The Role of Syntax in Reading Comprehension PDF
The Role of Syntax in Reading Comprehension PDF
1. General introduction
English language learners (ELLs) from Spanish speaking homes tend to have comparatively low
literacy achievements as early as first grade and continue to lag behind their English speaking peers
throughout the school years, even when instructed and assessed in Spanish (CTB/McGraw Hill, 1982,
1988; De la Rosa and Maw, 1990; Orfield, 1986). This lag in reading skills manifested in the earlier
grades is exacerbated at the middle school and high school levels when it is critical for students to
understand and manipulate large volumes of written text to learn subject matter. The cognitive
prerequisites for successful reading comprehension are already complex even when we consider this
process for monolingual children reading in their native language, particularly when they come from
low income families. In a widely cited study, Chall and Jacobs (1996) reported that by the time these
children reach fourth grade, their reading scores begin to decrease and continue to do so for the next
five years, a phenomenon that has been termed the fourth grade slump.
For inner city bilingual and bidialectal children, learning to read is even more complex, as they
have to negotiate two linguistic systems, acquire reading skills in a language not spoken at home, and
face the challenges of an overburdened public school system (NYT, March 28, 2002). In a report on
the necessity of research on reading comprehension (Snow 2002), the Rand Reading Study Group
points out that in order to successfully negotiate textual meaning the reader must bring at least the
following to the act of reading: cognitive capabilities (e.g. attention, memory), motivation (e.g.
purpose, interest), linguistic knowledge and experiences. Yet educators do not understand these
factors sufficiently, especially in the case of second language readers: the education field [does
not] know how to limit the particular challenges that second language readers face due to those
readers limited vocabulary and linguistic knowledge, nor do educators know how to build on those
readers first language comprehension abilities. (pg. xiv)
In this study, we look at the relationship between emerging language knowledge and reading skills
in the bilingual child. In particular, we investigate the role of the bilingual childs syntactic systems in
the emergence of reading readiness. Our focus is on reading comprehension and more specifically its
precursor skill, listening comprehension. The comprehension of written and aural text is an area of
literacy development that has received relatively little attention, especially when compared to the
investigation of decoding skills. Gough and Tunmer (1986) and Tunmer and Hoover (1993) were
among the first reading theorists to identify two main areas of cognition that contribute to the ability to
read and understand written text: decoding skills, consisting of the ability to identify speech sounds
and link these to individual letters; and listening comprehension, which is based on the readers ability
to recruit her mental grammar of the language and process sentences. Since syntax is a significant
component in processing and at the same time a domain of language that reading researchers are only
beginning to investigate, we have made it the main focus of our study. In the following section we
report some of the findings on the relationship between syntax and developing reading skills in
monolingual children.
2005 Gita Martohardjono, Ricardo Otheguy, Alison Gabriele, Michele de Goeas-Malone, Malgosia
Szupica-Pyrzanowski, Erika Troseth, Silvia Rivero, and Zoe Schutzman. ISB4: Proceedings of the 4th
International Symposium on Bilingualism, ed. James Cohen, Kara T. McAlister, Kellie Rolstad, and Jeff
MacSwan, 1522-1544. Somerville, MA: Cascadilla Press.
1523
Stein et al. (1984) who found that nondisabled readers demonstrated higher performance on complex
syntactic structures, such as adverbial and relative clauses, than reading disabled children. According
to Stein et al these results suggest that nondisabled children performed more like adults, in that they
were basing their interpretation of inherently complex sentences on a hierarchical, adultlike, structural
analysis, as opposed to the linear one characteristic of early language development. For example, they
found that there was an order of difficulty in the interpretation of relative clauses (OO > SO and OS)
for reading disabled children, while no implied order of difficulty was found for the comparison group.
Stein et al. defined the reading impaired childrens linguistic system as being partially delayed and
conclude that the language comprehension system of unskilled readers mirrors a deficit in
grammatical maturation and thus must rely on early forms of sentence interpretation or lower levels
of language processing.
Within the population of nondisabled readers similar findings have been presented by researchers
investigating the relationship between syntactic measures and reading skills. Tunmer, Nesdale and
Wright (1987) compared good, younger readers (in grade two) to poor, older readers (in grade four)
on four measures of reading ability (real word recognition, pseudo-word naming, reading fluency and
reading comprehension) as well as verbal intelligence. Tunmer et al. hypothesized that syntactic
awareness is causally associated with learning to read in two ways: First, syntactic awareness may
significantly aid the child in acquiring phonological recoding, which is understood as the ability to
translate letters into phonological form. This skill may enable beginning readers to recognize new
words, develop speed and automaticity in visual word recognition and indirectly support
comprehension. Second, it is plausible that syntactic awareness enables beginning readers to monitor
their comprehension processes more efficiently. The results of Tunmer et al. indicate that good,
younger readers scored significantly better than poor, older readers on two tests of syntactic awareness,
the oral cloze task and oral correction task. This further suggests that the older, unskilled readers were
developmentally delayed in syntactic awareness and that this delay may have altered reading
development. Compatible with this interpretation are the subsequent findings that the two measures of
syntactic awareness varied with reading level at each grade: the better readers of each grade scored
better on syntactic awareness tasks than the poor readers. In Tunmer et al.s view it is the
combination of both results, the higher performance of the good, young readers and the differences
among the chronological age matches, that points to a causal link between syntactic awareness and
reading acquisition (see also Tunmer and Bowey, 1984).
Bowey (1986a) presented fourth and fifth graders with two parallel sets of syntactically incorrect
sentences. Participants were told that all stimuli they would hear contained a mistake. In an elicited
imitation task participants were asked to repeat the sentence with the mistake, and in an elicited
correction task they were asked to fix the sentence. The difference between the number of elicited and
spontaneous corrections in both tasks was calculated and labeled a syntactic control, a nearly pure
measure of syntactic awareness. The results suggested that although this syntactic control did not
increase from fourth to fifth grade it was strongly correlated with both product and process measures
of reading proficiency as measured by standardized reading test performance as well the ability to
control and identify the acceptability of oral reading errors.
Boweys (1986a) results concerning the relation between syntactic awareness and reading
achievement in fourth and fifth graders were replicated, this time in a sample population of first to fifth
grade children, with both verbal ability and grade effects partialed out. Inspired by her previous
results, Bowey (1986b) investigated the development of metasyntactic skill (e.g. childrens ability to
correct grammatically incorrect sentences) and its relation to reading achievement.
Syntactic
awareness was significantly related to reading achievement. The significant correlation between
reading ability and deviant sentence recall, with random sentence recall effects statistically controlled,
remained significant with vocabulary, age and grade effects additionally controlled.
While most of the studies described above measured syntax in global ways, some studies have
focused on more specific sentence types. Waltzman and Cairns (2000), for example, looked at binding
and control and the reading abilities of good and poor readers in third grade and found that good and
poor readers differed with respect to their interpretation of pronominal relations. Overall the good
readers performed more adultlike (99% correct) than the poor readers (83%) on a comprehension task.
They also found a significant correlation between their independent measure of reading and knowledge
1524
of grammar. Waltzman and Cairns argue that it is highly unlikely that deviant (non adult) responses
of the children in their study could be associated with obstructions in phonological memory. The
methodology of their study was designed in such a way so as to put as little strain on phonological
short term memory as possible. The results further support the role of specific syntactic factors that are
independent of phonological short term memory and that underlie initial reading skills.
1525
Headless relative clauses (Look it mommy have on, OGrady, 1997:175) have been
documented in the speech of 2 year olds, and headed relatives (I want something that the cows eat
OGrady, 1997:175) in the speech of children between 2;6 and 3;0 (example from Hamburger and
Crain 1982:248). Subjectless infinitivals (I wanna cookie) are likely produced before the connective
and, but subordinate clauses with relative pronouns and subordinate clauses with temporal adverbials
are produced later in development. Although the order in which connectives are acquired varies, the
adverbial connectives before and after are likely two of the more difficult to acquire (Bowerman,
1979). Before and after may emerge as temporal adjuncts prior to their successful use as connectives.
1526
Cairns and Fiengo (1985) used an act out task to test 64 monolingual English speaking children ages
3;2 to 8;3 on sentences with relative and adverbial clauses. When presented with OS sentences of the
type The lion pushes the bear that is climbing up the ladder 90 percent of the children age 5;7 to 6;0
correctly chose the object of the matrix (the bear) to be the subject of the subordinate clause. 10
percent of them incorrectly chose the subject of the matrix (the lion) to be the subject of the
subordinate clause.
A leading hypothesis is simply that the more embedded a gap, the more difficult it is to interpret
it. An order of difficulty in comprehending gaps was proposed by de Villiers et al. (1979). Based on an
act out task, she found a hierarchy with subject gaps being easiest, object gaps being more difficult,
and object of preposition gaps being the most difficult. In a picture description task in which
production of relative clauses was elicited, Prez Leroux (1995) found resumptive pronouns filling in
object gaps, but no resumptive pronouns being used to fill in subject gaps. The filling in of gaps was
taken to be an indication of the difficulty of interpreting the gap.
1527
stage, children preferred the object of the matrix as the controller; in the third stage children sometimes
selected the matrix object and sometimes selected the matrix subject; and in the fourth stage, children
selected the correct controller (object for the tell complements, subject for the adverbials).
Hsu, Cairns, and Fiengo (1985) also propose a developmental hierarchy, largely paralleling the
hierarchy proposed by Chomsky (1969). Hsu et al. tested temporal adverbial clauses and control with
sentences like The boy hits the girl after jumping over the fence. Children ages 3;2 to 8;3 participated
in the study. The four stages Hsu et al. propose are: object oriented; mixed subject and object oriented;
approaching adult; and adult. A subject oriented stage or strategy (Tavakolian, 1977; Goodluck and
Roeper, 1978) is not uncontroversial, but Hsu et al. propose that it may precede the object stage
described in their study. The object oriented grammar was characterized as the minimal distance
principle in Chomsky 1969. Hsu et al. provide a more structural characterization of this phenomenon,
proposing that the closest c commanding NP controls the missing subject (thus, children who
misinterpret the object as the controller have not acquired the adult structure for these types of
sentences).
A coordinate strategy may also be employed in the acquisition of temporal adverbials (Cairns,
McDaniel, Hsu, DeFino, and Konstantyn, 1995) before the child learns to use words like before. As
mentioned earlier, before is one of the later connectives to emerge in language development.
4.2 Methodology
4.2.1 Subject/participant participant selection
All participants were kindergarteners from one of two New York City public elementary schools.
22 participants completed three syntax measures in Spanish and English and a standardized English
pre-reading test. In addition to the subordination and coordination syntax measure reported on here, a
1528
tense and aspect task (picture point) and a pronominal awareness task (picture point and act out) were
administered. 22 participants completed all tasks, and the results of 13 of those 22 are reported here.
For each of these 13 children, their scores on the Spanish versions of the syntax measures numerically
exceeded their scores on the English versions of the syntax measures. Analyzed as a group, the
Spanish score on the syntax measures is statistically higher than the English score on the syntax
measures. The mean age of the 13 participants was 5;9.
A questionnaire was conducted with a parent of each participant (in most cases, the mother). To
assess the level of literacy activities in the home and that the children participated in outside of school,
we asked questions like Did you bring your child to the library to borrow books any time during the
past two weeks?, Is there pencil and paper in your house so that the child can write?, and Do you read
the newspaper (a lot, sometimes, little, almost never)? Possible scores ranged from 0 percent to 100
percent. Interviews were conducted with parents of 12 of the 13 participants discussed here. The
literacy score ranged from 33 percent to 60 percent; the mean was 44 percent.
1529
(intransitive and transitive) where the order of events matches the temporal order of the sentence; and
OS (relativized object, subject relative pronoun) relative clauses (intransitive and transitive). When a
transitive verb was used, the object was inanimate.
Sentences with relative clauses and sentences with embedded clauses missing explicit subjects are
well researched in first language acquisition research and present a challenge to young learners
(Goodluck and Tavakolian, 1982; Cairns, McDaniel, Hsu, and Rapp, 1994). We tested only the
simplest relative clause type (OS), and the simplest temporal adverbial clause type (clause order
matched temporal order). Work has been done on the acquisition of tensed subordinate clauses (Cairns,
McDaniel, Hsu, and Rapp (1994); McDaniel et al. (1991), Lust et al. 1986), but we did not test this
structure here. The overt pronoun in these structures is ambiguous (may be controlled by the subject or
the object), while in the missing subject clauses tested here, there is always a correct and incorrect
response. An interesting observation to come out of research in tensed adverbials is the Pronoun
Coreference Requirement. Children who obey the PCR require that a pronoun be coreferential with a
noun phrase in that sentence (in other words, a pronoun may not refer to an agent not mentioned in the
sentence). However, the PCR seems to be a factor in the grammar of very poor readers, not normally
developing readers. Simple transitive sentences were used as fillers.
Examples of sentence types
Simple Transitive (filler)
The cat pushes the box.
El gato empuja la caja.
IP Coordination (intransitive)
The cat jumps and the bear runs.
El gato salta y el oso corre.
Object Coordination (transitive)
The monkey touches the bear and the cat.
El mono toca al oso y al gato.
Subject Coordination (intransitive)
The dog and the cat sleep.
El perro y el gato duermen.
Subordination (relative clause, intransitive)
The monkey hugs the dog who jumps.
El mono abraza al perro que salta.
Subordination (relative clause, transitive)
The bear hugs the cat who pushes the box.
El oso abraza al gato que empuja la caja.
Subordination (adverbial clause, intransitive)
The monkey pushes the dog before dancing.
El mono empuja al perro antes de bailar.
Subordination (adverbial clause, transitive)
The dog punches the bear before touching the box.
El perro golpea al oso antes de tocar la caja.
4.2.2.2 Scoring
A very conservative scoring procedure was used. No animal substitution errors or verb
substitution errors were permitted. A response counted as correct only if the entire sentence was acted
out as read to the child with no lexical errors.
1530
choice test is administered orally, and participants respond in a booklet, marking down their answers in
pencil. A second experimenter was present to make sure participants remained on task. At least one
day separated each session.
4.2.3.2 Scoring
The multiple choice test was scored using the key provided by Gates. Scores were totaled by
subsection and overall. Scores reported are mean percent correct.
5. Developmental results
5.1 Research questions
In addition to our primary investigation into the relationship between syntax and reading, we also
wanted to characterize the development of our participants L1 and L2. First we asked whether the
participants would display the same developmental patterns that have been described in studies of
monolinguals (primarily English-learning monolinguals). Second, we were interested in whether the
patterns they displayed in their L1 (Spanish) would similar to the patterns they displayed in the L2
(English).
Coordination
Subordination
Overall
L1 (Spanish)
85%
15%
45%
L2 (English)
71%
13%
38%
Figure 1. Mean percent correct overall and by sentence type on syntax measure
As you can see in the table above, performance on coordination in Spanish (85% correct) was
better than performance on subordination in Spanish (15% correct) (F (1,12) = 230.27, p < 0.01), and
performance on coordination in English (71% correct) was better than performance on subordination
in English (13% correct) (F (1,12) = 74.95, p < 0.01). Performance on Spanish coordination exceeded
performance on English coordination by a statistically significant amount (F (1,12) = 5.60, p < 0.05).
Performance on subordination in the two languages was similar. That is, although performance on
Spanish subordination exceeded performance on English subordination, the amount was not
statistically significant. Overall performance (coordination and subordination combined) in Spanish
(45% correct) was better than overall performance in English (38% correct) by a statistically
significant amount (F (1,12) =4.78, p < 0.05).
1531
L1 (Spanish)
94%
L2 (English)
94%
83%
58%
76%
62%
L1 (Spanish)
18%
23%
L2 (English)
19%
29%
13%
10%
12%
13%
7%
10%
10%
4%
1532
Spanish is 23% correct; the score on relative clauses with transitive verbs in Spanish is 13% correct.
The score on relative clauses with intransitive verbs in English is 29% correct; the score on relative
clauses with transitive verbs in English is 10% correct. The overall score on temporal adverbial clauses
in Spanish is 12% correct; the overall score on temporal adverbial clauses in English is 7% correct.
The score on temporal adverbial clauses with intransitive verbs in Spanish is 13% correct; the score on
temporal adverbial clauses with transitive verbs in Spanish is 10% correct. The score on temporal
adverbial clauses with intransitive verbs in English is 10% correct; the score on temporal adverbial
clauses with transitive verbs in English is 4% correct.
Our participants performed at about the same level when presented with subordinate clauses
(relative clauses and temporal adverbial clauses) in the L1 (Spanish) and the L2 (English). The
difference between the two was not statistically significant. In both languages, performance on relative
clauses with intransitive verbs was statistically significantly higher than performance on relative
clauses with transitive verbs. This is true on both Spanish (F (1,12) = 4.55, p= 0.054) and English (F
(1, 12) = 8.96, p < 0.05).
L1 (Spanish)
15%
34%
22%
29%
100%
Total
L2 (English)
13%
36%
15%
36%
100%
Figure 4. Responses Given for Sentences with Subordinate Clauses (Adverbial and RC) in L1 and L2
1533
participants interpret the relative clause as unnecessary to act out. OGrady (1997) reported that adults
sometimes omit the subordinate clause when asked to act out such sentences. For example, when asked
to act out a sentence such as The dog that the horse kissed jumped on the camel, adults took the dog,
and assumed that they selected the dog that the horse had kissed (but would not actually act out this
part of the sentence), and then had the dog jump on the camel.
Ten of the 13 participants made at least one reduction error, and seven of those 10 made a
reduction error in Spanish and English. Three types of reduction errors were observed: reduction with
no errors in the matrix performed; reduction with only a verb substitution in the matrix; and reduction
with further lexical errors (for example, selection of an incorrect subject or object). Looking only at
reduction errors, the table below shows the percentage of each error type.
Reduction Error Type
Matrix clause correct (reduction only)
Verb substitution
Further lexical errors
L1 (Spanish)
60%
21%
19%
100%
Total
L2 (English)
62%
15%
23%
100%
Figure 5. Percentage of reduction error types produced for all reduction errors on subordinate clauses
1534
structures are often tested in studies of monolinguals, but in this task only the simplest of these
structures was tested. With relative clauses, the simplest structure is sometimes referred to as object
control since it is the object of the matrix that is the subject of the relative clause.
As with the errors made in adverbial clauses, the errors our participants made when presented with
relative clauses largely pattern with errors described in studies of monolinguals. In no case (as with the
adverbials) did a participant select a character from outside the sentence to be the subject of the
relative clause. Participants often incorrectly selected the subject of the matrix to be the subject of the
relative clause. And as with the adverbials, some participants allowed the subject and object of the
matrix to simultaneously serve as the subject of the relative clause. Again, this coordinate NP error is
not one we have seen described in studies of monolinguals, and we hypothesize that the participants
who made this error may be providing evidence for an intermediate stage in development.
1535
Total
53
10.36
1536
100
90
80
70
60
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
100
90
80
70
60
50
40
30
20
10
0
0
10
20
30
40
50
60
70
80
90
100
7. Discussion
7.1 Developmental results
In our investigation of syntactic development in bilingual children we first asked whether or not
the same patterns observed in monolingual children can be discerned. With regard to coordination and
subordination, we can conclude that the acquisition sequence is the same, since our participants
1537
performed much better on coordination than they did on subordination. Furthermore, since this result
was found for both the L1 (Spanish) and the L2 (English), we can conclude that at least for the
sentence types we investigated, there is parallel development in the two languages of the bilingual
child. One reason for the similarity between L1 and L2 may be that the structures tested (coordination,
relative clauses and adverbial clauses) are nearly identical in Spanish and English. In future research, it
would be interesting to study the development of L2 reading skills when the L1 and L2 are
typologically very different (English and Japanese, for example). However, at this time, our concern
for the academic challenges faced by the population we tested override concerns for investigating
bilingual acquisition when the L1 and L2 are structurally dissimilar.
One result that pointed to a divergence from the monolingual pattern was our participants
superior performance on subject over sentence coordination. Recall that both Lust and Mervis (1980)
and Ardery (1980) found sentences like The dog kissed the horse and pushed the tiger easier than a
sentence like The tiger and the turtle pushed the dog. Our subjects, on the other hand, performed at
much higher levels on subject than on sentence coordination (94% vs. 76% in their L1). This may in
part be due to the fact that our subject coordination sentences contained intransitive verbs, rather than
the transitive verbs used in both Lust and Mervis and Arderys studies, thus not allowing for a direct
comparison between the monolinguals in their study and the bilinguals in ours. Note, however, that
our participants found object coordination, which necessarily contains transitive verbs, significantly
easier than sentence coordination, which contains intransitive verbs. This, together with the subject
coordination result, suggests that for our bilinguals the facility lay not in the type of verb used
(transitive vs. intransitive) but rather that the coordination of predicates (VPs) requires more
processing and resources than coordinating NPs.
A striking difference between monolingual children and our bilingual participants is the latters
overall low performance on subordination. The age of the monolingual children tested on coordination
and subordination ranged from about 3 to 5. The mean age of our participants was 5;9. On the OS
relative clauses, for example, our participants performed at 18% in their L1 (Spanish), compared to
70% for Goodluck and Tavakolians (1982) younger monolinguals. Clearly, our bilinguals lag behind
the monolinguals tested in the studies we reported earlier.
Although this may be taken as support for the claim that growing up bilingual somehow retards
cognitive development (see discussion in Hakuta 1986, Grosjean 1982), we believe it is instead a
reflection of the socioeconomic status of our particular population. First, our participants come from
inner-city schools where around 98% of the children are eligible for free lunchan indication of their
families low income level. Second, our participants all come from immigrant families, where in
general both parents work out of the house. Some of the parents of our participants did not go beyond
an elementary school education themselves. It is fair to assume that the general living and working
conditions of these families militate against providing their children with the optimal support for
language or literacy skills. Such unfavourable conditions stand in stark contrast to those of the
families of monolingual child populations typically used in L1 developmental studies who usually
have mid to high income levels. From a questionnaire we administered to assess the level of literacy
activities in the home we found that the literacy indices of our participants families was fairly low.
The questionnaire contained questions like Did you bring your child to the library to borrow books any
time during the past two weeks?, Is there pencil and paper in your house so that the child can write?,
and Do you read the newspaper (a lot, sometimes, little, almost never)? Possible scores ranged from 0
percent to 100 percent. Interviews were conducted with parents of 12 of the 13 participants discussed
here, and the literacy score ranged from 33 percent to 60 percent with a mean of 44 percent.
1538
comprehension skills. However, further analyses have provided us with strong evidence that this
correlation is clearly meaningful. First, all sub-sections of the Gates MacGinitie test are administered
orally, thus all sections, not just the Listening Comprehension subtest, recruit learners listening
comprehension skills. In addition, all sub-sections make use of relative clauses, one of our target
structures, in the instructions to the test-taker. For example, in the Oral Language Concepts sub-test,
which targets phonological awareness, a typical question asks the child to Listen for the sound that
begins with the same sound as dance. Despite the fact that all four sub-tests require listening
comprehension skills and comprehension of relative clauses, remarkably the only sub-tests which
correlated significantly with our syntax measure was the test of Listening Comprehension and the test
of Literacy Concepts. The Literacy Concepts subtest includes items that test childrens knowledge of
sequencing vocabulary. For example, children are instructed to Find the letter at the beginning of the
word. Knowledge of sequencing could clearly be connected to childrens ability to process structures
such as the temporal adverbial clauses that we tested in our test of subordination. There is a
straightforward explanation for the significant correlation between the two measures.
Importantly, there was not a significant correlation between knowledge of coordination and
subordination (in either the L1 or L2) and either the Oral Language Concepts subtest or the Letters and
Letter-Sound Correspondences sub-test of the Gates. These results suggest that what is correlating in
our results is not simply general listening skills, but rather knowledge of specific syntactic structures
and listening comprehension. Contrary to what has been argued by supporters of the sound based
Processing Deficit Hypothesis for example, our results strongly suggest that phonemic awareness and
word level decoding may not be as strongly related to syntactic processing as is currently believed (cf.
Crain and Shankweiler, 1986; Shankweiler and Crain, 1986, among others).
Additional analyses have suggested that it is not knowledge of syntax in general that is a good
predictor of pre-reading skills, but particularly knowledge of complex sentence structures. As we
mentioned earlier, participants were also tested on other measures of syntactic knowledge, specifically
tense and aspect and referential awareness. While we do not have the space to discuss the results of
these measures in detail, what is relevant to the present discussion is that these measures did not
indicate as strong of relationship between knowledge of syntax and pre-reading skills. In particular, in
the test of referential awareness, knowledge of reflexive pronouns and knowledge of personal
pronouns did not correlate significantly with listening comprehension. We tested referential awareness
using both picture-point and act-out tasks. These results provide further evidence that the relationship
we report cannot be reduced to the relationship between general listening skills. Furthermore, the noncorrelations with the test of referential awareness suggest that it is a certain type of syntactic
knowledge that is important for listening comprehension, specifically it is the ability to process
complex structures such as subordinate clauses, that might be directly relevant to the task.
Our results suggest that models of reading instruction which focus particularly on skills based
acquisition may be particularly effective in the population we tested. Models of reading instruction
which purposely avoid skills acquisition, and focus instead on whole pieces of literature and integrated
language experiences (cf. the vast literature on the Whole Language approach) may not serve the best
interests of the students. It has been argued in the educational literature that non-skills based
approaches are not effective in bilingual students or students of low socioeconomic backgrounds (cf.
Jeynes and Littell, 2000; de la Reyes, 1992). Our results provide further empirical support for their
arguments.
While the general relationship between syntactic skills and reading ability has been previously
documented for monolingual readers, our study is unique in that it examines this relationship in a
group of bilingual readers and investigates the relative contribution of knowledge of the first and
second language syntax to reading in the L2. Our main question was whether bilingual children with
a strong knowledge base in their first language would acquire pre-reading skills with greater success
than bilingual children whose knowledge of Spanish syntax was not as strong. Our results showed a
surprising role for the L1 syntax with respect to L2 reading skills. The L1 syntax measure (Spanish)
correlated with L2 (English) listening comprehension (r=.87, p=0.000), and the L2 syntax measure
(English) also correlated with L2 (English) listening comprehension (r=.70, p=0.007). However, the r
values shown above indicate that L1 (Spanish) syntax is the stronger contributor to L2 (English)
listening comprehension.
1539
A strong relationship was also observed between performance on sentences with subordinate
clauses and listening comprehension. The L1 (Spanish) subordination scores correlated with L2
(English) listening comprehension (r=.78, p=0.002), and the L2 (English) subordination scores also
correlated with L2 (English) listening comprehension (r=.57, p=0.038). However, the r values shown
above indicate that L1 (Spanish) subordination is the stronger contributor to L2 (English) listening
comprehension.
To further investigate the contribution made by the L1 syntax and the L2 syntax to L2 reading, we
ran a step-wise regression. Spanish (L1) subordination (mean percent correct) and English (L2)
subordination (mean percent correct) were the independent variables, and the dependent variable was
mean percent correct on the listening comprehension section of the (English) Gates MacGinitie test.
The two dependent variables accounted for 80 percent of the variance (R = .802) in the English
listening comprehension scores. When the Spanish subordination score was entered as the first
independent variable, adding the English subordination score explained an additional 3 percent of the
variance ( R Square = .609 with the Spanish subordination score only; R Square = .643 with the
addition of the English subordination score).
When the English subordination score was entered as the first independent variable, adding the
Spanish subordination score accounted for an additional 31 of the variance (R Square = .336 with the
English subordination score only; R Square = .643 with the addition of the Spanish subordination
score). Although the number of subjects is small (N=13), these regressions again point to the L1 as the
stronger contributor of success in L2 reading.
Previous research has shown that language development suffers when the educational
environment excludes the childs native language (or the language that is spoken at home). SkutnabbKangas and Toukomaa (1976) discuss the challenges in acquiring the L2 when development of the L1
is not strong. The threshold theory or threshold hypothesis (Toukomaa and Skutnabb-Kangas (1977),
Cummins (1976, 1979) puts forth the idea that a minimal threshold in the L1 must first be attained if
negative effects on cognitive development are to be avoided. Furthermore, the attainment of a second,
higher threshold is expected to have positive effects on cognitive development. Hoffman (1991) notes
that this model may lead to different results for different populations. If development of the L1 is
below the lower threshold, instruction only in the L2 may be detrimental. However, if development of
the L1 is already very strong, instruction in the L2 may not have detrimental effects.
Although we did not set out to test the ideas underlying the Threshold Hypothesis, the results
obtained in this investigation do point to a need for L1 support in the L2 classroom. While we found
that for our bilinguals the L1 and L2 are developing according to the hierarchies described in research
on monolinguals, their actual performance, even in the L1, was far below the performance of
monolinguals at younger ages. It seems quite likely that our participants are in danger of not attaining
the minimum level of proficiency in the L1 that is needed to avoid detrimental developmental effects
when learning the L2. We therefore take our results to suggest that reinforcing syntactic knowledge of
the L1 will have positive pedagogical effects on reading in the L2.
References
Adjmian, C. & Liceras, J. (1984). Accounting for adult acquisition of relative clauses: universal grammar, L1,
and structuring of the intake. In F. R. Eckman, L. H. Bell & D. Nelson (eds.), Universals of second language
acquisition, pp. 101-118. Rowley, Mass: Newbury.
Anderson, J. (1982). Language form and linguistic variation: papers dedicated to Angus McIntosh. John
Benjamins.
Ardery, G. (1980). On coordination in child language. Journal of Child Language, 7, 305-320.
Bentin, S., Deutsch, A. & Liberman, I. Y. (1990). Syntactic competence and reading ability in children. Journal of
Experimental Child Psychology, 48, 147-172.
Bever, T. (1970). The cognitive basis for linguistic structures. In J. R. Hayes (ed.), Cognition and the development
of language, pp. 274-353. New York: Wiley.
Bloom, L., Lahey, M., Hood, L., Lifter, K. & Fiess, K. (1980). Complex sentences: acquisition of syntactic
connectives and the semantic relations they encode. Journal of Child Language, 7, 235-261.
1540
Bowerman, M. (1979). The acquisition of complex sentences. In P. Fletcher & M. Garman (eds.), Language
Acquisition, pp. 285-306. New York: Cambridge University Press.
Bowey, J. A. (1986a). Syntactic awareness and verbal performance from preschool to fifth grade. Journal of
Psycholinguistic Research, 15 (4), 285-308.
Bowey, J. A. (1986b). Syntactic awareness in relation to reading skill and ongoing reading comprehension
monitoring. Journal of Experimental Child Psychology, 41 (2), 282-299.
Byrne, B. (1981a). Reading disability, linguistic access and short-term memory: comments prompted by Jorms
review of developmental dyslexia. Australian Journal of Psychology, 33 (1), 83-95.
Byrne, B. (1981b). Deficient syntactic control in poor readers: is a weak phonetic memory code responsible?
Applied Psycholinguistics, 2, 201-212.
Cairns, H. S., McDaniel, D., Hsu, J. R. & Rapp, M. (1994). A longitudinal study of principles of control and
pronominal reference in child English. Language, 70 (2), 260-288.
Cairns, H. S., McDaniel, D., Hsu, J. R., DeFino, S. P., Konstantyn, D. (1995). Grammatical and discourse
principles in childrens grammars: the pronoun coreference requirement. CUNYForum, 19, PhD Program in
Linguistics, CUNY Graduate Center.
Chall, J. S. & Jacobs, V. A. (1996). Stages of reading development (2nd ed.). Fort Worth, TX: Harcourt Brace.
Chomksy, C. (1969). The acquisition of syntax in children from 5 to 10. Cambridge: MIT Press.
Clark, E. V. (1971). On the acquisition of the meaning of before and after. Journal of Verbal Learning and Verbal
Behavior, 10, 266-275.
Coker, P. L. (1978). Syntactic and semantic factors in the acquisition of before and after. Journal of Child
Language, 5, 261-277.
Coker, P. L. & Legum, S. E. (1975). An empirical test of semantic hypotheses relevant to the language of young
children. In Working Papers on the Kindergarten Program: Quality Assurance, SWRL for Educational
Research and Development, Los Alamitos, CA.
Crain, S. & Shankweiler, D. (1986). Syntactic complexity and reading acquisition. Haskins Laboratories Status
Report on Speech Research, 86-87, 199-221.
Crain, S. & Shankweiler, D. (1988). Syntactic complexity and reading acquisition. In A. Davison & G. M. Green
(eds.), Linguistic complexity and test comprehension: readability issues reconsidered. Hillsdale, N.J.:
Erlbaum.
Cromer, W. & Wiener, M. (1966). Idiosyncratic response patterns among good and poor readers. Journal of
Consulting Psychology, 30, 1-10.
CTB/McGraw-Hill (1982). CTBS: comprehensive test of basic skills (forms U and V). Norms book, grades K-3.
Monterey, CA: CTB/McGraw-Hill.
CTB/McGraw-Hill (1988). SABE: Spanish assessment of basic education (technical report). Monterey, CA:
CCTB/McGraw-Hill.
Cummins, J. (1976). The influence of bilingualism on cognitive growth: a synthesis of research findings and
explanatory hypothesis. Working Papers on Bilingualism, 9, 1-43.
Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children. Review of
Educational Research, 40 (2), 222-251.
Cummins, J. (1981). The role of primary language development in promoting educational success for language
minority students. In California State Department of Education.
De la Reyes, M. (1992). Challenging venerable assumptions: literacy instruction for linguistically different
students. Harvard Educational Review, 62 (4), 427-446.
De la Rosa, D. & Maw, C. (1990). Hispanic education: a statistical portrait. Washington, DC: National Council
of La Raza.
De Villiers, J. G., Tager Flusberg, H. B., Hakuta, K. & Cohen, M. (1979). Childrens comprehension of relative
clauses. Journal of Psycholinguistic Research, 8 (5), 499-518.
Doughty, C. (1991). Second language instruction does make a difference: evidence from an empirical study of SL
relativization. Studies in Second Language Acquisition, 13 (4), 431-469.
Eckman, F. R., Bell, L. H. & Nelson, D. (1988). On the generalization of relative clause instruction in the
acquisition of English as a second language. Applied Linguistics, 9 (1), 1-20.
Flynn, S. (1983). Study of the effects of principal branching direction in second language acquisition: the
generalization of a parameter of universal grammar from first to second language acquisition. PhD
dissertation, Cornell University, New York.
1541
1542
Macaruso, P., Bar-Shalom, E., Crain, S. & Shankweiler, D. (1989). Comprehension of temporal terms by good
and poor readers. Language and Speech, 32 (1), 45-67.
McDaniel D., Cairns, H. S. & Hsu, J. R. (1990/91). Control principles in the grammars of young children.
Language Acquisition 1 (1), 121-139.
Menyuk, P. (1969). Sentences children use. Cambridge: MIT Press.
Menyuk, P., Chesnick, M., Liebergott, J. W., Korngold, B., DAgostino, R. & Belanger, A. (1991). Predicting
reading problems in at-risk children. Journal of Speech and Hearing Research, 34 (4), 893-903.
Munnich, E., Flynn, S. & Martohardjono, G. (1994). Elicited imitation and grammaticality judgment tasks: what
they measure and how they relate to each other. In E. E. Tarone, S. M. Gass & A. D. Cohen, (eds.), Research
methodology in second-language acquisition, pp. 227-243. Hillsdale, N.J.: Lawrence Erlbaum.
OGrady, W. (1997). Syntactic Development. Chicago: The University of Chicago Press.
Orfield, G. (1986). Hispanic education: challenges, research, and policies. American Journal of Education, 95, 125.
Pavesi, M. (1986). Markedness, discoursal modes, and relative clause formation in a formal and an informal
context. Studies in Second Language Acquisition, 8 (1), 38-55.
Prez-Leroux, A. (1995). Resumptives in the acquisition of relative clauses. Language Acquisition, 4, 105-138.
Prideaux, G. D. (1980). The role of perceptual strategies in the processing of English relative clause structures. In
Proceedings of the eighth international conference on computational linguistics, pp. 60-66. Tokyo.
Prideaux, G. D. (1982). The processing of Japanese relative clauses. The Canadian Journal of Linguistics, 27 (1),
23-30.
Scarborough, H. S. (1991). Early syntactic development of dyslexic children. Annals of Dyslexia, 41, 207-220.
Scarborough, H. S. (1998). Early identification of children at risk for reading disabilities: phonological awareness
and some other promising predictors. In B. K. Shapiro, P. J. Accardo & A. J. Capute (eds.), Specific Reading
Disability: A View of the Spectrum. pp. 77-121. Timonium, MD: York press.
Schumann, J. H. (1980). The acquisition of English relative clauses by second language learners. In R. C.
Scarcella & S. D. Krashen (eds.), Research in second language acquisition. New York: Newbury House.
Shankweiler, D. & Crain, S. (1986). Language mechanisms and reading disorder: a modular approach. Cognition,
24 (1-2), 139-168.
Shankweiler, D., Crain, S., Katz, L., Fowler, A. E., Liberman, A. M., Brady, S. A., Thornton, R., Lundquist, E.,
Dreyer, L., Fletcher, J. M., Stuebing, K. K., Shaywitz, S. E. & Shaywitz, B. A. (1995). Cognitive profiles of
reading-disabled children: comparison of language skills in phonology, morphology, and syntax.
Psychological Science, 6 (3), 149-156.
Sheldon, A. (1974). The role of parallel function in the acquisition of relative clauses in English. Journal of
Verbal Learning and Verbal Behavior, 13, 272-281.
Sheldon, A. (1977). On strategies for processing relative clauses: a comparison of children and adults. Journal of
Psycholinguistic Research, 6 (4), 305-318.
Skutnabb-Kangas, T. & Toukomaa, P. (1976). Teaching migrant children mother tongue and learning the
language of the host country in the context of the socio-cultural situation of the migrants family. Tampere,
Finland: Tukimuksia Research Reports.
Slobin, D. I. (1971). Psycholinguistics. Glenview, IL: Scott, Foresman.
Slobin, D. I. & Welsh, C. A. (1971). Elicited imitation as a research tool in developmental psycholinguistics. In C.
B. Lavatelli (ed.), Language training in early childhood education, Urbana, IL: University of Illinois Press.
Smith, S., Macaruso, P., Shankweiler, D. & Crain, S. (1989). Syntactic comprehension in young poor readers.
Applied Psycholinguistics, 10 (4), 429-454.
Snow, C. (2002). Reading for understanding: toward an R&D program in reading comprehension. Santa Monica,
CA: Science and Technology Policy Institute, RAND Education.
Stein, C. L., Cairns, H. S. & Zurif, E. B. (1984). Sentence comprehension limitations related to syntactic deficits
in reading-disabled children. Applied Psycholinguistics, 5 (4), 305-322.
Tarallo, F. & Myhill, J. (1983). Interference and natural language processing in second language acquisition.
Language Learning, 33, 55-76.
Tavakolian, S. (1977). Structural principles in the acquisition of complex sentences. PhD dissertation, University
of Massachusetts, Amherst, Massachusetts.
1543
Tavakolian, S. (1978). The conjoined-clause analysis of relative clauses and other structures. In H. Goodluck & L.
Solan (eds.), Papers in the structure and development of child language, pp. 37-83, Amherst: University of
Massachusetts, Linguistics Department, GLSA Publications.
Toukomaa, P. & Skutnabb-Kangas, T. (1977). The intensive teaching of the mother tongue to migrant children at
pre-school age. Research report #26, Department of Sociology and Social Psychology, University of
Tampere.
Tunmer, W. E. & Bowey, J. A. (1984). Metalinguistic awareness and reading acquisition. In W. E. Tunmer, C.
Pratt & M. L. Herriman (eds.), Metalinguistic awareness in children: theory, research, and implications, pp.
144-168. Berlin: Springer-Verlag.
Tunmer, W. E. & Hoover, W. (1993). Phonological recoding skill and beginning reading. Reading and Writing, 5
(2), 161-179.
Tunmer, W. E., Nesdale, A. R. & Wright, A. D. (1987). Syntactic awareness and reading acquisition. The British
Journal of Developmental Psychology, 5 (1), 25-34.
Vogel, S. A. (1975). Syntactic abilities in normal and dyslexic children. Baltimore, MD: University Park Press.
Waltzman, D. E. & Cairns, H. S. (2000). Grammatical knowledge of third grade good and poor readers. Applied
Psycholinguistics, 21 (2), 263-284.
1544