Pamela
Rosenthal Rollins
University
of Texas at Dallas
Callier Center for Communication Disorders
(in press, Journal
of Applied Psycholinguistics)
Correspondence
UTD/Callier Center
1966 Inwood Rd.
Dallas, TX 75235.
Telephone: 214-905-3153
Fax: 214-905-3
Abstract
This prospective longitudinal study examined the relationship between caregiver input to 9-month infants and subsequent language. Mothers-infant dyads were videotaped at 9, 12, and 30 month. Language comprehension (12 and 18 months) was measures by parent report and correlated with an independent language measure. Three maternal style variables were reduced from the 9-month data. Only Caregivers Contingent Comments (CCC) related to infants’ latter language. These findings held after infants’ skill with joint attention (CJA) was accounted for. Total number of words mother's used at 9 months predicted vocabulary, however, the predictive power was encapsulated in words mother used during CCC. Because studies have typically examined maternal input once infants CJA has emerged, this work contributes to current efforts to understand variations in very early language development.
Key words; Mother-Child interaction, Joint Attention, vocabulary comprehension, pragmatics style.
Introduction
There is mounting evidence
indicating strong links between the social environment during the second year
of life and later language development.
Studies have found contributions from the nature of the language
interactions in which the child engages: the quantity of maternal language
heard (Huttenlocher, Haight, Bryk, Seltzer & Lyons, 1991; Hart &
Risley, 1999), and/or the usefulness of that language as a source of
information (as mediated by parental maintenance of semantic contingency and
use of a conversation style, see Gallaway & Richards, 1994; Snow, 1989 for
reviews). Overwhelmingly, these studies have focused on language interaction
with children 12 months and older – the age at which other important
social/cognitive skills are emerging. In
particular, by 12 months typical infants routinely engage in coordinated joint
attention with their caregivers (Bakeman & Adamson, 1984; Carpenter,
Nagell, & Tomasello, 1998) a skill that is important to social-pragmatic
theories of language (Tomasello, 1999).
Unfortunately, there are very few studies devoted to understanding the communicative environments of infants younger than 12 months.1 The study by Carpenter et al. is a notable exception, however, these authors did not find an effect of maternal language in children as young as 9 months. Therefore, the relationship between the social/communicative environment and subsequent language accomplishments remains unclear in young infants whose skill at coordinated joint attention with caregivers is still fragile. The purpose of this prospective longitudinal study was to (a) understand the individual variation in caregiver input while interacting with their 9-month-old infants; (b) examine how caregiver input in these different communicative contexts may relate to subsequent language learning and (c) to explore whether maternal input to 9-month-old infants continues to be associated with linguistic skill once joint attention emerges. We hypothesize that linguistic processing strategies change as the infant ages. Specifically, language learning for 9-month-old infants is facilitated by caregivers' talk about the infant's focus of attention and only later (12 months) by a more specific placement of talk during periods of coordinated joint attention. Nine-month-old infants have been described as being more passive in their processing of linguistic information (Baldwin & Moses, 1996). Information acquisition, at 9 months of age is more likely to occur when there is an exact match between caregivers’ input, and the infants’ focus of attention to a reference in the immediate environment (Moore, 1998). Accordingly, linguistic input to 9-month-olds should facilitate language acquisition when embedded in adult-child interactions while the infant is attending to objects. On the other hand, 12-month-old infants routinely incorporate a third element into dyadic interactions, marking the emergence of the understanding that others have intentions and attentions which may be actively shared, followed or directed (Tomasello, 1999). This new achievement is thought to mark the beginning of intentionality (Baldwin & Moses, 1996; Tomasello, 1999) and has been found to facilitate the acquisition of language by creating a shared referential framework in which the child may ground the caregivers’ language input (Bruner, 1983; Nelson, 1973; Snow, Perlman, & Nathan, 1987; Tomasello, 1999).
Typical children rapidly move to understanding words in bouts of coordinated joint attention in a matter of a few months. The more swiftly the child progresses, the more obscure the lines between these two linguistic-processing strategies may seem. Nonetheless, understanding the contributions of more primitive socially constructed word learning is important when we consider individual variation in vocabulary for both typical children and those with language disorders. Children with autism provide a strong case for understanding the potential effects of early caregiver input. These children have protracted periods before coordinated joint attention emerges, and it is an even longer time before coordinated joint attention is part of the routine adult-child interaction (Baron-Cohen, 1989; Rollins, 1999). However, contrary to what social pragmatic theories of language would suggest, children with autism are capable of acquiring several words in the absence of joint attention (Rollins, 1999). If language is first rooted in a social system, then it is important to extend our knowledge of social correlates of language to infants who developmentally are not yet engaging in robust amount of coordinated joint attention.
To this aim, we have conducted a prospective longitudinal
study to explore the relationship between maternal language in different communicative
interactions and subsequent language developments. Specifically, we evaluated
maternal communicative input to 9-month-old infants, as well as child
coordinated joint attention at 9 months (when still fragile) and at 12 months
(when more robust) as predictors of language comprehension at 12 and 18 months
of age. In keeping with our objective,
we were particularly interested in measuring vocabulary comprehension (as
opposed to vocabulary production) because production is much more variable than
comprehension in the 12-18 month age range.
Some children produce more than 500 words and put together 2- to 3- word
sentences by their 18-month birthday while others use only a few words (e.g.,
"uh-oh!") at this age (see the standard errors in Carpenter et al.;
Fenson, Dale, Reznick, Thal, Bates, Hartung, Pethick & Reilly, 1993).
Because this type of variability is a characteristic in early vocabulary
production, it is difficult to differentiate those children who will ultimately
catch-up by 30 months of age from those children who will end up with language
disorders. On the other hand, vocabulary comprehension has been found to be a
strong predictor of later language (Thal, Tobias & Morrison, 1991; Thal,
2000). Unfortunately language
comprehension is difficult to assess during the second year of life. This has
led many investigators to turn to parent report measures like the MacArthur CDI
(Fenson, et al., 1993). The CDI has been considered a reliable and valid
measure of vocabulary comprehension by some (Bates, Bretherton & Snyder,
1988; McCartheren, Warren & Yoder, 1996; Ring, Erin Dyer. & Larry
Fenson, 2001;Thal, O’Hanlon, Clemmons, & Fralin, 1999) and has been used by
other investigators to address similar questions (Carpenter et al., 1998).
Nonetheless, the construct validity of parent report measures, like the CDI, has
been questioned. It is possible that in
studies such as ours, the relationship between maternal input to their
9-month-old infants and parent reported vocabulary comprehension at 12 and 18
months has more to do with parents and their attitudes toward their children's
language development than it does with specific language learning processes
used by the child. If a comprehensible pattern of relations between the
maternal input variables and later language outcomes are to be found, measures
collected by some method other than parent report must be included. Thus, we
turn to examining the relationship between maternal input variables and an
observational measure of each child’s productive language at 30 months. By providing a language measure independent
of parent report greater confidence may be placed on the validity of the
findings.
Thus, this paper seeks to contribute to current efforts to understand variations in early language development by examining two aspects of object-focused mother-infant communication, maternal language input and infant coordinated joint attention. We hypothesize that maternal input to the child’s focus of attention would be related to the language learning when the infants skill at coordinating joint attention was still fragile. Accordingly, we would expect that on average, differences in maternal input to 9-month-old infants would be associated with individual differences in the infants’ vocabulary at 12 months. Next, we explore whether early differences in maternal input to their 9-month-old infants continue to be associated with subsequent language outcomes once infants engage in robust amounts of coordinated joint attention, which can be actively recruited for language learning. Thus, we examine the multivariate relationship between caregivers’ input at 9 months, and the 12-month-old infants’ use of coordinated joint attention on subsequent language skill. Finally, we examine the relationship between parent report measures of language comprehension, when infants where 12 and 18 months of age, and an observational measure of each child’s productive language at 30 months.
Participants
The study presented here utilized a prospective
longitudinal design. Participants were 11
mother-child dyads recruited by mail as part of the UTD Language and
Communication Database Project. For the current study, only full-term (38-42
weeks gestation), monolingual, normal birth weight children (at least 2500
grams), who did not experience extensive medical complications at birth or
other major illnesses, hospitalizations or developmental disabilities, were
included. The participants (3 girls and 8 boys) were all from well-educated
families. Years of maternal education ranged from 12-18 with a mean of 16.45
years and a standard deviation 1.75 years.2 All children were learning
English as their first language, with none of the parents reporting that their
child experienced substantive exposure to another language (i.e., more than 7
hours per week). All participants were brought into the lab within one week of
their 9 month, 12 month, and 30
month birthdays and engaged in a parent-child interaction. In addition, parents
completed a commercially available parental report instrument, the MacArthur
Communicative Development Inventory: Words & Gestures (CDI: W&G)
(Fenson, et al., 1993). Each CDI: W&G was either mailed directly to the
parents along with a cover letter or was given to the parent during their visit
to the laboratory. Parents completed the CDI: W&G when the infants were
between 11-14 months (M = 12.27, SD = .79) and again when they
where between 17-18 months (M = 17.8, SD = .42) and returned them
to our facilities in a postage-paid envelope. Parents were told that they could
contact the project if they had any questions at any point.
Measures
Pragmatic Input. The Inventory of Communicative Acts-Abridged
(INCA-A) (Ninio, Snow, Pan & Rollins, 1994) was used to code mother's
communicative intents. This system is a shortened and modified version of the
system developed by Ninio and Wheeler (1984; see Ninio & Snow, 1996). The
system is based both on Speech Act Theory (Austin, 1962; Searle, 1976) and on
studies of face-to-face interaction (Goffman 1974; Streeck, 1980) that
emphasize the importance of socially constructed communicative interchanges.
Thus, the system identifies and codes communicative intent at two different
levels -- the level of the social interchange and the level of the utterance,
acknowledging the existence of an organization of talk at a level higher than
the single utterance (c.f. Streeck, 1980; Dore & McDermott, 1982). An interchange is defined as one or more
rounds of talk, all of which serve a unitary interactive function implicitly
agreed upon by the interlocutors. Within
this social-interchange, speakers express specific intents at the utterance
level. The INCA-A, then, actually
consists of two subsystems, each of which codes for a different component of
communicative intent. Since the system
was designed to provide exhaustive coding of the communicative attempts
expressed by children of varying ages (as well as their mothers), it can
reflect development and continuity across a wide age range.
Inspection of the interchange speech act combinations used by mothers to their
9-month-old infants revealed that they could be characterized by three
theoretically motivated categories: (a) child-centered acts that were sensitive
to the infants’ focus of attention, (b) other child-centered acts, and (c)
directive acts. We called the first
category Caregivers Contingent Comments (CCC).
Caregiver Contingent Comments were child-centered acts where the mother
discussed an object of joint focus of attention or narrated an ongoing
activity. These included utterances
coded in INCA-A as Discuss Joint Focus of Attention (DJF) in conjunction with
any speech act and Negotiate Immediate Activity (NIA) with a statement
(ST). The second category, other
child-centered acts, included routines in which the mothers engaged the child
in social exchanges (e.g., wave bye-bye, peek-a-boo) and feeling state
interchanges in which the mother talked about her own feelings or the child’s
feelings (“you like that?”). These included utterances coded in INCA-A as
Perform a Verbal Move in a Game (PRO on the interchange level or PR on the
speech act level), Mark an Event (MRK on the interchange level and MK on the
speech act level) and responses to questions in practiced formats (coded on the
speech act level); and Discuss Hearers, or Speakers’ Feelings (DHS, DSS
interchanges). Our last category was
child-directive acts which included directive interchanges in which the mother
communicated instrumentally to negotiate which actions would be carried out and
by whom (e.g., “Can you take the next one off?” “Put the blue one on.”), direct
attention in which the mother directed the infants attention to an object or
event (e.g., pointed to an object or event, “look at this”) and recent event
interchanges where the mother discussed an immediate past event (e.g., “it rolled
away”). Directives included utterances coded in INCA-A as Negotiate Immediate
Activity (NIA) with all speech act categories except for statements (ST, SI ),
proformatives (PR) and markings (MK); and Direct Hearers Attention (DHA
interchange) and Discuss immediate Recent Event (DRE interchange). Other
communicative acts prescribed by the Ninio et al., (1994) system were seldom
used by the mothers in the current study (e.g., all other interchanges with the
exception of Discuss Clarification [DCC interchange] accounted for less than 1%
of the 2338 maternal communicative
acts. DCC accounted 4% of the data).
Finally, because there was several interchange - speech act combinations
within each category of the three categories described above, we chose to numerically
reduce the information within each of the three categories using Principle
Component Analysis (PCA). PCA solves
many of the problems associated with compositing variables by other techniques
such as simple frequency counts. Specifically PCA (a) standardizes the
variables so that the variable with the largest variance does not dominate the
composite, (b) weights the standardized variables by the correlation between it
and the standardized sum so that all of the variables point in the same direction,
and (c) iterates the process for maximal internal consistency and reliability
(see Afifi & Clark, 1984). Thus, we were left with three pragmatic input
composites (a) caregiver contingent comments; (b) other child centered acts;
and (c) directives.
Lexical input at 9 months: Four measures of lexical input were calculated for each mother: the total number of words (TNW) the mother produced in 10 minutes (Klee, 1992; Miller, 1991) as well as the total number of words each mother used while engaging in each of the interchange speech act combinations associated with the three pragmatic input categories. Specifically, we calculated for each mother the total number of words used while engaging in contingent comments, other child-centered acts and child-directive acts.
Coordinated Joint Attention (CJA) at 9 and 12 months . CJA was coded by reviewing the original videotapes. The coding procedure strictly adhered to the rules formulated by Bakeman and Adamson (1984) and Carpenter, et al., (1998). Specifically, Coordinated Joint Attention episodes in which the mother and infant share attention to an object of mutual interest for at least 3 seconds, were coded.
CJA was coded when the infant looked from an object to the mother’s face and back to the same object. Episodes in which the infants clearly looked at mother’s face as a response to an adult action, vocalization or language were not coded as this would be more accurately described as an alternation of attention and not coordination of attention. The total duration of CJA in 10 minutes at 9 months and at 12 months were used as measures of CJA (see Carpenter, et al., 1998).
Child Vocabulary Comprehension at 12 and 18 months. The MacArthur Communicative Development Inventory, words and gestures (CDI: W&G) (Fenson, et al., 1993) was used to assess vocabulary comprehension at both ages. The CDI comprehension measure has been found to be a valid and reliable measure of vocabulary comprehension for children in this age range, and who have mothers of similar educational background (Bates, Bretherton & Snyder,1988; Ring, Erin Dyer. & Fenson, 2001; Thal, D. J., O’Hanlon, L., Clemmons, M. & Fralin, L. 1999). The CDI: W&G contains a vocabulary checklist of 396 words organized into categories such as animals, food and drink, body parts, people and action words. At both ages we computed a measure of vocabulary comprehension based on the sum of the items that parents marked on the checklist that their child "understands.” All 11 parents completed the CDI forms when the infants were between 11-14 months (M = 12.27, SD = .79) and but only 9 CDI’s where returned when infnats where between 17-18 months (M = 17.8, SD= .42). Because the “12 month” comprehension data spanned three months (11-14 months), we statistically controlled for age in the analyses using this data.
Child productive language skill at 30 months. Ten of the original 11 children were observed in the laboratory at 30 months. Each of these children’s productive language skill was assessed using the Index of Productive Syntax (IPSyn) (Scarborough, 1990). The IPSyn score assess emerging morphosyntactic skills and is a composite score, based on 100 utterances, which reflects the child's use of 56 syntactic and morphological forms, including elaborations within noun and verb phrases, questions, negations, and various sentence structures. Our rationale for using IPSyn was twofold: First, lexicalist approaches to grammar suggest vocabulary and grammar are continuous (see Bates & Goodman, 1997), and second, at 30 months of age vocabulary has typically leveled off and combining words and morphemes mark advances in language.
Reliability
A second rater independently coded 781/2338 or 33% of the mother’s communicative intentions. Cohen's Kappa statistic, which takes account of chance agreement, was calculated. The values for Kappa ranged from .80 to .73 (M= .77) which is considered “substantial” to “almost perfect” agreement according to guidelines established in Landis and Koch (1977). A second rater also coded 20% of the tapes for both episodes and duration of child-coordinated joint attention at 9 and 12 months. Point by points reliability ranged from 94% to 100% (M=96.7%) agreement at 9 months and 94% to 97% (M=96%) agreement at 12 months. Finally, a second rater independently re-scored all of the 30-month videotapes for IPSyn. A point-by-point reliability on the item level codes was 93.6%.
Descriptive statistics for the original pragmatic input measures, total number of words (TNW) in 10 minutes, as well as the total number of words associated with each of the maternal input composites, are found in Table 1. Mothers of 9-month-old infants varied considerably in their use of communicative contexts, their total verbosity (total number of words in 10 minutes) as well as their verbosity while engaging in each of the communicative contexts.
Consistent with previous studies, infants’ use of coordinated joint attention at 9 months was just beginning to emerge. All of our 9-month-old infants engaged in at least one episode of CJA while interacting with their mothers. While the mean total duration spent in CJA at 9 months was 21.9 seconds, there was enough variability in this measure (ranged of 5.93 seconds to 78.02 seconds; SD = 21.6) to suggest a null finding with CJA at 9 months could not be attributed to a floor effect. As we expected, infants engaged in CJA for considerably longer durations at 12 months. On average, during the 10 minutes of toy mediated dyadic play interactions, infants engaged in 191.9 seconds of coordinated joint attention at 12 months (range of 51.4 – 215.3 seconds; SD = 112.3).
All child language measures were also variable. It is noteworthy that the mean percentile ranking for the 12-month-old children was lower than expected from the general population (40.4 percentile vs. 50. percentile). Nonetheless, the children represented a full range of abilities from the 0 percentile to the 86th percentile (M=40.4, SD=26.2). At 12 months, vocabulary comprehension ranged from 9 to 154 words (M= 69.4, SD = 43.9). By 18 months, vocabulary comprehension ranged from a minimum of 78 words to a maximum of 337 (M = 191.9, SD = 98.3 words).
Next, we used correlation and regression analyses to understand the relationship between each of the maternal input variables when infants were 9 months, infants’ coordinated joint attention at 9 and 12 months, and subsequent language outcomes.4 Results revealed, controlling for age, only caregivers’ contingent comments in conversation with 9-month-old infants was predictive of infants’ language comprehension at 12 months (r = .89, p= .003). Specifically, on average, mothers who spent more time talking about objects in the child’s focus of attention at 9 months had children with better vocabulary comprehension at 12 months.
Further, we found, on average, the total number of words mothers used in 10 minutes predicted children's vocabulary comprehension at 12 months, controlling for age (r=.87, p=.004). The pattern of results, however, suggested that mothers’ total number of words while engaging in caregiver contingent comments was driving the effect. Specifically, vocabulary comprehension at 12 months of age was associated with mothers’ total words when engaging in contingent comments at 9 months (r=.89, p=.004), but not mothers’ total number of words within other child-centered acts and child-directive acts.
Thus far we have found that maternal input to the child’s focus of
attention is associated with initial language learning when the infant’s skill
at coordinated joint attention is still fragile at best. Next, we explored whether early differences
in maternal input to 9-months-old infants continued to relate to subsequent
language outcomes, once infants have more developed social/cognitive skills.
Estimated correlations were consistent with many previous studies: there was a positive concurrent relationship between CJA and words understood at 12 months (r=.69, p= .03 controlling for age); and a predictive relationships between CJA at 12 and words understood at 18 (r= .67, p = .05). In keeping with Spearman’s Law of reliability that states a measure cannot correlate with another measure higher than it does itself, words understood at 12 months had the highest predicative relationship with words understood at 18. Noteworthy, for the hypothesis under investigation is the strength of the relationship between CCC at 9 months and vocabulary comprehension at 18 months is strikingly similar the strength of the relationship between CJA at 12 months and vocabulary comprehension at 18 months (r = .70, p = .03, r = .67, p= .05 respectively).
Furthermore, multiple regression revealed that 64% of the variance in words understood at 18 months was attributable to the combined effects of caregivers’ contingent comments at 9 months and CJA at 12 months F(2, 6) = 5.32, p=.05). This suggests that the way mothers talk to their prelinguistic infants continues to be associated with language comprehension even after coordinated joint attention is used with some frequency.
In order to assuage the
alternative hypothesis that early maternal input is related to mother’s
judgment about the infant’s comprehension, (as opposed to the infants true
language abilities) an independent measure of child language was introduced.
More specifically, we investigated the relationship between infants’
comprehension (as measured by parent report) at 12 and 18 months, with child’s
IPSyn scores obtained during mother-child conversation at 30 months of age. The
Children’s IPSyn scores at 30 months ranged from 28 to 72 (M=54.9, SD=
13.6). We found that there was a positive relationship between parent report of
infants’ comprehension at 12 and at 18 months and children’s’ IPSyn scores at
30 months (r= .78, p= .02; and p=.87, p= .002 respectively). Furthermore, we
found that on average, mothers who spent
more time using contingent comments when the infant was 9-months, had children
with higher IPSyn scores at 30 months (r=.74, p=.01). Coordinated joint
attention at 12 months was not related to IPSyn at 30 (r = .53, p= .12) but the
null effect was due to one child, without her there was an estimated
correlation r= .92, p = .001).
Discussion
This study highlights the relationship between caregiver input to 9-month-old infants and subsequent language developments. Although the positive effects of child-centered utterances have been demonstrated repeatedly for infants in the second year of life (Carpenter, et al., 1998; Rollins & Snow, 1998; Tomasello & Todd, 1983 to name a few) this association has not been found for infants as young as 9 months. As expected, we did not find infants’ use of coordinated joint attention at 9 months predictive of early vocabulary comprehension. Thus we found that even before infants routinely engage in coordinated joint attention, the social/linguistic environment was associated with later language development. At 9 months, infants engage in very few instances of coordinated joint attention, and do not appear to be actively recruiting their newly learned social skill to promote language learning. Thus, these findings are in accord with theories that suggest that 9-month-olds use a more passive linguistic information processing strategy (Baldwin & Moses, 1996).
Moreover, we did not find caregiver contingent comments to 9-month-old infants to be a fleeting influence. Caregivers’ Contingent Comments related to language outcomes at 12, 18 and 30 months even when joint attention was accounted for. It is noteworthy that our child language measures at 12 and 18 months where obtained through parent report, and as such may have been biased by the mothers’ attitudes towards infants’ language abilities. However, our parent report measures were highly correlated with an independent observational language measure obtained when the children were 30 months. We interpret this finding as evidence of the validity of our parent report measures of language. Stated differently, these findings do not simply indicate that mothers who think more highly of their children’s language abilities speak more supportively. Rather, they suggest that a relationship between caregiver input to 9-month-old infants and later language development. Specifically at 9 months of age the caregivers’ temporally contingent comments about the infants’ focus of attention is associated with infants’ ability to perceive the relationship between words and their referents. This finding was particularly striking when you consider we studied a group of educated mothers, and the pattern of results remained unchanged when maternal education was partialed out (refer to footnote 4). That is, we would expect maternal input to be more variable if we had sampled families with more diverse social-economic backgrounds (Hart & Risley, 1999).
The type of learning described here does not imply that infants are developing true language at 9 months. As Tomasello (1999) points out “Infants this young may on occasion, of course, learn to associate one of these noises with a perceptual event in much the same way a household pet may understand that the sound dinner heralds the arrival of food. But this is not language. Sounds become language for young children when and only when they understand that the adult is making that sound with the intention that they attend to something” (p100-101). Thus the issue is not whether infants are learning true language at 9 months but whether the social environment is important for later language. Nine-month-old infants are probably only capable of comprehending a few words as a result of immediate contingent reinforcement. However, the current study suggests that the early social/linguistic environment, prior to the influences of infants’ ability to share reference is related to children’s later language skills. We offer two explanations for the association. First, the more vocabulary comprehension infants have by the time they are able to share reference with their caregivers, the more distal from the immediate here and now of the communicative episode the shared object or reference can become. This is important as the infant becomes increasingly mobile. Thus mothers, and infants with their increasing skill at coordinated joint attention, can begin to share reference from across the room. What we are suggesting here is a dynamic assembly of early learning supported by environmental influences with infants developing understanding of others as intentional beings.
An alternative explanation is mothers who are initially adept at using contingent comments also use more facilitative features (e.g., child centered conversation style, semantically contingent utterances) as the child gets older. Carpenter et al., (1998) found only moderate correlations between maternal input to 9 month-old infants and maternal input to the same infants at 12 and 15 months (.52 and .56 respectively) thus accounting for about 25% of the variability. Nonetheless, it would be important to address this question empirically.
Another area of further research suggested by the current study is the relationship between maternal talkativeness and vocabulary in slightly older infants. Our study suggests the relationship between overall caregiver talkativeness and subsequent language development may be unduly influenced by caregivers’ verbosity within specific communicative contexts. This finding has important ramification for evidenced based treatment approaches in early intervention. As such, understanding whether or not communicative context has biased the relationship between overall maternal talkativeness and language outcomes in slightly older children is important.
A weakness of this study was the
sample size is small. However, because
the study is both longitudinal and micro-analytic in nature, the results add to
our understanding of a socially constructed communication system and lend
further support to a social pragmatic theory of language development. Of course,
these correlations do not prove causality.
The prospective nature of these cross lagged correlations however,
suggests for normally developing children at least, caregivers’ contingent
comments to 9-month-old infants contribute to the development of early language
skill.
A more powerful test of our hypothesis would have been to study rate of vocabulary development. Unfortunately we were unable to take advantage of growth modeling techniques because vocabulary scores were collected at only two time points. It is noteworthy, however, that past research has shown that vocabulary growth is linear from 12 to 18 months, thus rate of change would not yield sufficient additional information in a study of this size.
In summary, this paper highlights the vital role of caregivers’ input to their 9-month-old infants and subsequent language skill. That is, even before an infant can actively attend to the referent that the caregiver is talking about, caregivers’ comments on the infants’ focus of attention are associated with subsequent language. Furthermore, the facilitative effects of caregivers’ contingent comments on the infants’ focus of attention did not disappear once a child used robust amount of coordinated joint attention. Thus, this study contributes to the current efforts to understand variations in very early language development.
Appendix
Toy list
1) happy bird toy (large wobble toy
that jingles when tipped)
2) big bird stuffed animal
3) ring stacker
4) book on colors
5) plastic book of faces with wholes for eyes
6) hat
7) hand-shaped chew toy
8) foot-shaped chew toy
9) plastic pig, cow and sheep figurines
10) plastic toy
car
11)
candy-striped rattle with blue balls in it
12) doll
Twelve month toys
Included all toys listed for nine months plus
10) telephone
13) ball
14) toy (like carousel) that goes around when top is pushed.
References
Afifi, A.A., & Clark, V., (1984). Computer-aided multivariate analysis. NY, New York: Van Nostrand Rienhold Company.
Austin, J.L. (1962). How to Do Things with Words. Cambridge, MA: Harvard University Press.
Bakeman, R., & Adamson, L., (1984). Coordinating attention to people and objects in mother-infant and peer-infant interaction. Child Development, 55, 1278-1289.
Bates, E., Bretherton, I., & Snyder, L. (1998). From first words to grammar: Individual differences and dissociable mechanisms. New York: Cambridge University Press.
Bates, E. & Goodman, J. (1997). On the inseparability of grammar and the lexicon: Evidence from acquisition, aphasia, and real-time processing. Language and Cognitive Processes, 12 507-584.
Baldwin, D., & Moses, L. (1996). The ontogeny of social information gathering. Child Development, 67, 1915-39.
Baron-Cohen, S., (1989). The autistic child's theory of mind: A case of specific developmental delay. Journal of Child Psychology and Psychiatry 30, 285-298.
Baron-Cohen, S., Baldwin, D., & Crowson, (1997). Do children with autism use the speaker’s direction of gaze strategy to crack the code of language? Child Development, 68, 48-57.
Bruner,
J. (1983). Child's Talk: Learning to use language. New York: Norton.
Carpenter, M., Nagell, K., & Tomasello, M. (1998). Social cognition, joint attention, and communicative competence from 9-15 months. Monographs of the Society for Research in Child Development, 63.
Dore, J., & McDermott, R.P. (1982). Linguistic indeterminacy and social context in utterance interpretation. Language, 58, 374‑398.
Fenson,
L., Dale, P., Reznick, J.S., Thal, D., Bates, E., Hartung, J., Pethick, S.,
& Reilly, J. (1993). The MacArthur Communicative Development
Inventories: Users guide and technical manual. San Diego, CA: Singular
Publishing Group.
Gallaway, C. & Richards, B.J. (1994) Input and interaction in language acquisition. Cambridge University Press.
Goffman,
E. (1974). Frame analysis: an essay on the organization of experience.
Harmondsworth: Penguin.
Goldfield,
B. (1990). Pointing, naming, and talk about objects: Referential behaviour in
children and mothers. First Language, 10, 231-242.
Hart, B., & Risley, T. (1999). The social world of children: Learning to talk. Baltimore, MD. Paul Brookes Publishing Co.
Huttenlocher, J., Haight, W., Bryk, A., Seltzer, M., Lyons, T. (1991). Early vocabulary growth: Relation to language input and gender. Developmental Psychology 27, 236-248.
Klee, T. (1992). Development and diagnostic characteristics of quantitative measures of children's language production. Topics in Language Disorders, 12, 18-41.
Landis, J. & Koch, G. (1977). The measurement of observer agreement for categorical data. Biometrics, 33, 159-174.
MacWhinney, B. (1991). The CHILDES Project: Computational Tools for Analyzing Talk. Hillsdale, N.J.: Lawrence Erlbaum.
McCathre, R., Warren, S.F., & Yoder, P.J. (1996). Prelinguistic predictors of later language development. In K.N. Cole, P.S. Dale & D. J. Thal (Eds.), Assessment of Communication and Language. (Vol. 6) Communication and Language Intervention Series. Baltimore, MD. Paul H. Brooks Publishing Co.
Miller, J. F. (1991). Quantifying productive language disorders. In J.F. Miller (Ed.), Research on Child Language Disorders, (pp. 211-220). Austin, TX. Pro-ed.
Moore (1998). Commentary, Social
cognition in infancy. Monographs of
the Society for Research in Child Development, 63, 167-174.
Nelson, K. (1973). Structure and strategy in learning to talk. Monographs of the Society for Research in Child Development, 38, (serial No. 149)
Ninio, A. & Snow, C.E. (1996). Pragmatic Development. Boulder: Westview.
Ninio, A., Snow, C.E., Pan, B.A., & Rollins, P.R. (1994). Classifying communicative acts in children's interactions. Journal of Communications Disorders, 27, 157-188.
Ninio, A., & Wheeler, P. (1984). A manual for classifying verbal communicative acts in mother-infant interaction. Working Papers in Developmental Psychology, No. 1. Jerusalem: The Martin and Vivian Levin Center, Hebrew University. Reprinted as Transcript Analysis, 1986, 3, 1-82.
Ring,E.D.,& Fenson, L.(2000).The correspondence between parent report and child performance for receptive and expressive vocabulary beyond infancy. First Language, 20, 141-159.
Rollins,
P.R., (1999). Early Pragmatic Accomplishments and Vocabulary Development in
Preschool Children with Autism. American Journal of Speech Language
Pathology, 8, 85-94.
Rollins, P. R., & Snow, C.E. (1998). Shared attention and grammatical skills in typical children and children with autism. Journal of Child Language, 25, 653-674.
Scarborough, H. (1990). Index of productive syntax. Applied
Pycholinguistics, 11, 1‑22.
Searle, J. (1976). A classification of illocutionary acts. Language in Society 5, 1-23.
Snow, C.E., (1989). Understanding social interactions and language acquisition: Sentences are not enough. In M. Bornstein & J. Bruner (Eds.), Interactions in human development. (83-103). Hilldale, NJ: Erlbaum.
Snow, C.E., (1989). Understanding social interactions and language acquisition: Sentences are not enough. In M. Bornstein & J. Bruner (Eds.), Interactions in human development. (83-103). Hilldale, NJ: Lawrence Erlbaum Associates
Snow, C.E., Perlmann, R., & Nathan, D., (1987). Why routines are different: Toward a multiple-factors model of the relation between input and language acquisition. In K. Nelson & A. van Kleeck (Eds.), Children’s Language (Vol. 6). Hillsdale, NJ: Lawrence Erlbaum Associates
Streeck, J. (1980). Speech acts in interaction: a critique of Searle. Discourse Processes, 3, 133‑154.
Tomasello, M., (1999). The cultural origins of human cognition, Cambridge, MA: Harvard University Press.
Tomasello, M., & Kruger, A. C. (1993). Joint attention on actions: Acquiring verbs in ostensive and non-ostensive contexts. Journal of Child Language, 19, 311-333.
Tomasello,
M., Strosberg, R., & Akhtar, N., (1996). 18 month old children learn words
in non-ostensive contexts. Journal of Child Language, 23, 157-176.
Tomasello, M. & Todd, J.
(1983) Joint attention and lexical acquisition style. First Language 4, 197-212
Thal, D. (2000). Late-talking Toddlers: Are they at risk? San Diego, CA: San Siego State University Press.
Thal, D. J., O’Hanlon, L., Clemmons, M. & Fralin, L. (1999). Validity of a parent report measure of vocabulary and syntax for preschool children with language impairment. Journal of Speech Language & Hearing Research, 42(2), 482-496.
Thal, D.J., Tobias, S., & Morrison, D. (1991). Language and gesture in late talkers: A 1-year follow-up. Journal of Speech and Hearing Research, 34, 604-612.
.
.
This research was supported by grants from
the Timberlawn Foundation and UTD School of Human Development faculty
initiative. I would like to thank the children and families who participated in
the study, Sandi Schniepp and Diane Moulder for their assistance with
transcription, coding and reliability. To Robert Stillman for insightful
discussions and comments on early drafts of the paper. Portions of the analyses
discussed here have been presented at the 2001 Society of Research on Child
Language Disorders and are included in the proceedings from the 2001 Early Lexicon Acquisition Conference held in Lyon France.
Endnotes
1 It is noteworthy that
some researchers have begun to link other aspects of the infants’ social
environment at 9 months, such as mother’s responsiveness to the child’s
affective signals, with subsequent language outcomes (Bornstein &
Tamis-LeMonda, 1997).
2 Maternal education was used as an indicator of SES because it can be easily and reliably obtained, and it has been found to be strongly correlated with other measures of SES (Hart & Reisley1999).
4 It is noteworthy that
all of the assumptions of correlation and regression analyses were met in our
analyses. Furthermore, except for the one instance where it is noted, there
where no outliers driving the results. In addition, the pattern of results
remained unchanged when analyses where conducted controlling for maternal
education.
Table 1: Univariate statistics for original maternal pragmatic input variables and lexical input variables (n=11)
|
|
Variable |
Mean |
Standard deviation |
Minimum |
Maximum |
Skewness |
|
CCC |
Joint focus |
42.9 |
20.0 |
13 |
73 |
0.08 |
|
|
Narrate ongoing activity |
18.0 |
8.94 |
5 |
32 |
0.17 |
|
Other Child-Centered |
State feelings |
11.8 |
10.4 |
0 |
32 |
0.84 |
|
|
Routine |
48.8 |
15.3 |
21 |
74 |
-0.04 |
|
Directives |
Behavioral Directives |
29.7 |
16.4 |
7 |
58 |
0.23 |
|
|
Attention Directives |
22.2 |
11.18 |
6 |
41 |
0.30 |
|
|
Recent Event |
7.45 |
5.98 |
0 |
16 |
0.16 |
|
|
Total words in CCC |
224.7 |
96.5 |
80 |
376 |
0.09 |
|
|
Total words in Other child centered |
110.3 |
53.6 |
45 |
198 |
0.46 |
|
|
Total words in directives |
264.7 |
114.7 |
131 |
473 |
0.52 |
|
|
Total words in 10 minutes |
756.1 |
328.2 |
374. |
1440.0 |
0.90 |