© 2009 The Linguistic Association of Canada and the United States (lacus).
The content of this article is from lacus Forum 26 (published 2000). This article and
others from this volume may be found on the Internet at http://www.lacus.org/volumes/26.
YOUR RIGHTS
This electronic copy is provided free of charge with no implied warranty. It is made available
to you under the terms of the Creative Commons Attribution-NonCommercial license
version 3.0
(http://creativecommons.org/licenses/by-nc/3.0/)
Under this license you are free:
•
•
to Share — to copy, distribute and transmit the work
to Remix — to adapt the work
Under the following conditions:
•
•
Attribution — You must attribute the work in the manner specified by the author
or licensor (but not in any way that suggests that they endorse you or your use of the
work).
Noncommercial — You may not use this work for commercial purposes.
With the understanding that:
•
•
Waiver — Any of the above conditions can be waived if you get permission from the
copyright holder.
Other Rights — In no way are any of the following rights affected by the license:
• Your fair dealing or fair use rights;
• The author's moral rights;
• Rights other persons may have either in the work itself or in how the work is
used, such as publicity or privacy rights.
Notice: For any reuse or distribution, you must make clear to others the license terms of
this work. The best way to do this is with a link to the web page cited above.
For inquiries concerning commercial use of this work, please visit
http://www.lacus.org/volumes/republication
Cover: The front cover of this document is licensed under the Creative Commons
Attribution-No Derivative Works 3.0 license (http://creativecommons.org/licenses/bynd/3.0/) and may not be altered in any fashion. The lacus “lakes” logo and University
of Alberta logo on the cover are trademarks of lacus and the University of Alberta
respectively. The University of Alberta logo is used here with permission from the
trademark holder. No license for use of these trademarks outside of redistribution of this
exact file is granted. These trademarks may not be included in any adaptation of this work.
THE DEVELOPING LEXICON:
MORPHOSYNTACTIC KNOWLEDGE IN FIFTH GRADE CHILDREN
Martha Smith & Elena Nicoladis
University of Alberta
Much attention has been given to ascertaining the nature of the mental lexicon
through the use of literate adult subjects (generally university undergraduates).
Psycholinguists have also given a good deal of attention to the nature of lexical
development in pre-literate children. In between early child language acquisition
and the nature of the adult mental lexicon, however, is a vast and relatively unexplored frontier. The language skills that are mastered in the first few years of life are
so impressive that a child of five already appears to be a master of his/her language.
Beyond this rapid initial period of linguistic growth, however, development appears
gradual and difficult to determine. Perhaps as a consequence, comparatively little
attention has been devoted to understanding the development of the mental lexicon beyond the early years. However, as educational research has made clear, children’s vocabularies grow at astounding rates through the grade school years
(Jenkins & Dixon 1983, Nagy & Anderson 1984, Nagy & Herman 1987). Thus, by
implication, the nature of the ‘mental lexicon’ changes and develops and it is of
some import to understand the nature of this development.
In this paper, we are concerned with an aspect of lexical knowledge that is
undergoing development in the grade school years. Specifically, we are concerned
with differences between the morphosyntactic knowledge of fifth grade children (10
year olds) and that of literate, educated adults. To investigate morphosyntactic
knowledge we probed children’s and adults’ intuitive knowledge of the syntactic functions of derivational suffixes. The study of derivational suffixes can give us insight
into the lexicon-syntax interface because suffixes are obviously parts of lexical items
that have some semantic content as well as syntactic functions. In the word teacher,
for example, the -er indicates an agent (semantic content) and also marks the word
as a noun (syntactic function). Clearly, knowing something about the syntactic
functions of derivational morphology is part of lexical knowledge.
The syntactic aspect of derivational knowledge (cf. Tyler & Nagy 1989) involves
knowledge (tacit or explicit) that derivational suffixes mark words for ‘part of speech’
or word class membership (syntactic category like noun, verb, etc.). For example,
most words ending with the suffix -ize are verbs (e.g. sympathize, visualize, traumatize) while the suffix -ity marks nouns (e.g. majority, inferiority, possibility, security).
One effect of this characteristic of derivational suffixes is that related words play different roles in sentences (e.g. national is an adjective while nationality is a noun).
190
martha smith & elena nicoladis
Several major complications arise in an attempt to develop methodology to
assess knowledge of the syntactic functions of suffixes. Perhaps the most formidable is that these syntactic category indicators are not invariant, an observation that
is frequently pointed out in literature on the nature of morphology (Matthews,
1991; Henderson, 1989). In many cases, a suffix is not a reliable guide to word class
information. The same form, for example, can have several functions such as the
suffix -er which functions differently in the words slower (adjective) and teacher
(noun). Syntactic category is primarily determined by word order in English so that
a word like question can be either a noun (to ask a question) or a verb (to question
someone) depending on how it is used in a sentence. Even though a word may end
in a suffix that usually marks nouns, for example, other sentence context information like surrounding grammatical elements (a, the, with, very) and word order are
needed to establish the part of speech of the word. An additional complication is
that the combination of letters or sounds that make up derivational suffixes are not
unique to suffixes; that is, there is orthographic (and often phonological) overlap
between suffixes in derived words and letter/sound sequences in many other words.
For example, the words exhaustive, offensive and creative can properly be viewed as
derived words ending in the suffix -ive which marks adjectives; many other English
words, however, also end in the sequence -ive which cannot be viewed synchronically in the same way (e.g. captive, motive, forgive).
Because of the probabilistic nature of suffixes, we looked at the performance of
educated adults on a nonsense word task and then used adult performance as the
target for children. This rationale subsumes information about exposure to suffixes
in context; fifth grade children can be assumed to have had less exposure than educated adults but the exposure should be similar in nature. If, for example, words
ending in the sound/letter sequence -ive are used as adjectives 90% of the time in
actual language usage (oral and written), an educated adult would presumably have
a strong perception of -ive being an adjective ending. The development of children’s
intuitions should be in the same direction depending on the amount of exposure
to such tokens (among other factors).
Although frequency information on the relative informativenss of suffixes (e.g.
the percent of -ive words that are adjectives) would illuminate the acquisition picture, we found this information difficult to find and translate in terms that would
inform our findings. A backwards dictionary, for example, can identify words in the
language that end in various letter sequences and provide word class information
about the entry. Token frequencies for each type would need to be compiled to provide a relevant estimate for adults. Such an estimate, however, would not be reasonable for fifth grade children. Thus, for the purposes of this study, we did not
attempt to elucidate the role of frequency information in children’s performance.
Rather, we assumed that educated adult intuitions took this information into
account and that the standard set by adults was the standard to which children
could be reasonably compared. This approach is similar to that adopted by other
child language researchers who have estimated natural productivity of lexical items
the developing lexicon
191
by examining adult performance on an experimental task (cf. Clark & Berman,
1984). Thus, the research question was as follows: Do 10 year old children understand (i.e. have they acquired knowledge of) the lexical category information
implicit in derivational suffixes? That is, given the nature of English derivational
suffixes, how does the performance of fifth grade children compare to that of literate adults on a task designed to probe intiuitve knowledge of the syntactic functions
of English derivational suffixes?
1. method
1.1. participants. The target population for this study was native English speaking
fifth grade students in a large (population approximately 626,000) midwestern city
in Canada. To ensure a sample that was broadly representative of the population,
the city was divided into five geographical areas and several schools were randomly
selected from each area. The sample was drawn from 15 fifth-grade classrooms that
supplied a possible 402 participants representing all levels of SES. All students who
received parental permission and were present at the time of testing participated in
the project but, for the purposes of this analysis, data were excluded from students
who were not native English speakers or who were identified by school criteria as
learning disabled or linguistically delayed in some way (since the study was limited
to normally developing fifth grade children).
The final sample consisted of 301 fifth grade students. Females accounted for
55% of the sample while males made up 45% of the sample. The mean age of this
group was 10 years, 4 months and the median age was 10 years, 3 months. While ages
ranged from 9 years, 8 months to 11 years, 11 months, the majority of students’
ages (90%) fell between 9 years, 9 months and 10 years, 9 months. Thus, there was
considerable homogeneity in age in that the bulk of the sample was less than a year
apart in age.
Adult participants were 29 graduate students who were attending a non-credit
workshop on language analysis in a well-known research university on the East coast
of the USA. All students were native speakers or had an extremely high level of
English competence as required for graduate school admission. Thus, these adults
represent highly educated English speaking adults from a variety of backgrounds.
1.2. materials. This study utilized a nonsense word production task where the
objective was to get children to produce sentences using nonsense words that ended
in real English suffixes (that presumably coded part of speech information). The
stimuli for the task consisted of 10 nonsense words with real English derivational
suffixes: four noun (-ment, -ness, -ity, -(t)ion), three adjective (-less, -able, -ive) and
three verb (-ize, -ate, -ify) suffixes were included. The pseudo-stems for the nonsense words were created with respect for the constraints of real English words (e.g.,
str- is an allowable word initial consonant cluster whereas tzr- is not). The resulting
stimuli were, therefore, possible but non-occurring words in English (see Smith,
1998, for further discussion).
192
martha smith & elena nicoladis
Three training examples were constructed to show students that nonsense words
could be put into a variety of acceptable sentences (e.g. That girl is prastful. They
had a prastful time at school). The training examples used suffixes that are generally
mastered quite early in development; the suffix -er, for example, is comprehended
and produced by children as young as 3–5 years old (Clark and Hecht, 1982; Clark
& Cohen, 1984). Two of the three suffixes used in the training examples had more
than one syntactic function (e.g. -ly is found on adverbs such as slowly as well as
adjectives such as lovely). Thus one training example could show the same suffixed
nonsense word in several different syntactic roles. The task directions were constructed to convey the idea that, just like real words could be used in sentences, the
same could be done with the nonsense words in the task. It was clearly conveyed
to students that (1) since the words were not real words, they could mean whatever
the student wanted them to mean, (2) there was no particular ‘right’ answer and
(3) the experimenter was only interested in the kinds of sentences people made
when given ‘fake’ words. Full task directions and further discussion about the
design of the task can be found in Smith (1998).
1.3. procedure. The children’s nonsense word production task was administered as
the third task in a series of tasks tapping morphological knowledge. After working
through the training examples, the student was asked to try to make up some sentences with some fake words. The student was then presented with each of the 10
stimuli items in turn (all students saw and heard the stimuli in the same randomized order). Presentation was both visual (nonsense words were clearly printed on
cards of a uniform size) and oral (the examiner pronounced the word several
times) with an emphasis on listening to the pronunciation of the word. Each student was given as much time as needed to come up with a sentence. All sentences
were equally accepted with the exception of cases where students changed the nonsense word (such as by adding an inflectional ending). In these cases, the student
was asked to try again without changing the nonsense word. The examiner repeated
each sentence for student verification and tape clarity. If a student’s sentence was
not clearly heard, he/she was asked to repeat it. If a student had difficulty thinking
of a sentence, the item was set aside and the student was encouraged to give it
another try after the other items had been completed. The task was tape recorded
and the students’ sentences transcribed.
Some minor changes were made in the administration of the task to adults. An
adult version of the task was administered to 29 graduate students in one half hour
session. Participants were not told the purpose of the task prior to the collection of
their productions. The adult version of the task used the same stimuli items as the
children’s task but the graduate students were directed to write their sentences on a
response sheet that was provided. Task directions and training examples were
adjusted for presentation to a group of adults. Students saw the nonsense word in
the developing lexicon
193
print (each stimulus item was written on the chalkboard and displayed on a large card
as introduced) and heard the item (the experimenter pronounced each item several
times while giving the students time to write a sentence using the nonsense word).
1.4. coding. After transcription, a three-tiered scoring system was applied. Full
points (two points) were assigned to productions where the participant (child or
adult) had used the nonsense word unambiguously in accordance with the major
syntactic role of the appended suffix. That is, it was clear from word order and other
grammatical elements in the sentence (a, the, very, etc.) that the suffixed form was
used in the syntactic role indicated by the suffix. For example: The fladment was
destroyed in the accident. No points (0 points) were assigned if the nonsense word
was unambiguously used in a role that was clearly not in accordance with the usual
role of the suffix. For example: I tropanity every day (the word with the noun suffix -ity is used as a verb). Ambiguous constructions that did not include the possibility of correct usage were also assigned a score of zero. For example, in the
sentence ‘The trees and grass are grastify’ the nonsense word is either a noun or an
adjective but cannot be a verb as indicated by the suffix (-ify). Partial points were
assigned to ambiguous cases where the ambiguity included the possibility that the
suffixed word was functioning in accordance with the usual role of the suffix. For
example, the nonsense word in the sentence ‘I feel tropanity today’ could be either
a noun or an adjective and, since -ity indicates a noun, the production was given
partial points. Productions assigned partial points, however, were not included in
the analyses presented in this paper.
1.5. elimination of problematic elements. Some cases or items were eliminated
from the analyses presented in this paper. On a few items, several children produced
more than one sentence. This situation may have resulted from self-correction or
from the experimenter’s request to repeat a sentence or to use the word that was
given (rather than add an inflection such as using grastified instead of the given
item, grastify). These items were difficult to score because it was unclear whether to
count the first sentence, second sentence or the best sentence. For the purposes of
this paper, then, these items were eliminated from the analysis.
A number of cases were also eliminated from the analysis on the basis of
response bias. That is, some children showed a bias to form sentences with only one
kind of framework such as all adjective frames (‘He is very X’) or definition-type
noun frames such as ‘Tropanity is XYZ.’ Children who produced sentences with
only one kind of framework in 9 or 10 out of 10 possible cases were eliminated from
further analyses. In total, 54 cases were eliminated for this reason (49 children
showed a bias for adjective frames and 5 children showed a bias for noun frames).
One child’s data could not be analyzed because of a tape malfunction. As a result,
the items available for analysis were produced by the remaining 246 children.
194
martha smith & elena nicoladis
100
Children
Adults
90
80
70
60
50
40
30
20
10
0
-less -able
-ive
-ment -ness
- i t y -tion
-ize
-ate
-ify
Figure 1. Percentage of adults and children’s responses that were assigned full
points, by suffix.
2. analysis and results. Figure 1 addresses the main research question of whether
adults and 10-year old children use nonsense words with real English derivational
suffixes in the same type of sentence frames.
Figure 1 shows the percentage of adult and children productions that were
assigned full points. From this figure, it is clear that children had many fewer productions that were assigned full points than adults did. A series of Chi-square
analyses, comparing the ‘zero points’ and ‘full points’ responses by children and
adults showed that there was no significant difference in the distribution of
responses for the item suffixed in -less (χ2 = 0.46, n.s.). However, all other differences were highly significant.
From the data presented in Figure 1 it is clear that literate adults treated the nonsense words as having syntactic category information according to the appended
suffixes and (almost unanimously) they constructed sentences unambiguously using
the nonsense words as those respective parts of speech. The children’s morphosyntactic knowledge (that is, knowledge of syntactic category information in derivational suffixes), however, is not on a par with that of literate, educated adults. Quite
clearly, this aspect of linguistic knowledge is still developing in children of this age.
Some interesting differences emerged with respect to types of suffixes. Most
notably, mastery of verb suffixes appears consistently low while mastery of adjective
suffixes is consistently high (in one case, on a par with adult usage). The most variability in usage appears in the group of noun suffixes where children are doing better
with the suffixes -ity and -tion than with -ment and -ness. Interestingly, the pattern of
performance in the children’s data is mirrored in the adult data. For example, sentences containing the -ment item were assigned full points less often than sentences
containing the other suffixed items in both the child and adult data and both children
and adults obtained full points most frequently for the item suffixed in -tion.
195
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Verb
Adjective
Adjectives
Nouns
Ad
ul
ts
ld
re
n
Ch
i
Ad
ul
ts
ld
re
n
Ch
i
Ch
i
Ad
ul
ts
Noun
ld
re
n
Percentage of answers with no points
the developing lexicon
Verbs
Figure 2. Sentence slot used when no points given.
A second research question addressed the question of the kinds of ‘errors’ that
children and adults made. That is, when a production was scored as zero points,
how were the adults and children using the suffixed nonsense words? Did children
and adults make the same kinds of errors and was there anything systematic about
these productions? Figure 2 illustrates the answers to these questions.
Figure 2 shows how the participants used the suffixed nonsense words in productions that were scored as ‘no points.’ In interpreting this graph it is important to
recall that adults made very few ‘errors.’ Thus, the adult columns are based on only
a few cases whereas the children’s columns represent a substantial proportion of the
children’s productions.
The data in Figure 2 reveal similarities between the child and adult errors. In the
case of nonsense words ending in adjective suffixes (-less, -able, -ive), the predominant ‘error’ is use as a noun. This is clearly the case for both children and adults. In
the case of items with noun suffixes (-ment, -ness, -ity, -tion), the children’s errors
tend to be primarily adjectives. The similarity with the adult data is not revealed here
because adults made only three errors in total (one adjective and two verb usages).
Both the child and the adult data show that there is a slight preference for noun
frames in the ‘errors’ associated with items ending in verb suffixes (-ize, -ate, -ify).
These productions that were scored as ‘no points’ may reflect some kind of
knowledge about syntactic category information in suffixes. One possibility is that
these responses show (implicit) awareness that the category information in suffixes
is not invariant. For example, although -less is primarily an adjective suffix, words
ending in -less can sometimes be used as nouns as in the following sentences: The
people were homeless (Adj) and They gave food to the homeless (Noun). Similarly,
-ate is primarily a verb suffix but it is not uncommon to find words ending in
(orthographic) -ate used as nouns or adjectives as in the case of The graduate
(Noun) was compassionate (Adj). Thus, these errors may reflect implicit knowledge
196
martha smith & elena nicoladis
of possible roles that words with these suffixes can play and the data reveal that children’s errors are in the same direction as those of adults suggesting that children of
this age are acquiring this linguistic knowledge.
Several other analyses were conducted to address the question of whether participants were utilizing other information about lexical category in this task and
whether other differences could be detected between the adult and child data.
Specifically, we were interested in the quality of the productions that were assigned
full points. For example, the nonsense word ‘snarkness’ is used as a noun in the sentence That person has much snarkness (adult production) and in the sentence My
dog has a big snarkness (child production). Both of these productions were therefore assigned full points although, subjectively, there seems to be something qualitatively ‘better’ about the adult sentence. The question that arises, then is the
following: When children get full points for their productions, does it mean they
have attained adult-like proficiency?
In order to address the issue of production quality we considered other characteristics of one noun suffix (-ness) and two verb suffixes (-ize and -ify). In the first
case (noun suffix), an examination of a list of real words ending in -ness indicated
that this suffix is almost always used to form mass, rather than count, nouns (as
in the words correctness, forgiveness, illness and sadness). We questioned, therefore,
whether participants who used the nonsense word ending in the suffix -ness as a
noun (thereby getting full points) would also use the item as a mass (rather than
a count) noun. In addition, we questioned whether there would be any difference
in this respect between adults and children.
To address the questions concerning the count/mass noun distinction, we tabulated the instances where children and adults produced sentences using the nonsense word ending in -ness as a count noun and those where the item was used as a
mass noun. Of course, the fact that we were working with nonsense items required
an operational definition of ‘count’ and ‘mass’ noun. All productions using the -ness
nonsense word as a noun were coded in one of three ways: (1) The stimulus item
was considered a count noun if the article ‘a’ appeared before it, as in the sentence
That dog has a big snarkness. (2) The item was counted as a mass noun if no article
or the word ‘some’ or ‘much’ was used in front of the word as in the sentence I was
swimming and I made a lot of snarkness.’ (3) Items were counted as unclear if the
words ‘the’, ‘his’, ‘her’ or ‘their’ appeared in front of the word as in the sentence He
was scared of the snarkness. Only the clear cases of count and mass nouns were considered in the analysis.
Of the clear cases, almost all the adults and the majority of children used the
stimulus item as a mass (rather than a count) noun. The difference between adult
and child usage was not statistically significant (χ2 = 2.14, n.s.). This result implies
that adults and children who have learned the syntactic role of the -ness suffix have
also acquired the semantic nuance of the suffix (it is used on mass nouns). These
results are more suggestive than conclusive however as the operationalization of
‘count’ and ‘mass’ nouns was limited and the majority of cases were, in fact, unclear.
the developing lexicon
197
Thus, differences between child and adult productions may have been masked by
the limitations of the analysis. An examination of productions that were tabulated
as count nouns is suggestive. Very few adults used the -ness stimulus item as a count
noun (only 2 out of 29) but even those who did showed an appreciation of the
semantic nuance (The doctor responded with a snarkness in his voice. A snarkness
covered the valley and chilled us to the bone). In contrast, many child uses of the
stimulus item as a count noun lacked this appreciation of ‘snarkness’ as an intangible quality and indicated a countable, ‘thing-like’ nature (e.g. He has a very big
snarkness. I have a pet—it’s a snarkness. He was afraid of the shark so he sometimes
called it a snarkness instead of a shark). Further work is needed to explore the acquisition of semantic nuances of various suffixes but an examination of productions
suggests that 10 year old children have not yet attained adult-like proficiency even
if statistically significant differences were not detected in the child and adult data
noted here. The data also raise the issue of some syntactic distinctions being
acquired before some semantic nuances. It is possible, for example, that children
must first acquire (implicit) knowledge of the syntactic function of a suffix like
-ness (it marks a noun) before they learn the semantic nuances associated with the
suffix (it marks a ‘mass’ versus a ‘count’ noun).
A second analysis regarding qualitative differences in productions given full
points focused on two verb suffixes (-ize and -ify). An examination of real verbs
ending in these suffixes revealed that these suffixes are generally used on transitive
verbs (those requiring a direct object) as can be appreciated by an attempt to use
the following verbs in a sentence (categorize, neutralize, recognize, visualize, mystify,
specify, terrify, fortify). Since real words ending in these suffixes tend to be transitive,
we questioned whether those participants who used the nonsense words ending in
these suffixes as verbs would also use them with a direct object in their sentences.
Further, we questioned whether there would be discernible differences in how children and adults used the stimuli items (when they received full points for using the
items as verbs).
To address these questions, we first examined all cases where participants used a
nonsense word as a verb. Table 1 (overleaf) shows the tabulation of children and
adults who used nonsense words as verbs and the number of times the nonsense
verbs were used with and without direct objects. Three of the stimuli used as verbs
had verb suffixes (-ize, -ify and -ate); ‘other’ cases are stimuli not specifically suffixed with a verb ending but nonetheless used as a verb (e.g. I’ll tropanity that).
Table 1 suggests that adults have a preference for using verbs transitively whether
or not a direct object is required. Although the pattern is similar for children when
the stimuli require a direct object (-ize and -ify items), no such tendency appears in
the children’s data when the stimuli do not require a direct object. This suggests
that children are learning the argument structure associated with transitive verbs.
To examine this issue more closely, we focused on the two stimuli that represented
transitive verbs. We calculated how often a direct object appeared in sentences using
the nonsense words ending in -ize and -ify as verbs. Transitive use of the nonsense
198
martha smith & elena nicoladis
Stimuli
bromize
grastify
straferate
All others
Participants
Children
Adults
Children
Adults
Children
Adults
Children
Adults
No Direct Object
25
8
29
6
39
9
27
0
Direct Object
42
15
χ2 = 0.29
66
22
χ2 = 13.54
27
15
χ2 = 57.76
9
3
χ2 = 17.25
Table 1. Number of children and adults who used nonsense words as verbs with or
without direct objects.
verbs included sentences like ‘I can bromize words’ and ‘That graph will grastify your
sight.’ Productions given full points for verb usage but without a direct object
include sentences like ‘Let’s bromize together’ and ‘When reading maps, I always grastify.’ Figure 3 shows the percentage of children and adults who used a direct object
when they used a nonsense word ending in -ize and -ify as a verb.
Figure 3 reveals that the majority of both children and adults who used the stimuli as verbs also used them as transitive verbs. With an -ize nonsense word, children
formed transitive sentences equally as often as adults (the difference is not statistically significant, χ2 = 0.29, n.s.). The difference between children and adults is statistically significant in the case of the -ify stimuli (χ2 = 13.54, p < .01), indicating that
adults formed transitive sentences slightly more often than the children did.
Interestingly, these results suggest that, although the majority of children have not
acquired -ize and -ify as verb markers even by 10 years of age (see Figure 1), those
that did use these suffixed nonsense words as verbs also used them as transitive
verbs (Figure 3). Thus, along with the syntactic category information, information
about argument structure has been acquired by the minority of children that
(implicitly) recognized the nonsense words as verbs.
conclusion. The data from this study confirm that literate adults have strong intuitions about the syntactic functions of derivational suffixes and the semantic
nuances of these suffixes. This suggests that derivational suffixes may have some
status independent of stems in the mature mental lexicon.
Similarities and differences between the data produced by children and adults
suggest, however, that suffixes may only gradually attain a status independent of
stems in the mental lexicon. The data show, for example, that in many cases 10-year
old children did not use the suffixed nonsense word in accordance with the primary
syntactic role of the suffix. This suggests that these children did not understand
(implicitly) or did not have strong intuitions about the syntactic functions of
derivational suffixes. Further, even when children do show an appreciation of the
the developing lexicon
199
percentage with direct object
10 0
90
80
70
Children
Adults
60
50
40
30
20
10
0
-ize
-ify
Figure 3. The percentage of children and adults who used a direct object with two
of the nonsense verbs.
syntactic category suggested by the suffix, there may still be some nuances inherent
in suffixes that may not be mastered. So, for example, only approximately a third of
the children in this large sample appeared to have mastered verb suffixes (appreciated the syntactic category of the verb suffixes) and of those who did, some
appeared to still be learning the direct object distinction associated with those suffixes. However, once the syntactic category information is acquired, other semantic
nuances seem to be well underway suggesting a strong lexical-syntactic interface.
This study suggests that the lexicons of these children are developing like those
of educated adults. As a group, these children’s intuitions about syntactic category
are moving towards those displayed by the adult group. Similarly, intuitions about
subtle nuances (like mass noun, direct object distinctions) associated with real
words with these suffixes are evolving towards adult standards. Children’s errors on
this task are also in the same direction as adults suggesting that they are aware of
other information about suffixes that adults know such as the fact that the same
suffix can appear on words with different syntactic roles. While there is still much
to be learned about the subtle interactions involved in the lexical-syntactic interface, this study suggests that morphosyntactic knowledge in fifth grade children is
still developing and that intuitive awareness of syntactic category information plays
an important role in the growth of lexical knowledge.
1
The data for this research were collected as part of Martha Smith’s dissertation study
that was partially supported by a Spencer Mentor Grant to Catherine Snow.
200
martha smith & elena nicoladis
REFERENCES
Clark, Eve V. & Ruth A. Berman. 1984. Structure and use in the acquisition of
word formation. Language 60:542–90.
Clark, Eve V. & Sophia R. Cohen. 1984. Productivity and memory (vs. meaning)
for newly formed words. Journal of child language 2:611–25.
Clark, Eve V. & Barbara F. Hecht. 1982. Learning to coin agent and instrument
nouns. Cognition 12:1–24.
Grimshaw, Jane. 1990. Argument structure. Cambridge ma: MIT Press.
Henderson, Leslie. 1989. On mental representation of morphology and its diagnosis by measures of visual access speed. In Lexical representation and process,
ed. by William Marslen-Wilson, 357–91. Cambridge ma: MIT Press.
Jenkins, Joseph R. & Robert Dixon. 1983. Vocabulary learning. Contemporary
educational psychology 8:237–60.
Matthews, P. H. 1974. Morphology, 1st ed. New York: Cambridge University Press.
———. 1991. Morphology, 2nd ed. New York: Cambridge University Press.
Nagy, William E. & Richard C. Anderson. 1984. How many words are there in
printed school English? Reading research quarterly 19:304–30.
Nagy, William E. & Patricia A. Herman. 1987. Breadth and depth of vocabulary
knowledge: Implications for acquisition and instruction. In The nature of
vocabulary acquisition, ed. by Margaret G. McKeown & Mary E. Curtis, 19–35.
Hillsdale nj: Erlbaum.
———, ——— & Richard C. Anderson. 1985. Learning words from context.
Reading research quarterly 20:233–53.
Ravid, Dorit & Avraham Avidor. 1998. Acquisition of derived nominals in
Hebrew: Developmental and linguistic principles. Journal of child language 25:
229–66.
Smith, Martha L. 1998. Sense and sensitivity: An investigation into fifth grade
children’s knowledge of English derivational morphology and its relationship to
vocabulary and reading ability. Cambridge ma: Harvard Graduate School of
Education dissertation.
Tyler, Andrea & William Nagy. 1989. The acquisition of English derivational
morphology. Journal of memory and language 28:649–67.