
r
y
c
h
e
c
k
l
i
s
t
i
n
c
l
u
d
e
s
g
r
a
m
m
a
t
i
-
c
a
l
f
u
n
c
t
i
o
n
w
o
r
d
s
l
i
k
e
p
r
e
p
o
s
i
t
i
o
n
s
,
a
r
t
i
c
l
e
s
,
a
u
x
i
l
i
a
r
y
v
e
r
b
s
,
p
r
o
n
o
u
n
s
.
a
n
d
c
o
n
j
u
n
c
-
t
i
o
n
s
.
P
e
r
h
a
p
s
a
l
l
t
h
a
t
w
e
r
e
a
l
l
y
h
a
v
e
i
n
f
i
g
u
r
e
1
0
.
2
i
s
a
t
a
u
t
o
l
o
g
i
c
a
l
r
e
l
a
t
i
o
n
o
f
O K>
o
1 ^
O
UI
| _
O
Q
O
1 -
1
N)
O
O
O
0
t "
T -
g
o
01 _
o
o
1
i>
t
1
\
G
i
v
e
n
t
h
e
p
o
w
e
r
o
f
t
h
i
s
r
e
l
a
t
i
o
n
s
h
i
p
,
w
e
m
i
g
h
t
s
u
s
p
e
c
t
t
h
a
t
a
n
o
t
h
e
r
k
i
n
d
o
f
a
r
t
i
f
a
c
t
j
^
i
t
c
o
m
p
l
e
x
f
o
r
m
s
.
H
e
n
e
e
t
h
e
f
u
n
c
t
i
o
n
i
n
f
i
g
u
r
e
1
0
.
2
f
o
l
l
o
w
s
c
h
i
l
d
r
e
n
r
i
g
h
t
i
n
t
o
t
h
e
v
e
r
y
h
e
a
r
t
o
f
g
r
a
m
m
a
t
i
c
a
l
d
e
v
e
l
o
p
m
e
n
l
,
w
h
e
n
p
r
o
d
u
c
t
i
v
e
c
o
n
t
r
o
l
o
v
e
r
c
r
u
c
i
a
l
m
o
r
p
h
o
l
o
g
i
-
c
a
l
a
n
d
s
y
n
t
a
c
t
i
c
s
t
r
u
c
t
u
r
e
s
i
s
w
e
l
l
u
n
d
e
r
w
a
y
(
B
r
o
w
n
.
1
9
7
3
)
.
W
e
a
l
s
o
n
o
t
e
t
h
a
t
t
h
i
s
p
o
w
e
r
f
u
l
f
u
n
c
t
i
o
n
i
s
n
o
t
a
n
a
r
t
i
f
a
c
t
o
f
a
g
e
.
b
e
c
a
u
s
e
i
t
r
e
m
a
i
n
s
v
e
r
y
s
t
r
o
n
g
w
h
e
n
a
g
e
i
s
p
a
r
t
i
a
l
e
d
o
u
t
o
f
t
h
e
c
o
r
r
e
l
a
t
i
o
n
(
F
e
n
s
o
n
e
t
a
l
.
.
]
9
9
4
)
.
I
n
d
e
e
d
.
a
g
e
i
s
a
s
u
r
p
r
i
s
i
n
g
l
y
p
o
o
r
p
r
e
d
i
c
t
o
r
o
f
b
o
t
h
v
o
c
a
b
u
l
a
r
y
a
n
d
g
r
a
m
m
a
r
w
i
t
h
i
n
t
h
i
s
1
6
-
3
0
-
m
o
n
t
h
w
i
n
d
o
w
.
f
o
r
t
h
i
s
l
a
r
g
e
s
a
m
p
l
e
o
f
h
e
a
l
t
h
y
E
n
g
l
i
s
h
-
s
p
e
a
k
i
n
g
c
h
i
l
d
r
e
n
.
T
a
k
e
n
t
o
g
e
t
h
e
r
,
a
g
e
a
n
d
v
o
c
a
b
u
l
a
r
y
s
i
z
e
a
c
c
o
u
n
t
f
o
r
7
1
.
4
%
o
f
t
h
e
v
a
r
i
a
n
c
e
i
n
g
r
a
m
m
a
t
i
c
a
l
c
o
m
p
l
e
x
i
t
y
.
W
h
e
n
a
g
e
i
s
e
n
t
e
r
e
d
i
n
t
o
t
h
e
e
q
u
a
t
i
o
n
a
f
t
e
r
v
o
c
a
b
u
l
a
r
y
s
i
z
e
i
s
c
o
n
t
r
o
l
l
e
d
.
i
t
a
d
d
s
a
s
t
a
t
i
s
t
i
c
a
l
l
y
r
e
l
i
a
b
l
e
b
u
t
e
x
c
e
e
d
i
n
g
l
y
s
m
a
l
l
0
.
8
%
t
o
t
h
e
t
o
t
a
l
v
a
r
i
a
n
c
e
a
c
c
o
u
n
t
e
d
f
o
r
.
H
o
w
e
v
e
r
,
w
h
e
n
v
o
c
a
b
u
l
a
r
v
s
i
z
e
i
s
e
n
t
e
r
e
d
i
n
t
o
t
h
e
e
q
u
a
t
i
o
n
a
f
t
e
r
a
g
e
i
s
c
o
n
t
r
o
l
l
e
d
,
i
t
a
d
d
s
a
w
h
o
p
p
i
n
g
3
2
.
3
%
t
o
t
h
e
v
a
r
i
a
n
c
e
i
n
g
r
a
m
m
a
r
s
c
o
r
e
s
.
(
F
o
r
s
i
m
i
l
a
r
r
e
s
u
l
t
s
i
n
I
t
a
l
i
a
n
,
s
e
e
C
a
s
e
l
l
i
e
t
a
l
.
,
t
h
i
s
v
o
l
u
m
e
.
)
Grammatical complexity scores
0 5 0 o o M r o S o w ^ a i m o M
i i i i i i i i t i i i i i
| i
9 <>
\^ +
v> en
CD CD
*
i>
%.
%s
%,,
%;,
^%x
^%x
"^SJx
TV>
o
f
t
h
e
i
r
l
a
n
g
u
a
g
e
b
y
3
-
3
.
5
y
e
a
r
s
o
f
a
g
e
.
i
n
c
l
u
d
i
n
g
p
a
s
s
i
v
e
s
,
r
e
l
a
t
i
v
e
c
l
a
u
s
e
s
,
a
n
d
o
l
h
e
r
"
j
i
1
S
ro
IB
8
-g^
CD
X
"*
m
a
l
l
y
d
e
v
e
l
o
p
i
n
g
c
h
i
l
d
r
e
n
a
r
e
a
b
l
e
t
o
p
r
o
d
u
c
e
m
o
s
t
o
f
I
h
e
b
a
s
i
c
g
r
a
m
m
a
l
i
c
a
l
s
t
r
u
c
t
u
r
e
s
^ j
T
o
u
n
d
e
r
s
t
a
n
d
t
h
e
r
e
l
e
v
a
n
c
e
o
f
t
h
i
s
f
i
n
d
i
n
g
.
i
t
i
s
i
m
p
o
r
t
a
n
t
t
o
k
e
e
p
i
n
m
i
n
d
t
h
a
t
n
o
r
-
1
4
2
I
n
t
r
o
d
u
c
t
i
o
n
t
o
W
o
r
d
L
e
a
m
i
n
g
K
3
744
Introduccin to Word Learnina
o.
i
01
w
0
O
)
>%
i
S
o
"5.
1
to
o
<5
m
O
36-
34
:
32-
30
:
28-
26:
24-
22
:
20:
18 :
1 K
\ D
14:
12:
10:
8:
6:
4 :
2:
o-
1
-
Figure 10
50 100 150 200 250 300 350 400 450 500 550 600 650 700
Vocabulary size
,6 Relationship between grammatical complexity & vocabulary size for one late
talker and one early talker.
both cases (i.e., results collapsed over many different children at each data point). We
might therefore ask whether the growth curves in the longitudinal study look similar
for individual children, or whether the commonalities in figure 10.5 represen! group
trends that mask sharp dissociations in at least some individual cases. To address this
issue. we graphed the individual grammar-on-vocabulary functions for each of the 2 7
children. Results suggest a remarkable degree of similarity between these individual
growth curves and the range of curves (from the 1 Oth to the 90th percentile) summa-
rized at the group level in figure 10.3a. To Ilstrate this point. we present the grammar-
on-vocabulary functions for two individual children in figure 10.6, for each session
between 16 and 30 months. These children were selected to represent extremes in rate
of vocabulary growth, including one very late talker and one very early talker. Even
though the two children overlap perfectly in age, there was absolutely no overlap in
their vocabulary size across this period: the late talker produced only 2 72 words by the
last session at 30 months, while the early talker already had a vocabulary of 315 words
at 17 months, when administraron of the grammar scales began. It is clear from figure
10.6 that both children are making progress in grammar that is directly commensu-
rate to their lexical abilities. even though they reach their respective grammar-on-
vocabulary levis at widely different ages withi n this period of development.
We have also had an opportunity to compare the grammar-on-vocabulary functions
for two languages. Hnglish and It al i an (Caselli et al.. 19991. In both languages. the
On the Inseparablity of Grammar and the Lexicn
O English complexity
Italian complexity
145
<50 51-100 101-200 201-300 301-400 401-500 501-600
Vocabulary level
>600
Figure 10.7 Grammatical complexity as a function of vocabulary size in English vs. Italian.
grammar scale of the CDI contained 3 7 pairs of sentences designed to pick up contrasts
that are known to develop in that language between 17 and 30 months. However,
the conten of the respective grammar scales is entirely different (i.e., they are not
translations). Despite these differences, figure 10.7 shows that the same lawful and
powerful function is observed in two languages that differ markedly in their grammati-
cal structure.
We are convinced by these data that there is a powerful link between grammar and
lexical growth in this age range, a nonlinear growth function that holds for both cross-
sectional and longitudinal designs, at both the individual and the group level. These
results (even for individual children) are collapsed across a range of different gram-
matical structures. What does the relationship look like when we look at specific aspects
of the grammar? Presumably, because we know that different grammatical structures
come in at different points within this developmental window, we ought to expect indi-
vidual forms to display different degrees and (perhaps) different types of "lexical depen-
dence." For example. individual grammatical structures might require a different
"critical mass" across the whole vocabulary, or they might require a critical number of
lexical tems within a specific class.
Some insights into this issue come from Marchman and Bates (1994). who used the
MacArthur norming data to investgate the relationship between the number of verbs
that chiidren use and their progress on the verb morphology subscales on the CDI (i.e..
the checklists of irregular, regular and overregularixes forms noted above). This study
was motivated by the ongoing controversy between the connectionist "single mecha-
nism" approach to grammatical development (fi l man, t hi s volume: Klman et al.. 1 996:
MacWhinney. 1993; Plunketl & Marchman, 1993) and the symbolic "rote vs. rule"
146
Introduccin to Word Learning
Stem + corred irregular
O 1-9 10-1920-2930-3940-4950-5960-6970-7980-8990-99100+
Size of verb vocabulary
Figure 10.8 Reported production of 16 irregular verbs as a function of vocabulary size (from
Marchman & Bates. 1994).
approach(Brown. 1973; Marcusetal., 1992;Pinker, 1991). Connectionist simulations
of the acquisition of past-tense morphology have shown that a single mechanism
can display many of the same behaviors that characterize grammatical learning in
children. These include an initial stage in which the system produces high-frequency
irregular past-tense forms correctly (e.g., CAME, WENT), a lengthy intermedate phase
m which these corred forms coexist with occasional overregularizations (e.g., COMED.
GOED). followed by an asymptotic convergence on the use of correct regulars and irreg-
ulars. However. as critics of the single-mechanism account have noted (e.g., Marcus
et al., 1992), these changes are linked to the size and composition of verb vocabulary
(e.g.. high-frequency irregulars tend to domnate in the early stages; the appearance of
overregularizations correlates with an increase in the proportion of regulars to irre-
gulars in the networks "vocabulary"). This raises an important question: is past-tense
learning in real chldren also tied to changes in vocabulary size?
Figure 10.8 from Marchman and Bates iliustrates the relation between number of
verbs in the childs vocabulary (based on the subset of verbs that are used in the vocabu-
lary checklist and n the past-tense scale) and three forms of past-tense marking- xero
stem (the child is reported to use the verb in the dtation form only). correct irregulars
On the Inseparability of Grsmmar and the Lexicn 147
and ncorrect overgeneralizations. This figure reveis a strong nonlinear relatonship
between verb vocabulary size and successive phases in the developrnent of past-tense
morphemes. similar to the relationship observed in connectionist simulations of this
learning process.
This demonstration of a link belween verb vocabulary and past-tense morphology
is the only example we have right now of a link between specific grammatical struc-
tures and their requisite "critical mass" of lexical tems. A great deal more work could
be done n this rea, to determine the lexical prerequisites (if any) for specific gram-
matical forms. which may vary substantially within and across languages. For now. we
turn to a different set of questions. Is the powerful grammar-vocabulary relationship
universal for anyone acquiring English. or can we tind atypical populations in which
dissociations between lexical and grammatical developrnent are observed?
II. Grammatical Development and the Lexicn in Atypical Populations
Although grammatical developrnent invariably follows a stage in which single words
are acquired, the findings reviewed so far suggest that, aside from this temporal lag.
there is no dissociation between grammar and lexical development in normal children.
In view of the claims that have been made about the dissociability of these domains in
adults. it seemed important to us to determine whether there are any individual
children or any specific pediatric populations who deviate from the function displayed
in figures 10.2-10.6.
In fact. the literature on older children with language disorders gives us reason to
expect selective impairments in early grammar. Specific Language Impairment or SLI
is defined as a delay in expressive language abilities that is at least 1 standard deviation
below the mean for the child's chronological age. in the absence of mental retardation,
frank neurological impairment. social-emotional disorders (e.g., autism), or any other
serious biomedical risk factors that could account for the delay. Although the definition
of SLI presupposes that language can be dissociated from the rest of cognition, the
speciicity of Specific Language Impairment is still controversial (Bishop, 1997; Gopnik.
1990: Leonard, 1998). For example. many investigators report that children with SLI
score significantly below age-matched controls on at least some nonlinguistic mea-
sures, including mental imagery. symbolic play. shifting attention. and the ability to
detect rapid temporal changes in auditory stimuli (see Bishop and Leonard for reviews).
There is considerably more agreement about the nature of the language impairment
in SLI. Within every linguistic domain that has been studied to date, the expressive
and/or receptive abilities of children with SLI are qualitatively similar to those of
younger normal children. but delays seem to be greater for grammar.
The vulnerability of grammar in SLI has led to some radical proposals. For example.
Gopnik (1990) and Pinker (1991) described a family with what appeared to be a con-
genital versin of SLI, manifested in a specific dissociation between regular and irregu-
lar grammatical morphology, with relative sparing of irregular forms. They suggested
that such a dissociation is possible because irregulars are processed in trie lexicn, vvhile
regulars are handled by a seprate grammatical processor. However. a more compre-
hensive study of the same family by Vargha-Khadem et al. (1995) has shown that the
I4c9 Introduccin to Word learning
affected members of this family are equally impaired on both regular and irregular mor-
phemes, as well as on a host of other language and nonlanguage measures. Marchman
et al. (1995) have investigated the proposed regular/irregular past-tense dissociation
in a large sample of children with SU. They report significant impairmcnts in past-tense
morphology for SLI children compared with age-matched controls. However, they find
no evidence whatsoever for a dissociation between regular and irregular forms. 'The
errors produced by children with SLI are quite similar to those of younger normal
children - there are just more of them (especially omissions).
Regardless of one's position on this question, we shall see later on that the selective
vulnerability of grammatical morphology is not specitic to SLI. It has been observed in
a number of other child and adult populations, including children with Down Syn-
drome (Chapman, 1995; Contardi & Vicari, 1994). For present purposes. the point is
that a certain degree of dissociability has already been observed between grammatical
morphology and other aspects of language in older children and adults with language
impairments. What we want to determine next is whether analogous patterns of dis-
sociation are observed in the period when grammar first develops.
Late ana early talkers
Within a larger program looking for patterns of association and dissociation within and
across linguistic and cognitive domains. Thal and her colleagues have examined pat-
terns of lexical and grammatical development in infants and preschool children at the
extreme ends of the normal distribution (Thal, 1991; Thal et a!., 1996). This includes
"late talkers." defined as children in the bottom lOth percentile for expressive vocabu-
lary between 18 and 24 months of age. in the absence of the same exclusionary factors
that are used to diagnose SU in older children. It also includes "early talkers," defined
as children in the top lOth percentile for expressive vocabulary between 12 and 24
months of age. For our purposes here, we are interested in the relationship between
grammar and vocabulary in these extreme groups. If late talkers constitute an early
variant of SLI. then we might expect to find that grammar lags behind vocabulary level.
compared with children who reach the same vocabulary size closer to the normative
age. Alternatively, we might find some late talkers who have managed to develop gram-
matical abilities well in advance of their lexical levis. The same two extremes may also
be observed among the early talkers: children whose grammatical abilities are still tied
to chronological age. despite their lexical precocity. and children who are "grammar
geniuses." attaining levis of sentence complexity that are even greater than their
abilities in the lexical domain.
These were reasonable hypotheses, but they have not been supported by the data. In
study after study. grammatical development appears to be tied to lexical level even in
children at the far ends of the continuum, in patterns similar to those displayed by the
two extreme longitudinal cases in figure 10.6. Some further insights into this issue
come from case studies of children with extremely precocious language development
(Thal et al.. 1996). In one of the children in this study, grammar did appear to lag
behind vocabulary level, suggesting some degree of dissociation. However. a detailed
comparison of the free-speech data and parent report data revealed an unexpectedly
strong association between vocabulary development and inflectional morphology for
On the Inseparability of Grammar and trie Lexicn 749
both these children. even though one of them has barely moved out of the single-word
stage!
This conclusin is illustrated in Table 10.2 (from Thal et al., 1996), which provides
examples of the utterances produced by MW (1 7 months od with an expressive vocabu-
lary of 596 words in the CD1) and SW (21 months od with an expressive vocabulary
of 627 words on the CDI). With an MLU of 2.1 3, MW is right whcre we would expect
her to be in grammar, given her vocabulary size (equivalent to performance by an
average 28-30-month-old child in both domains). By contrast, SW has just begun to
combine words (MLU 1.12) despite her huge vocabulary. In fact. her grammatical
abilities are quite average for a 21-month-old child. At first glance it appears that
SW represents a striking dissociation between grammar and vocabulary. However, the
examples n Table 10.2 reveal a very curious phenomenon: production of words
with contrasting inflections (e.g., 'falling . . . fell") in single-word utterances. This is a
very odd phenomenon for children acquiring English, although it has been observed
in children acquiring a highly inflected language such as Turkish (Slobin. 1985).
Applying the criteria for morphological productivity developed by Brown (1973).
Thal et al. discovered that both children have about as much control over English
morphology as we would expect to find in a 2.5-year-old child. In fact, SW was actu-
ally more advanced than MW in grammatical morphology (i.e., productive control over
more morphemes according to Brown's rules), although both children are well within
the range that we would expect for children with more than 500 words (Marchman &
Bates. 1994).
If the difference between MW and SW does not represen! a clear dissociation
between grammar and vocabulary, how can we explain their striking differences in
utterance length? Thal et al. note that SW produced carefully articulated single words.
By contrast, MW was observed to use longer utterances that often appeared formulaic
in nature. Her parents indicated that MW could remember and produce a number of
songs and idiomatic expressions (e.g., "No way, Jos," or ''You little monkey!"). In fact,
to the surprise and amusement of her mother and the experimenten MW produced a
novel juxtaposition of two established formulae during one of the experimental ses-
sions: "No way, you monkey!" Thal et al. tentatively conclude that these two children
differ primarily in the size of the unit that they are able to store in auditory memory,
and/or the size of the unit that they are able to retrieve and reformulate in speech
production (see also Peters, 1983). As we shall see shortly, this kind of processing
account will prove useful n explaining the apparent dissociations observed in some
clinical populations.
Early focal lesione
A different perspectiva on the relationship between early grammar and the lexicn
comes from studies of infants and children with early focal brain lesions to the left or
right hemisphere. usually due to pre- or perinatal stroke (Bates et al.. 1997: Reilly et
al.. 1998). When cases with intractable seizures or other medical complications are
excluded. most studies of this population report language abilities that are well within
the normal range. regardless of lesin side. si/e, or site (Bates et al.. 1999: Kisele &
Aram, 1995; Feldman et al.. 1992: Vargha-Khadem el al.. 1994). As a group. children
Table 10.2 Examples of language production by two very early talkers (from Thal. Bates,
Zappia, & Oroz, 1996)
MW:
Age: 1 7 months od
Vocabulary 5 9 6 words
Vocabulary age: 30 months
MLU: 2.13
MLUage: 28 months
Where cup went?
Where chair went?
Teddy bear went?
Baby doing?
Wanna walk e baby.
Wanna put it on.
Wanna go ride it.
Want mom get off.
Daddy take her. (referring to self)
Help with the apple.
Can't get the teddy bear.
Teddybear the bath.
Too much carrots on the dish.
Move it around.
Clean e bottom.
Put ne sof.
Put in eye.
Momrny wear hat.
Mommy smell it.
Mommy read the book.
Mommy sit down.
Find Becky.
See Becky in the morning.
Becky is nice.
Saw Becky and goats.
SW:
Age: 2 1 months od
Vocabulary: 62 7 words
Vocabulary age: >30 months
MU! 1.1 9
MLtJ age: 20 months
Pretty.
Cute.
Big.
Round.
Dry.
Hungry.
Wet.
Different.
Enough.
Else.
More.
Minute.
Brushing.
Hiding.
Baby crying.
Hold.
Hold it.
Dropped t.
Bring it.
Falling.
Fell.
Talk.
Talking.
VVash'em.
Shirt on.
Teddy up.
Mommy shoe.
On the In&eparability of Grammar and the Lexicn
151
with unilateral brain damage tend to perform below normal controls. bul do not qualify
for a diagnosis of aphasia. This conclusin holds even for chiklren whose injuries
involve the classical language xones, and for some children who have had the entire left
hemisphere removed. Jn fact, when children with early lesin onset are studied after
5-7 years of age, current evidence suggests that there are no significant differenccs at
all between children with left- vs. right-hemisphere injury.
In view of all this evidence for plasticity, what can children with focal brain injury
tell us about the relationship between grammar and the lexicn? The stereotypical vicw
of aphasia in adults might lead one to expect that these children would provide evi-
dence for a dissociation between grammar and the lexicn - at least in the early stages,
when they are trying to get grammar off the ground. Instead. children with early brain
injuries show a tight relationship between dficits observed in grammar and the lexicn
across the years in which these skills develop (Bates et al., 1997; Vicari et al., 2000).
Before 5-7 years of age. brain damage does exact a very heavy cost: regardless of side
or site of injury, infants with early unilateral injuries are often 10-20 months behind
the norm for their age. Furthermore, we can find specific effects of lesin side and site
during the years in which language is first acquired - but surprisingly, these effects bear
little resemblance to the lesion-symptom correlations observed in adults with aphasia.
For example, initial delays in "cracking the code" (word comprehension and gesture
from 10-20 months) are actually greater in infants with damage to the right hemi-
sphere, exactly the opposite of what is typically seen in brain-injured adults. Children
with damage involving the frontal lobes are often especially delayed in word and
grammar production - but it doesn't seem to matter whether the frontal lobe injury
occurs on the left or right side! So far, there is only one clue to the left-hemisphere spe-
cialization for language that usually emerges n uninjured adults: between 1 and 5
years of age (which is a fairly long time), delays n word production are greatest in chil-
dren whose lesions include the temporal lobe (a sensory rea) in the left hemisphere. It
may seem strange that a sensory rea is associated with production delays, but Bates
et al. (1999) point out a possible reason for this correlation: When children are learn-
ing to talk for the first time, they have to carry out a much more detailed perceptual
analysis of the speech signal ("hearing for production") than they need to figure out
what those words mean in the first place ("hearing for understanding"). By 5-7 years
of age (including longitudinal samples), all of these effects seem to disappear. Although
brain-injured children are still slightly behind normal controls, there are no left-right
differences (at least not in most well-controlled studies), and the children have almost
all moved into the normal range.
One final conclusin is most important for present purposes: If children with early
focal brain damage are delayed in vocabulary, they are delayed in grammar - and vice
versa. There seems to be no evidence in this population for a specific lesin that selec-
tively affects grammar (leaving vocabulary intact) or selectively affects vocabulary
(leaving grammar intact). To Ilstrate this last point. figure 10.9 shows the relation-
ship between grammar and vocabulary for 19 individual children in the Bates et al.
(1997) study. compared with the means for normal controls at different vocabulary
levis between 19 and 31 months of age from the MacArthur COI norming study
(Fenson et al., 1 994). We have plotted grammatical complexity against vocabulary si/e
n this figure in a form that faciltales comparison between the focal lesin data and the
752 Introduccin to Word Learning
I
E
S
O
A +Left temporal damage
A Left damage with temporal sparing
o Right damage
On the Inseparability of Grammar and the Lexicn 153
50 100 150 200 250 300 350 400 450 500 550 600 650 700
Vocabulary size
Figure 10.9 Grammar as a function of vocabulary size in children with focal brain injury
(Unes = lOth. 50th & 90th percentile for normis).
other populations considered so far. Seprate symbols are provided to distinguish cases
with left-hemisphere injuries involving the temporal lobe, left-hemisphere injuries that
spare the temporal lobe, and right-hemisphere damage. The three unes in figure 10.9
represen! the 1 Oth, 50th. and 90th percentiles for grammar as a function of vocabu-
lary size in the Fenson et al. normative sample. It should be clear from this figure that
children with focal brain injury display the normal nonlinear relationship between
grammar and vocabulary, even though some of them are markedly delayed on both
(clustered in overlapping symbols in the bottom left quadrant). Of course there is
some variance around this function, but the variance is no greater than we observe
with normal children. Eighteen out of 19 focal-lesion cases fall within the 10-90
window for normal children. and one falls outside; we would expect between 1 and 4
cases to fall outside that window f we were drawing children randomly from the
normal population. In short, there is no evidence for a dissociation between vocabulary
and grammar in this phase of development, even in children who have suffered pre- or
perinatal injuries to the classical language zones within the left hemisphere.
Williams Syndrome and Down Syndrome
Williams Syndrome (WMS) and Down Syndrome (DNS) are genetically based forms of
mental retardation (Bellugi et al.. ] 994: Contardi & Vicari, \ 994; Giannotti & Vicari.
1994: Mervis et al.. 1999: Miller. 1992). In both groups. mean [Qs generally hover
between 40 and 60. although a broader range of JQ scores can be observed at every
stage from infancy through adulthood. People with WMS and DNS tend to end up in
the same special classrooms and residential centers; some adults are able to hold down
simple Jobs, but they are rarely able to live independently. Despite these similarities in
global IQ and life experience, recent studies have revealed sharp contrasts between the
two groups. For our purposes here, we are particularly interested in the claim that
WMS and DNS represen! a double dissociation between language and nonlinguistic
cognition. and between lexical and grammatical aspects of language processing.
Children with DNS are markedly delayed in the acquisition of language. More
importantly. their language abilities at virtually every stage (including the adult steady
State) fall below the levis that we would expect based upon their mental age (Chapman,
1995; Miller. 1992). Furthermore, children and adults with DNS appear to be
especially impaired in the production of bound and free grammatical morphemes.
constituting a form of congenital agrammatism that is even more severe than the
selective delays in grammatical morphology reported for children with Specific
Language Impairment. The function word omissions and structural simplifications
produced by older children with DNS are especially salient in a richly inflected
language like Italian (Contardi & Vicari, 1994), resulting in profiles that are quali-
tatively similar to much younger normal children. although rates of omission are
actually higher in DNS than controls when mean length of utterance is controlled
(Fabbretti et al., 1997).
By contrast. older children and adults with WMS display levis of linguistic knowl-
edge and language use that are surprisingly good when they are compared with the low
levis of performance shown by children with DNS at the same mental age (Bellugi
et al., 1994; Karmiloff-Smith et al., 1995; Mervis et al., 1999). This does not mean that
individuis with WMS are "language savants." Those studies that have used normal
controls have shown that the linguistic performance in WMS falls invariably below
chronological age - which is, of course, not surprising for subjects with an IQ score
around 50. When children with WMS are compared with younger normis matched
for mental age, the picture is mixed. On most measures, performance s closely tied to
mental age (Capirci et al., 1996: Giannotti & Vicari. 1994: Volterra et al., 1996). Per-
formance above mental age is observed in children with WMS in two kinds of tasks:
language tasks that draw heavily on auditory short-term memory (Wang & Bellugi,
1994), and/or language tasks that elicit affectively charged and colorful speech (Reilly
etal.. 1991).
The peculiar patterns of sparing and impairment that are observed within the
language domain in WMS are complemented by equally interesting patterns of sparing
and impairment outside of language. For example, people with WMS and DNS differ
markedly in basic measures of information processing. WMS tend to be extremely sen-
sitive to sound (i.e., hyperacusis). and they perform significantly better than IQ-matched
individuis with DNS on tests of auditory short-term memory (Contardi & Vicari, 1994;
Vicari et al., 1996: Wang & Bellugi. 1994). DNS is often associated with mild to mod-
rate dficits in hearing, but DNS children show significantly better performance than
age-matched WMS on gestural tasks. and DNS adults perform significantly better than
WMS adults on measures of visual short-term memory. These differences in processing
modality may be relevant to the contrasting grammatical profiles observed in WMS and
DNS (see below).
154 Introducton to Word Learning
Based on findings like these, the initia) view of WMS as language savants has giren
way to an interest n this syndrome for the insight that this syndrome offers into Ian-
guage deve/opment under different conditions of information processing (especially
auditory processing), affect and cognition (see Karmioff-Smith, this volume). One of
the most interesting questions revolves around the relationship between grammar and
the lexicn in people with WMS. Earlier papers describing anguage in WMS suggested
that grammar may be "spared" while semantics and pragmatics are deviant or "odd"
(Bellug et al., 1994: Reiiiy et al.. 1991). Pinker (1991) has made the further sugges-
tion that WMS may be particularly adept in the use of regular morphemes, resulting
in overgeneralization errors on irregular forras (e.g.. "goed" instead of "went").
I lowever, more recent evidence suggests that grammar and lexical semantics are both
abnormal in the WMS population. Rubba and Klma (1991) have shown that English-
speaking WMS produce peculiar substitutions of prepositional forms. Karmiloff-Smith
and colleagues have provided evidence suggesting that French-speaking WMS find it
difficult to generalize regular gender morphology to novel words - even though they
are very good at repeating the same novel iteras. Comparative studies of WMS and DNS
in Italian suggest that both groups are markedly impaired in the production of gram-
matical morphemes, including both regular and irregular forms. However, the two
groups differ markedly in the kinds of errors they produce. As noted above, Italian
speakers with DNS tend to err by omisson and simplification, yieldng profiles similar f
to those that are observed in Italian Broca's aphasics, Italians with SLI, and very young
talian-speaking normis. By contrast, Italian speakers with WMS have been shown to I
produce a range of morphological substitutions that are rarely observed in normis at
any age, bearing a distant resemblance to the substitution errors observed in Italian
Wernicke's aphasics (Bates et al., 1991).
In view of the many contrasts observed in older children and adolescents with WMS
syndrome vs. Down Syndrome, it is interesting to ask just how early the two groups
seprate in their profiles of Ianguage developmen t. Is the Ianguage advantage in WMS
(or the Ianguage disadvantage in DNS) present from the very beginning, or does it
emerge only after some critica! cognitive or linguistic milestone is reached? This ques-
tion was addressed by Singer Harris et al. (1997), who used the MacArthur CD to
obtain early Ianguage data from a large sample of children with WMS or DNS between
one and six years of age. In the period of development covered by the infant scale
(equivalent to normal children between 8 and 16 months), WMS and DNS were equally
and severely delayed in both word comprehension and word production. The predicted
separation between WMS and DNS did not emerge until the period of development
covered by the toddler scale (equivalent to normal children between 16 and 30
months). Both groups were still delayed by approximately two years at this point, with
no significant difference in overall vocabulary size. However, Singer et al. found strik-
ing differences in the emergence of grammar. Interestingly, this difference relects a
DNS disadvantage rather than a WMS advantage.
To faciltate comparison across groups. we have plotted the data for individual WMS
and DNS from Singer Harris et al. in figure 10.10, in the same format adopted through-
out this chapter. Within WMS. grammatical development appears to be paced by vo-
cabulary size, in the normal fashion. In fact. when these children are compared with
lexically matched normal controls from the CDJ sample, the relationship between
On the Inseparabilty of Grsmmar and the Lexicn
155
o Williams Syndrome
Down Syndrome
50 100 150 200 250 300 350 400 450 500 550 600 650 700
Vocabulary size
Figure 10.10 Grammar as a function of vocabulary size in children with Williams vs. Down
Syndrome (lines = lOth. 50th & 90th percentile for normis).
grammar and vocabulary size is identical. following the same nonlinear accelerating
function described above for normis and for children with focal brain injury. In short,
there is no evidence for a dissociation between grammatical and lexical development in
WMS - at least not in this early phase of grammatical development.
By contrast, the DNS sample provides our best evidence to date for a significant
dissociation between grammar and the lexicn. In particular, DNS children scored
significantly below the grammatical levis displayed by normal children and by WMS
matched or vocabulary size (Figure 10.10). However, the term "dissociation" must be
used with caution here. Correlational analyses show that grammar and vocabulary
are tightly related in both children with WMS and children with DNS (correlating
around +.70). There is a real difference between these groups, but it lies in the amount
of vocabulary that children in each group seem to need in order to get grammar off
the ground: WMS need the same amount required by normal children of the same mental
age, while DNS require more than the normal amount of vocabulary in order for grammar to
emerge.
As we noted earlier. there is a double dissociation in these two groups between audi-
tory short-term memory (signiflcantly better in WMS) and visual short-term memory
(significantly better in DNS). It appears that DNS suffer from a selective impairment in
one or more aspects of auditory processing, a dficit that is superimposed u pon their
more general cognitive delay. Under these circumstances. it is perhaps not surprising
the DNS are selectively impaired in the ability to detect. store, and/or retrieve those
aspects of their l i ngui st i c i nput t hat are lovvest in what I.eonurd (1998) refers to as
56 Introduction to Word Learning
"phonological substance," or salience. This argument is similar to one that Tallal and
hercolleagues have used to explain the selective grammatical dficits displayed by older
children with SU (Tallal et al., 1985). We will return to this issue later on, with refer-
ence to selective impairments of grammar in brain-injured adults.
Explaining the Link
Why is the relationship between grammar and the lexicn so strong in this period of
development? The same basic nonlinear relationship appears in longitudinal and cross-
sectional data, in at least two dramatically dfferent languages (see Caselli et al., this
volume), in different domains of grammar. in children who are developing on a normal
schedule and in children who are developing at an aberrant rate. We do find a viola-
tion of this relationship in children with Down Syndrome, but the deviation seems to
involve not a dissociation between grammar and the lexicn, but a pathological differ-
ence in the amount of vocabulary required in order to learn grammar. There is no a
priori reason why this had to be the case. Por example, many adults who try to acquire
a second language in the classroom find themselves in a situation in which their stock
of grammatical rules outstrips their limited vocabulary. The dependence of early
grammar on vocabulary size is so strong and the nonlinear shape of this function is
so regular that it approaches the status of a psychological law. akin to the reliable psy-
chophysical functions that have been observed in perception (e.g.. Weber's Law,
Fechner's Law). But explanation by legislatura is not very satisfactory, and it is parti-
cularly unsatisfactory if better explanations are available. We can offer at least five
reasons why grammar and vocabulary track each other so closely in this stage of lan-
guage development. None of them are mutually exclusive.
1 Perceptual bootstrapping. Nusbaum and Henly (1992) have proposed that
efficient word perception requires a certain amount of top-down processing, per-
mitting the listener to weed out inappropriate candidates from a large pool of
tems that overlap (at least partially) with the blurred word tokens that so often
occur in fluent speech. To the extent that this is true. it is probably even more
true for the perception of grammatical function words and bound inflections.
For a variety of reasons. these units are particularly hard to perceive (Herrn &
Bates. 1997; Hurlburt&Goodman. 1992:Leonard. 1998).Theytendtobeshort
and low in stress even in speech that is produced slowly and deliberately. In infor-
mal and rapid speech, speakers have a tendency to exploit the frequency and pre-
dictability of function words and bound morphemes by giving them short shrift.
deforming their phonetic structure and blurring the boundaries between these
morphemes and the words that surround them. In fact, when grammatical func-
tion words are clipped out of connected speech and presented in isolation. adult
native speakers can recogni/e them no more than 40-50% of the time (Herrn
& Bates. 1997). This is true of speech directed to children as well as speech
directed to adults (Goodman & Nusbaum. 1994). Under these circumstances. we
should not be surprised that young children are unable to acquire grammatical
On the Inseparability of Grammar and trie Lexicn 157
forms until they have a critical mass of conten words. providing enough top-
down structure to permit perception and learning of those "littlc sounds" t hat
occur to the right or left of "real words."
2 Logical bootstrapping. Studies in several different languages have shown that
verbs and adjectives are acquired later than nouns (see Caselli el al., this volume).
Except for a few terms like "up" and "no" that can stand alone, function words
tend to appear later still, well after the first verbs and adjectives appear. Fur-
thermore, many relatively early prepositions (e.g., "up") may not be used in the
same way by children as by adults. Adults use them to specify a relation between
objects or a location. Children on the other hand use them to refer to events
(Tomasello & Merriman, 1995) instead of using them as "grammatical glue." It
has been suggested that this progression from ames to predication to grammar
is logically necessary, based on a simple assumption: Children cannot under-
stand relational terms until they understand the things that these words relate.
One can arge about the extent to which this assumption holds for individual
structures, but it may provide a partial expianation for the dependence of
grammar on lexical growth.
3 Syntactic bootstrapping. The perceptual and logical bootstrapping accounts
both presuppose that the causal link runs from lexical growth to grammar.
However, studies from several different laboratories have shown that children
between 1 and 3 years of age are able to exploit sentential information to learn
about the meaning of a novel word (Goodman et al., 1998; Naigles et al.. 1993;
Tomasello, 1992). Naigles et al. (1993) refer to this process as "syntactic boot-
strapping," although t has been shown that children can use many different
aspects of a sentence frame for this purpose, including sentence-level semantics.
morphological cues, word order, and prosody. It is therefore possible that the
accelerating function in Figures 10.2-10.10 is due in part to the effect of the
child's emerging grammar on lexical growth.
4 Nonlinear dynamics of learning in a neural network. The above three
accounts all support a link between lexical and grammatical development, but
it is not obvious from these accounts why the function ought to take the non-
linear form that appears so reliably across populations and age levis. We noted
earlier that the nonlinear functions governing the relation between verb vocabu-
lary and the emergence of regular and irregular past-tense marking appear in
a similar form in English-speaking children and in neural network simulations
of past-tense learning (Elman, this volume; Marchman & Bates 1994; Plunkett
& Marchman. 1993). This is only one example of a more general point: Multi-
layered neural networks produce an array of nonlinear growth functions,
reflecting the nonlinear dynamics of learning and change in these systems. The
kinds of critical-mass effects that we have proposed to underlie the relation
between lexical and grammatical growth may be a special case of this more
general approach lo the nonlinear dynamics of learning (Thelen & Smith.
1994).
5 Lexically based grammar. Finally. as we noted at the outset of this paper. the
historical trend in modern linguistics has been to place in Ihe lexicn more and
15& Introduction to Word learning
more of the work that was previously carried out in a seprate grammatical com-
ponent. The powerful relation between grammatical and lexical development
that we have observed here is precisely what wc would expect if grammar is an
inherent part of the lexicn.
Points 1-4 all pertain to learning. Point 5 is a stronger claim, extending to the
relationship between grammar and the lexicn in the adult steady statc. The data
that we have reviewed so far may be relevan! only to the early stages of language
development, the period in which the fundamental properties of language-specific
morphosyntax are laid down. It is entirely possible that a modular distinction between
grammar and the lexicn may emerge at a latter point in development. in accordance
with the processes of "modularization" described by Karmiloff-Smith (1992): see also
Bates et al., 1988). This question is best addressed by looking at the literature on
language disorders in older children and adults. where strong claims about the
modularity of grammar and the lexicn have been made. Our review of this literature
suggests that the lexicalist approach stands up to the evidence for older children
and adults as well. For every clinical group for whom grammatical dficits have
been reported. lexical dficits have also been demonstrated. In fact, a dficit in word
finding (called "anomia") is the one symptom that is present in every form of adult
aphasia (Goodglass, 1993). Conversely. in every clinical group for whom lexical
dficits are reported, some kind of grammatical problem is observed as well. A par-
ticularly interesting observation about these links is the fact that grammar and the
lexicn tend to break down in the same way within individual patients. Non-fluent
Broca's aphasics tend to err by omission; they leave out grammatical inflections
and function words. but they also tend to leave out conten words (especially verbs).
This omission profile is also seen (usually in a milder form) in children with Down
Syndrome and children with Specific Language Impairment. Fluent Wernicke's
aphasics tend to err by substitution, putting in the wrong conten word (or a blend
of words). In richly inflected languages, the same tendency (called paragrammatism)
is observed for inflections and function words. This "substitution profile" is also
observed in children with Williams Syndrome, if they are acquiring a language
with enough morphology for such substitutions to surface (Volterra et al., 1996).
Finally. patients with Alzheimer's disease and patients with milder forms of aphasia
rarely make frank grammatical errors of any kind (omission or substitution).
However, they tend to avoid production of complex syntactic structures (e.g.. the
passive. Bates et al., 1995b), a grammatical symptom that mirrors their problems
in lexical production (i.e., an overreliance on pronouns and other "light forms"). A
mild versin of this "avoidance profile" is also seen in normal aging. In short, grammar
and the lexicn continu to travel together in adulthood. and they break down in
similar ways.
Grammar and the lexicn develop together in infancy, and they break down together
in brain-injured adults. In our view, the most parsimonious explanation for these facts
would be a mental/neural architecture in which grammar and the lexicn are repre-
sented together, and handled by the same mechanisms for learning and processing - in
short, a lexicalist view of the sort that has gained considerable acceptance in modern
linguistics.
On the InseparaWity of Grammar and the Lexicn
Notes
159
1 "Observation of early stages of language acquisilion may be quite misieading in this regard.
It is possible that at an early stage mere is use of languagelike expressions. but outside the
framework imposed. at a later stage of intellectual maturation, by the faculty of language -
much as a dog can be trained to respond to certain commands, though we would not
conclude. from this. that it is using language." (Noam Chomsky, Reflections on language
(p. 53). New York, Pantheon Books, 1975).
2 We also looked at many other metrics of grammatical development, including propositional
complexity and morphological productivity. After all that work we were surprised to find that.
at least in this period of development, MLU is so highly correlated with other, more
sophisticated measures, that there was no point in using any other estmate of grammar in
correlational analyses with other variables; for a discussion of this point. see Bates et al..
1988.
References
Bates. E., Bretherton, I.. & Snyder. L. (1988). Fromfirst words to grammar: Individual differences and
dissociable mechanisms. New York: Cambridge University Press.
Bates. E.. Dale. P., & Thal, D. (1995a). Individual differences and their implications for theories
of language development. In P. Fletcher & B. MacWhinney (Eds.), Handbook of child anguage
(pp. 96-151). Oxford: Basil Blackwell.
Bates, E.. Marchman. V. Harris, C., Wulfeck, B.. & Kritchevsky. M. (1995b). Production of
complex syntax in normal ageing and Alzheimer's Disease. Language and Cognitive Processes,
11 (5).487-539.
Bates. E.. Thal. D.. Trauner, D.. Fenson. J., Aram. D.. Bisele. J.. & Nass, R. (1997). From flrst words
to grammar in children with focal brain injury. In D. Thal & ]. Reilly (Eds.). Special issue on
Origins of Communication Disorders. DevelopmentalNeuropsychology. 13 (3). 275-343.
Bates. E.. Vicari. S.. & Trauner, D. (1999). Neural mediation of language development:
Perspectives from lesin studies of infants and children. In H. Tager-Flusberg (Ed.), Neurode-
velopmental disorders (pp. 533-581). Cambridge, MA: MIT Press.
Bates. E.. Wulfeck. B., & MacWhinney. B. (1991). Crosslinguistic research in aphasia: An
overview. Brain and Language. 41. 123-148.
Bellugi. U., Wang. P.P.. & Jernigan. T.L. (1994). Williams Syndrome: An unusual neuropsycho-
logical protile. In S. Broman & J. Grafman (Eds.), Atypical cogntive dficits in developmental
disorders: Implications for brain function. Hillsdale. NJ: Erlbaum.
Bickerton. D. (1984). The language bioprogram hypothesis. The Behavioral and Brain Sciences. 7.
173-187.
Bishop. D.V.M. (1997). Uivammon understanding: Development and disorders of comprehension in
children. Hove, UK: Psychology Press.
Bloom. L. (1991). Language development from two to three. New York: Cambridge University Press.
Bresnan, J. (2001). Lexical fimctiona syntax. Malden. MA: Blackwell.
Brown, R. (1973). A flrst anguage: The early stages. Cambridge. MA: Harvard University Press.
Capirci. O.. Sabbadini. L., & Volterra. V. (1996). Language development in Williams Syndrome: A
case study. Cotmitin' Ncwopsi/r/io/oiji/. J 3 ( 7) , 1017-1039.
Caselli. M.C., Casadlo. P, & Bates. E. (1999). A comparison of the transition from tirst words to
grammar in English and Italian. Journal of (.'luid Limauaie. 26, 69-11 1.
16O
Introduccin to Word Learning
Chapman. R.S. (1995). Language developmenl n children and adolescents with Down
Syndrome. In Paul Fletcher & Brian MacWhinney (Eds.). The handbook of child language
(pp. 664-689). Oxford: Basil Blackwell.
Chomsky. N. (1965). Aspects of the theorij of syntax. Cambridge, MA: MIT Press.
Chomsky. N. (1 975. Reflections an language. New York. Pantheon Books. 1975).
Chomsky, N. (1981). Lectures on governmenl and binding. New York: Foris.
Chomsky. N. (1995). The minimalist program. Cambridge. MA: MIT Press.
Contardi, A..& Vicari, S. (Eds.). (1994). Le persone Down: Aspetti neuropsicologid, educativi e sociali.
Milano: FrancoAngeli.
Dale. RS. (1991). The validity of a parent report measure of vocabulary and syntax at 24 months.
Journal of Speech and Hearing Sdences. 34. 565-571.
Dale, RS.. Bates. E., Reznick. S.. & Morisset. C. (1989). The validity of a parent report instrument
of child language at 20 months. Journal of ChildLanguage, 16, 239-249.
Eisele. }., & Aram. D. (1995). Lexical and grammatical development in children with
early hemisphere damage: A cross-sectional view from birth to adolescence. In P. Fletcher
& B. MacWhinney (Eds.), The handbook of child language (pp. 664-689). Oxford: Basil
Blackwell.
Elman, J.. Bates, E., Johnson. M.. Karmiloff-Smith, A.. Parisi. D.. & Plunkett. K. (1996). Rethnk-
ing innateness: A connectionist perspective on development. Cambridge. MA: MIT Press/Bradford
Books.
Fabbretti, D., Pzzuto. E., Vicari, S., & Volterra. V (1997). A story description task in children
with Down syndrome: lexical and morphosyntactic abilties. ournal of Intelectual Disability
Research. 41 (2), 165-179.
Feldman. H., Holland. A., Kemp. S., SJanosky, J. (1992). Language development after unilateral
brain injury. Brain and Language. 42, 89-102.
Fenson. L. Dale, P, Reznick, J., Bates, E., Thal, D.. & Pethick. S. (1994). Variability in early com-
muncative development. Monographs of the Sodety for Research in Child Development, Serial
No. 242. Vol. 59. No. 5.
Fillmore, C.J., Kay, R, & O'Connor, C. (1988). Regularity and idiomaticity in grammatical
constructions: The case of Let Alone. Language, 64. 501-538.
Fodor, J.A. (1983). The modularity of mina: An essay onfaculty psychology. Cambridge, MA: MIT
Press.
Garrett, M. (1992). Disorders of lexical selection. Cognition. 42, 143-180.
Giannotti, A., & Vicari, S. (Eds.). (1994). U bambino con sndrome di Williams. Miln: FrancoAngeli.
Goldberg, A.E. (1995). Constructions: A construction grammar approach to argument structure.
Chicago: University of Chicago Press.
Goodglass. H. (1993). Understanding aphasia. San Diego: Academic Press.
Goodman, J.C. (199 5). The shape of change: Longitudinal evidence about language development.
Sodety for Research in Child Development Abstrais. 10. 111.
Goodman. J.C.. McDonough, L., & Brown. N.B. (1998). The role of semantic context and memory
in the acquisition of novel nouns. Child Development. 69 ( 5) , 1330-1344.
Goodman, J.C.. & Nusbaum, H.C. (Eds.). (1994). The development of speech perception: Tnmsitions
from speech sounds to spoken words. Cambridge. MA: MIT Press.
Gopnik. M. (1990). Feature-blind grammar and dysphasia. Nature. 344 (6268), 715.
Herrn, D.. & Bates, E. (1997). Sentential and acoustic factors n the recognition of open- and
closed-class words. Journal of Memory and Language, 37 (2), 217-239.
Hurlburt, M.S., & Goodman. J.C. (1992). The development of lexical elTecis on children's
pilneme identilications. In ].]. Ohala, T.M. Nearey. B.L. Derwing. M.M. Hodge, & t.E. Wiebe
(Eds.), ICSLP 92 Proceedings: 1992 International Conference on Spoken Language Prncessing
(pp. 3 37-340). Banff. Alberta. Canad: University of Alberta.
r
On the Inseparability of Grammar and the Lexicn ]Q"\
Jahn-Samilo. J. (1995). Language comprehension and production in children from 8 to 30
months of age: A comparison of parent report and laboratory measures from a longitudinal
study. Sodety for Research in Child Development Abstracts, 10, 112.
Jakobson. R. (1 968). Child language aphasia andphonological universals. Paris: Mouton.
Karmiloff-Smilh, A. (1992). Beyond modularity: A developmenlal perspective on cognitive science.
Cambridge. MA: MIT Press.
Karmiloff-Smith, A.. Klima, E.S.. Bellugi. U., Grant, J.. & Baron-Cohen, S. (199 5). [s there a social
module? Language. face processing, and theory of mind in subjects with Williams Syndrome.
Journal of Cognitive Neiirosdence, 7 ( 2 ) . 196-208.
Langacker. R. (1987). Foundations of cognitive grammar. Stanford: Stanford Universily Press.
Leonard. L.B. (1998). Children with spedfic language impainnent. Boston MA: MT Press.
MacDonald. M.C.. Pearlmutter. N.J.. & Seidenberg. M.S. (1994). Lexical nature of synlactic
ambiguity resolution. Psychological Review, 101 (4), 676-703.
MacWhinney. B. (1993). Connections and symbols: Closing the gap. Cognition. 49 ( 3) , 291-
296.
MacWhinney. B., & Bates. E. (Eds.). (1989). The crosslinguistic study of sentence processing. New
York: Cambridge University Press.
Marchman. V., & Bates. E. (1994). Continuity in lexical and morphological development: A test
of the critical mass hypothesis. Journal of Child Language, 21 (2), 339-366.
Marchman. V, Wulfeck, B., & Weismer, S.E. (1995). Productive use of English past-tense morpho-
ogy n children with SLI and nonnal language (Tech. Rep. No. CND-9514). La Jolla: University
of California. San Diego. Center for Research in Language. Project in Cognitive and Neural
Development.
Marcus. G., Pinker, S.. Ullman. M., Hollander. M., Rosen. T.J., &Xu. E (1992). Overregularization
in language acquisition. Monographs of the Sodety for Research in Child Development, 57.
Mervis, C. Morris. C., Bertrand, J.. & Robinson. B. (1999). Williams syndrome: findings from
an integrated program of research. In H. Tager-Flusberg (Ed.), Neurodevelopmental disorders
(pp. 65-110). Cambridge. MA: MIT Press.
Miller, J. (1992). The development of speech and language in children with Down Syndrome.
In E. McCoy and I. Lott (Eds.), Clinical care for persons with Down Syndrome (pp. 121-138).
New York: Academic Press.
Naigles, L., Gleitman, H., &Gleitman. L.R. (1993). Children acquire word meaning components
from syntactic evidence. In E. Dromi (Ed.), Language and cognition: A developmental perspective
(pp. 104-140). Norwood. NJ: Ablex.
Nelson, K. (1973). Structure and slrategy in learning to talk. Monographs of the Sodety for
Research in Child Development. 38 (1 & 2. Serial #149).
Nusbaum. H.C., & Henly, A.S. (1992). Listening to speech through an adaptive window of analy-
sis. In B. Schouten (Ed.). The processing of speech: From the auditory periphery to word recognition.
Berln: Mouton-De Gruyter.
Peters. A. (1983). The units of language acquisition. Cambridge: Cambridge IJniversity Press.
Pinker, S. (1991). Rules of language. Science. 253. 530-535.
Plunkett, K.. & Marchman. V. (1993). From rote learning lo system building: Acquiring verb
morphology in children and connectionist nets. Cognition. 48. 21-69.
Pollard. C.. & Sag, I. (1994). Head-drive phrase structure grammar. Center for the Study of
Language and Information. Report # CSLI-88-1 32. University of Chicago.
Reilly. J.. Bales. E.. & Marchman. V. (1998). Narrativo discourse in children with early focal brain
injury. In M. Dennis (l i d. ). Special issue. Discourse in children with anomalous brain develop-
menl or acquired brain injury. Brain and Language. 6! ( 3 ) . 3 35-375.
Reilly. J.. Klima. F... & Bellugi. U. ( 1991) . Once more with feeling: al'fect and language in atypical
populalions. Developmental Psychopathology. 2. 367-391.
J62 Introduccin to Word learning
Rubba, ]., & Klima, E.S. (1991). Preposition use in a speaker with Williams Syndrome: Some
cognitive grammar proposals. Center for Research in Language Newsletter Val. 6. No. 1. University
of California, San Diego.
Shore, C.M. (1995). Individua! differences in language development. Thousand Oaks. CA: Sage
Publicalions. Inc.
Singer Harris, N., Bellugi. U.. Bates. E., Jones. W.. & Rossen. M. (1997). Contrasting proiles of
language development in children with Williams and Uown Syndromes. In 1). Thal & }. Reilly
(Eds.). Specal issue on origins of Communication Disorders. Developmental Neuwpsychology,
13( 3) . 345-370.
Slobin. D. (Ed. ). (1985-1997). The crosslinguistic study of language acquisition (Vols. 1-4).
Hillsdaie, NJ: Erlbaum.
Tallal. R.Stark. R..&Mellits. D. (1985). Identification of language-impaired children onthebasis
of rapid perception and production skills. Brain and Language. 25. 314322.
Thal. D. (1991). Language and cognition in late-talking toddlers. Topics in Language Disorders. 11.
33-42.
Thal. .. Bates. E., Goodman, J.. & Jahn-Samilo, J. (1997). Continuity of language abilities in
late- ar.d early-talking toddlers. In D. Thal & J. Reilly (Eds.), Special issue on Origins of Com-
munication Disorders, Developmental Neuropsychology. 13 (3), 239-273.
Thal. D.. Bates. E., Zappia. M.].. & Oroz. M. (1996). Ties between lexical and grammatical devel-
opment: Evidence from early talkers. Journal of Chd Language. 23 (2). 349-368.
Thelen. E.. & Smith. L.B. (1994). A dynamic systerns approach to the development of cognition and
action. Cambridge. MA: MIT Press.
Tomastllo. M. (1992). First verbs: A case study of early grammatical development. Cambridge. UK:
New York: Cambridge University Press.
Tomasello. M.. & Merriman, W. (Eds.) (1995). Beyond ames for things: Yoitng children's acquisition
of verbs. Hillsdaie. NJ: Lawrence Erlbaum.
Vargha-Khadem. E, Isaacs. E., & Muter, V. (1994). A review of cognitive outcome after unilat-
eral lesions sustained during childhood. Journal of Child Neurology, 9 (SuppL), 2S67-2S73.
Vargha-Khadem, E, Watkins, K.. Alcock. K.. Fletcher. E, & Passingham. R. (1995). Praxic and
nonverbal cognitive dficits in a large family with a genetically transmitted speech and lan-
guage disorder. Proceedings of the National Academy of Sciences USA, 92. 930-933.
Vicari, S., Albertoni, A.. Chilosi. A.. Cipriani, E. Cioni, G.. & Bates. E. (2000). Plasticity and reor-
ganization during early language learning in children with congenital brain injury. Cortee. 36.
31-46.
Vicari. S.. Brizzolara, D.. Carlesimo, G.A., Pezzini. G.. & Volterra. V. (1996). Memory abilities in
children with Williams syndrome. Cortex. 32 (3). 503-514.
Volterra. V. Pezzini. G., Sabbadini. L, Capirci. O.. & Vicari. S. (1996). Linguistic abilities in [tallan
children with Williams syndrome. Cortex. 32 (4). 663-666.
Wang. E. & Bellugi. U. (1994). Evidence from two genetic syndromes for a dissociation between
verbal and visual-spatial short-term memory. Journal of Clnica! and Experimental Neuropsy-
chology. 16. 317-322.