Social Influences in the First Years of Life

Catherine S. Tamis-LeMonda Tonia N. Cristofaro Eileen T. Rodriguez Marc H. Bornstein

INTRODUCTION

Children's understanding and production of their first words is undeniably one of the most heralded achievements in early development. The onset of language officially marks the transition from "infancy" (which derives from the Latin root infans, meaning "unable to speak") to early childhood, and radically alters the child's social world. Words enable children to share meanings with others and to participate in cultural learning in unprecedented ways. The fact that all normally developing children acquire language suggests that there exist universal properties that support children's language development, including the opportunity to communicate with others and direct and indirect access to analyzable language models (e.g., Meisel, 1995). Nonetheless, children vary enormously in the course of their language growth in terms of how quickly they achieve specific milestones, their receptive and productive vocabulary sizes, and in their eventual levels of competence in syntactic, semantic, and pragmatic aspects of language (Bloom, 1993; Hoff-Ginsberg, 1997; Tamis-LeMonda & Bornstein, 2002). For example, some children speak as early as 8 months while others express their first words several months later (e.g., Acredolo, Goodwyn, Horobin, & Emmons, 1999; Dapretto & Bjork, 2000). Bloom's (1993) research on the language achievements of12 children illustrates the substantial variation in the age of onset of children's first words. Although children expressed their first conventional words (i.e., adult word forms) at the start of the second year on average, some children began to speak as early as 10 months while others did not achieve the first word milestone until 18 months.

Early differences in children's language are not transitory. Language achievements in the first years of life portend children's later abilities in literacy, school readiness, and cognitive development (e.g., Chall, Jacobs, & Baldwin, 1990; Bornstein & Haynes, 1998). In particular, children's oral language skills, which include spoken vocabulary and discourse, have been positively associated with reading and phonemic awareness (e.g., Beals, DeTemple, & Dickinson, 1994; Scarborough, 1989; Whitehurst & Lonigan, 2003), as well as with more developed abilities to construct meanings from conversations, reading, and writing (Beals et al., 1994). Research indicates strong associations between early vocabulary development and later reading comprehension (e.g., Anderson & Freebody, 1983). Moreover, parents' assessments of children's language skills between 3 and 4 years of age relate to teachers' assessments of vocabulary skills in second and fourth grades (e.g., Dickinson & DeTemple, 1998). Children with poor language skills are at a higher risk of literacy failure on school entry than are children with more developed oral language abilities (e.g., Snow, Burns, & Griffin, 1998). These effects are far reaching, such that children who exhibit delays at the onset of schooling are more likely to experience grade retention, special education placement, and failure to complete high school (Adams, 1990; McGill-Franzen, 1987).

The magnitude of associations from early child language to later school performance has incited decades of inquiry into the sources of early individual differences. Research on the social context of language development has underscored the role of more sophisticated language partners, most notably parents, in guiding novice toddlers toward understanding the meanings expressed in shared conversations (Bruner, 1974, 1983; Vygotsky, 1962). A cornerstone of this research has been the careful description of children's language environments in terms of the words and phrases that children hear and the ways that those words are presented to children. This literature has produced irrefutable evidence for the centrality of children's exposure to language for their receptive and productive language growth, literacy, and cognitive development (e.g., Bloom, 1993; Bornstein, 1985; Carpenter, Nagell, & Tomasello, 1998; Hart & Risley, 1995; Huttenlocher, Haight, Bryk, Seltzer, & Lyons, 1991; Landry, Smith, Miller-Loncar, & Swank, 1997; Nelson, 1973, 1988; Snow, 1986; Woodward & Mark-man, 1998).

In addition to studying the contributions that children themselves make to their language progress and learning (e.g., Bornstein et al., 2004), we have investigated the influence of parents' verbal responsiveness on children's early language development. Parents' verbal replies that are prompt, contingent, and appropriate to children's activities (i.e., responsiveness) predict children's language achievements within and across time. Promptness refers to the timing of parents' responses relative to children's behavior. Prompt responses facilitate children's processing of incoming verbal information due to the temporal contiguity between auditory and other perceptual experiences. Contingency refers to the dependence of parents' reactions on child behaviors. Contingent responses are those that evolve out of moments of shared attention and bear direct relevance to what the child is doing (e.g., "Look at the doll!" as the child looks to a doll). Contingency can be contrasted with asynchronous replies (e.g., "Look at the doll!" as a child reaches for a ball), which even if prompt, may actually interfere with the process of mapping words to referents, particularly during early stages of language (Tamis-LeMonda & Bornstein, 2002; Tomasello & Farrar, 1986). Appropriateness refers to maternal replies that are positively connected to the child's behavior. A mother who states, "You're feeding your bear!" as a child directs a spoon to the mouth of a toy bear is providing constructive information, whereas a mother who admonishes "Put that down!" is not.

In this chapter, we describe variation in children's language achievements across the first 3 years, and document findings on the significance of parents' responsiveness for children's early language. The chapter begins with a brief overview of the methods and measures that we have used to study children's language and parents' responsiveness, followed by a presentation of research findings across successive periods of language growth. Research is ordered chronologically, moving from infants' understanding of words (~9-10 months), to production of first words (~13 months), to the "vocabulary spurt" (~18 months), to first "sentences" and decontextualized uses of language (end of the second year). We end by describing our research on the language and cognitive development of 2- and 3-year-old children from low-income families, as children living in poverty are typically found to lag behind their more advantaged peers in their language abilities already in the toddler years. This recent work emphasizes the roles of both fathers and mothers in children's language development, and extends conceptualizations of responsive parenting to include the broader learning environments of young children.

METHODS FOR STUDYING CHILDREN'S LANGUAGE AND PARENTING

Over the past 15 years, we have conducted a series of longitudinal investigations on children's early language and communicative development in the context of parenting in both middle- and low-income populations. Middle-income families were recruited from private pediatric groups in New York City, and families from low-income backgrounds were recruited from community agencies as part of the Early Head Start National Evaluation Study (Love, Kisker, Ross, Schochet, Brooks-Gunn, Paulsell, et al., 2002). Samples ranged from 40 to over 1,000 children across studies. Families were visited in their homes every few months where children were assessed for their language gains and parents were coded on their verbal responsiveness. In the first year, visits occurred when infants were 5-6 months and/or 9 months, depending on the study. In the second year, families were revisited when children were 13-14, 17-18, and 22-24 months of age, and certain cohorts were seen again when children were 3 years of age.

Mother-child and sometimes father-child dyads were observed during free play at each assessment. Children and parents were provided with standard sets of toys and were videotaped for 10 to 20 minutes. Parents were instructed to remain with their children and to interact with them in whatever way was most natural. They were told they could use any or all of the toys provided, but not to introduce other toys into the play session.

Assessments of Children's Language

Measures of children's language and other forms of communication were based on maternal report and/or coding of children's language from verbatim transcriptions of the videotaped free-play sessions. Transcripts were coded for children's word types and tokens, semantic usage, and syntactic complexity (e.g., MLU) using the Systematic Analysis of Language Transcripts (SALT; Miller & Chapman, 1993) as well as special-purpose coding systems.

Maternal reports of children's language were based on the MacArthur Communicative Development Inventories (MCDI; Fenson et al., 1994), which were modified in various studies to access more detailed information about the situations in which children used specific words and phrases. In one longitudinal study of children's language growth from 9 to 21 months, which is later discussed in detail, data on children's language were obtained through biweekly interviews with mothers (Tamis-LeMonda, Bornstein, Kahana-Kalman, Baumwell, & Cyphers, 1998; Tamis-LeMonda, Bornstein, & Baumwell, 2001). To facilitate these interviews, we provided mothers with a packet of language inventories at the end of the 9-month home visit and scheduled weekly times to discuss children's language progress. Packets included versions of the Early Language Inventory (ELI; Bates, Bretherton, & Snyder, 1988) and MacArthur CDI (Fenson et al., 1994) as well as checklists exemplifying various semantic and syntactic uses of words and phrases (Tamis-LeMonda & Bornstein, 1994). At younger ages (i.e., from the start of the study to around 13 months), a subset of items from the MCDI was used (specifically, the ELI on which the MCDI was based), as children at these ages expressed few words and their receptive language was still limited. Early interviews lasted approximately 15 to 20 minutes, whereas interviews at later ages, which utilized the full MCDI, took up to 2 hours.

During telephone interviews, we asked the mother whether her child understood and/or expressed each word/phrase on the list, and in the case of an affirmative response, further probed as to whether the child's understanding and/or production of the word was context "restricted" (e.g., understanding "dog" only as referring to the child's dog; saying "dog" only in imitation or to the child's dog) or "flexible" (e.g., understanding and/or saying "dog" in reference to all dogs).

After probing for specific words/phrases, we asked about various other language milestones, including children's combination of words into simple sentences (e.g., see dog), and whether children used specific grammatical and semantic speech acts. From these interviews, children's receptive and productive vocabulary sizes were computed at each age, as was the timing of various language milestones (e.g., onset of understanding; first words in production; the timing of 50 flexible words in productive language; combinatorial speech).

In certain of these studies, children were also assessed using standardized tests of performance, including the Peabody Picture Vocabulary Test, Third Edition (PPVT-III; Dunn & Dunn, 1997) as well as the Mental Development Index (MDI) of the Bayley Scales of Infant Development—Second Edition (BSID-II; Bayley, 1993). The PPVT, a widely used measure of receptive vocabulary, assesses children's listening comprehension at various ages (for our purposes, focus was on children of 36 months). Children are presented with four pictures per testing plate and asked to point to the one that corresponds to the assessor's spoken word. Items are administered in sets of 12 and testing continues until children reach a ceiling, defined as 8 or more errors within a set of 12. Children in many of our studies were also administered the MDI at 14, 24, and 36 months of age as a measure of their mental status. The MDI assesses abilities in cognitive, language, and personal-social areas of development through items that code for children's vocalizations and language, memory, classification abilities, etc. Both the PPVT and MDI, although standardized tests, capture children's language abilities relative to normative samples.

Coding of Parents' Responsiveness

Parents' responsiveness was coded at either a macro-level (i.e., global ratings) or micro-level (i.e., coded for each parent and child behavioral turn). For macro-level coding systems, coders viewed videotaped free play sessions and coded maternal (or paternal) responsiveness on 5- or 7-point scales ranging from never responsive to always responsive.

For micro-level coding schemes, the play sessions were transcribed verbatim and included notes about target actions of the child and parent. From these transcripts (along with viewing of the videotapes), the frequencies of children's vocalizing, bidding toward mother, looking at objects, and playing were obtained. These child behaviors provided a starting point for coding whether or not the parent responded to each child behavior. Responsiveness to target behaviors was defined as a positive and meaningful change in the parent's behavior that was temporally contiguous (i.e., occurring within 5 seconds), contingent on a change in the child's behavior, and appropriate. We adopted a 5-second window for responsiveness based on research demonstrating the duration for various basic temporal events (e.g., lines of poetry, spoken sentences, breath cycles, and communicative movements) falls between 2 and 7 seconds (see Jaffe, Beebe, Feldstein, Crown, & Jasnow, 2001, for discussion; Rovee-Collier, 1995). As an example, if the child looked at a cup (target act = exploring) and the mother said, "cup," the mother was credited with being responsive; similarly, if the child said, "cup" (target act = vocalizing) and mother responded, "Yes, that's a cup," she would be credited with responding. For each instance of responsiveness, coders noted what the mother was responding to (response target) and the precise nature of her response (response type). The categories of child behaviors, as well as maternal responses to these individual behaviors, were mutually exclusive. From these data, frequency counts were obtained on infant target acts, mothers' responses to each target act, and various types of responses (e.g., descriptions, questions).

CHILDREN'S EARLY LANGUAGE ACHIEVEMENTS

In this section, findings from the various studies that we have conducted are presented in developmental order. Presentation begins with research on children's early understanding of words (receptive language), followed by children's first words in production, vocabulary spurt, and first sentences (combinatorial speech). Each of these subsections opens with a description of the target language achievement, followed by research on the role of parents in supporting the specific achievement.

Understanding Words: The Dawning of Language

The first months of life can be conceptualized as a protracted period of social-communicative "lessons." People have long been captivated by the rhythmic mutuality that characterizes the earliest infant-caregiver interactions (e.g., Bateson, 1975, 1979; Bloom, 1998a; Fogel, Messinger, Dickson, & Hsu, 1999). Caregivers and babies take turns gazing, vocalizing, pausing, smiling, and moving in exquisite synchrony, as though engaged in a communicative "dance" (Lock, 2001; Stern, 1985). Through their participation in these moment-to-moment exchanges, babies are socialized into the communication system (e.g., Hsu & Fogel, 2001). Already by 6 months of age, infants take an active lead in initiating interactions with familiar people, and they exhibit heightened responsiveness to their partners' cues (e.g., Lock, 2001). These early protoconversations (Bateson, 1975; Bloom, 1998a) lay the foundation for later conversations and verbal exchanges (e.g., Ginsburg & Kilbourne, 1988; Locke, 1995), which retain many of the temporal and affective qualities of earlier social communications, yet are marked by the infant's increasing appreciation of the meanings of the words that others use during social interactions.

It is somewhere around 9-10 months that the transition from prelinguistic to linguistic communications occurs, a milestone that is rooted in infants' growing social-cognitive understandings of others (Tamis-LeMonda & Adolph, in press). Through daily participation in reciprocal exchanges, infants acquire an understanding of agency and intentionality (e.g., Lock, 2001; Tomasello, 1995). By the end of the first year, infants are thought to appreciate that people can share attention toward objects and events in the outside world, an understanding that has been referred to as "secondary intersubjectivity" (Stern, 1985; Trevarthen, 1993). This recognition enables children to benefit from social interactions in new ways, and expands their opportunities for learning enormously (Baldwin & Moses, 1996; Tomasello, 1993). Children look where adults look, imitate adults' actions with novel objects, and reference adults in ambiguous situations. At this same time, infants begin to exhibit deictic gestures, such as pointing and showing, suggesting that they are engaged in purposeful communication and wish to convey their intentions or emotions to others (e.g., Bates, Benigni, Bretherton, Camaioni, & Volterra, 1979; Camaioni, Aureli, Bellagamba, & Fogel, 2003). These episodes of joint attention and shared understanding are central to children's social cognitive and language development (e.g., Baumwell, Tamis-LeMonda, & Bornstein, 1997; Bornstein & Tamis-LeMonda, 2001; Camaioni, 2001; Homer & Tamis-LeMonda, in press).

In light of these social-cognitive achievements, it is unsurprising that the 9-10 month period has also been referred to as the "dawning of language" as reflected in infants' rudimentary understanding of simple words and phrases (Bates et al., 1979; Volterra, Bates, Benigni, Bretherton, & Camaioni, 1979). These initial developments in language comprehension can be distinguished from the more sophisticated understanding of words and phrases that is exhibited by slightly older children. Specifically, infants' early understanding of words is routinized and depends on the contextual cues that accompany speech acts; infants appear to have a very rudimentary grasp of the meanings of words, and early word learning is often restricted to a specific context and specific actions (e.g., Barrett, 1995; Bloom, 1973; Camaioni, 2001; Camaioni et al., 2003; Tomasello, 1992; Snyder, Bates, & Bretherton, 1981).

We have documented the transition from restricted to flexible understanding over the period of 9 to 13 months based on data obtained from biweekly language interviews conducted with mothers (as described above). Virtually all infants displayed restricted understanding at the onset of language, with words becoming increasingly flexible with development. In the initial 2 months of interviews, the majority of infants demonstrated at least rudimentary understanding of their name and the names of familiar people (e.g., caregivers and siblings; mama, dada), basic foods (bottle, Cheerios), simple commands (e.g., "No!"), and performatives and games (e.g., "bye-bye," "clap hands," and "where's baby?"). Although mothers reported that their infants understood these simple words/phrases, further probing revealed that infants' understanding depended on the cues that accompanied these verbal expressions. For example, in several instances, infants responded to phrases such as "pattycake" or "clap hands"

with the appropriate clapping behavior; however, their responses were cued by parents' own clapping, nodding of the head, smiles, and melodic chanting of "pattycake." Parents' facial expressions, vocal intonations and gestures cajoled infants to join in on the fun, and lent familiarity to the continued repetition of the words associated with "pattycake."

In other instances, infants responded to words/phrases in generalized, diffuse ways, again suggesting only a tenuous grasp of the word's true meaning. For example, several infants were reported to laugh, flail their arms, and kick their legs in response to hearing phrases that contained the word "out" or "outside" (e.g., "We're going out. Out? Out? Do you want to go out?"), as though anticipating an impending excursion. However, these verbal phrases were typically accompanied by the assemblage of a stroller, packing a bag, and walking towards the door. Again, infants' reactions appeared to be context-dependent, making it questionable whether their emotional reactions of enthusiasm signaled true understanding of the meaning of the word "out." (To illustrate this limited nature of early understanding, Table 4.1 presents a diary excerpt taken from the first language interview conducted with a mother of a girl who was 9.5 months at the time of the call.)

On the other hand, these early glimmers of understanding were meaningful precursors to more reliable forms of receptive and productive language, illustrating the ways in which infants' experiences with repeated and familiar events provide the structure for interpreting others' speech. Infants map words onto existing concepts or cognitive structures, which form the building blocks to language (e.g., Bloom, 1998a; Hoff-Ginsberg, 1997). By participating in familiar social and cultural routines, children develop scripts, or generalized event representations of their life experiences that are comprised of specific information about actors, actions, and the objects or recipients of their actions (e.g., Nelson, 1981, 1986).

Moreover, parents' reports about their infants' receptive vocabularies were neither arbitrary nor overly inclusive. Rather, mothers were very selective about the words they claimed their infants "understood." Each week mothers reported a select handful of words that elicited consistent behavioral responses from their infants. A typical pattern was for infants' restricted understanding of a specific word to shift to flexible understanding within 1 to 2 weeks, and many of the first words "understood" were those that appeared in children's productive language weeks or months later.

As expected, across the developmental period of 9 to 13 months, an increasing proportion of the words that children understood were flexible or independent of contextual cues. Thus, the word "pattycake" would eventually induce infant clapping in the absence of parents' clapping, and "Let's go out" would impel infants to toddle to the door before strollers appeared. Moreover, context-flexible words functioned to "bootstrap" the acquisition of new words. As children came to understand the meaning of specific verbs, for example, the nouns and noun phrases to those actions were more readily understood (e.g., "eat" facilitates mapping new food terms—"Eat your peas!", "Eat your Cheerios!", Eat your apple!").

TABLE 4.1

Infants' Restricted Understanding of Early Words: Example Diary Excerpt at 9 Months

Word/Phrase Reported to Be Understood

Infant's Response

Cues Necessary for Understanding

"No"

"Good girl" "Good job" "Stop" "Cheerios" "Clap your hands" "Peek-a-boo"

Stops and looks at speaker Smiles and gets happy Smiles and gets happy Hesitates

Smiles and/or bangs hands on table

Claps

Laughs

Uses a raised tone of voice

Claps or smiles

Claps or smiles

Uses a raised tone of voice

Shows item

Sings the phrase or claps first Covers infant with blanket or appears from behind furniture

THE ROLE OF PARENTS IN EARLY UNDERSTANDING OF WORDS

What role do parents play in children's early understanding of words? At the most basic level, parents, as all humans, express words simultaneous with their actions. Indeed, the context-restricted nature of infants' early understanding reveals the universal dependencies that exist between language and action. Infants parse the action streams that they observe into meaningful units, and link those units to the words that are used to describe them (Baldwin, in press). Returning to the example of clapping, as infants view the motion of hand-repeatedly-meeting-hand and hear the chanting words of "clap hands," speech acts come to be associated with perceptual experiences. Parents, therefore, provide infants with gestures, actions, and emotions that support natural links between words and referents.

However, in addition to availing infants of these naturally occurring connections, parents differ in the ways they verbally communicate to their young children, in terms of the timing of their responses and the content of what they say, and these supports affect the course of children's receptive (and later productive) language. As noted earlier, parents' verbal responsiveness is thought to be especially supportive of children's understanding of words because responsive verbal information is offered at times of psychological salience. Adults who label objects and events that are the target of children's attention constrain interpretive possibilities, bolstering conceptual connections between words and referents (e.g., Baldwin & Markman, 1989; Bloom, 1993; Bloom, Margulis, Tinker, & Fujita, 1996; Carpenter et al., 1998; McCune, 1995; Rogoff, Mistry, Radziszewska, & Germond, 1992; Snow, 1986; Tomasello & Farrar, 1986). For example, a mother might gesture toward a novel object to elicit her child's attention and then name the object at precisely the point when her child shifts attention toward the object ("Ball. This is the ball. It's blue."). Or, she might wait for her child to demonstrate interest in an object, and use that interest as a springboard for providing verbal information (e.g., "Yes, that's a ball you have in your hand"). In both instances, the adult is being responsive by cueing into the child's interests such that the child need not "guess" the topic of conversation or rely solely on natural constraints (Nelson, 1988).

We offer empirical support for the role of parents' responsiveness in children's receptive language. In one investigation dyads were visited in their homes at 9 months of age and were visited again 4 months later (Baumwell et al., 1997). The question of interest was whether mothers who were responsive to their 9-month-olds would have toddlers who later exhibited larger receptive vocabularies.

From videotaped sessions of mother-infant play, the base rate frequencies of infant vocalizations, looks, and bids to mother, and play and exploration of toys were coded. Each maternal reaction to each infant behavior was classified into one of five categories: (1) verbal responses (mother replies promptly, contingently and appropriately within 5 seconds to a change in infant behavior—e.g., "doggie" as child shifts attention to a dog); (2) elaboration (mother builds on her prior response by providing additional information—e.g., "Furry dog!", after her first response to the child's interest in the dog); (3) focus shift (mother attempts to redirect her child's attention to something new—e.g., "Look at the bus!" as child looks at a dog); (4) prohibitions/reprimands (mother verbally restricts her child's actions—e.g., "Stop"); or (5) miss (mother ignores or shows no reaction to child's behavior). In addition, "refocus" was coded if a mother attempted to redirect her unfocused child's attention to the play materials (e.g., "Look at the dog!"—as child is not focused on anything in particular).

Analyses revealed two factors of maternal behaviors at each age (9 and 13 months): Sensitivity (a factor that loaded on mothers' verbal responsiveness, elaborations, and refocusing attempts) and Intru-siveness (a factor that loaded on focus shifts, prohibitions, and misses). Sensitivity predicted children's language comprehension at 13 months, but Intrusiveness did not. Specifically, maternal sensitivity to 9-month-olds uniquely accounted for a significant 15% of the variance in the size of infants' flexible language comprehension at 13 months after controlling for infants' 9-month language and mothers' later responsiveness. Together, infants' comprehension at 9 months and mothers' sensitivity at 9 and 13 months accounted for 37% of the variance in infants' 13-month language comprehension. In contrast, Intrusiveness accounted for a nonsignificant 1% variance in the size of children's receptive vocabularies. The fact that Intrusiveness contained verbal information (e.g., mothers' talking about objects/events outside of their children's focus and verbal prohibitions), but did not predict children's language, highlights the special significance of maternal responsiveness for children's early understanding of words and phrases.

PRODUCTION OF FIRST WORDS

Somewhere around the start of the second year children produce their first formal words. Prior to this point, children have been adding words to their receptive lexicons for several weeks or months; consequently, early in development children's production of words lags behind what they are able to comprehend (e.g., Bates et al., 1979; Childers & Tomasello, 2002). Nonetheless, children's first spoken words signal an important transition in language development and greatly affect their social environments. Parents enthusiastically welcome children's first words and are quick to respond to their children's new verbal achievements.

Similar to the patterns documented for receptive language, infants' productive language shifts from being context-restricted to context-flexible. In the early stage of word production, children often imitate the words of others. Rather than spontaneously generating words on their own, children parrot back words that they have just heard in their daily interactions (e.g., Mother: "Where's your ball Johnny?" Child: "Ball"; Mother: "Say bye-bye!" Child: "Bye-bye"). Additionally, children produce words in reference to a limited range of exemplars in narrow contexts. A child might say the word "dog," but only to indicate her own dog. Moreover, words are used sporadically, sometimes appearing and then seemingly disappearing from children's lexicons. Thus, the word "dog" might be used one week, not used for several weeks after that, and might "reenter" the lexicon with regularity weeks later. Consequently, early progress in productive language growth is gradual and effortful over the course of the first several weeks or even months of achievement (Bloom, 1998b).

Over the next few weeks and months, word production becomes more generative, regular, as well as more generalized. Words that were once only used in imitation are now produced spontaneously across a broad range of contexts. Children will now use the word "dog" to refer to dogs in general, pictures of dogs, the memory of dogs, and in anticipation of getting a new dog.

THE ROLE OF PARENTS IN EARLY WORD PRODUCTION

To what extent is children's early production influenced by parents' verbal responsiveness? We have used two approaches to study the role of parents' verbal responsiveness at the early stages of word production. In the first approach, parents' responsiveness was examined in relation to the size of infants' productive vocabularies at the start of the second year (i.e., 13 months), the point in development when most children are expressing their first words. In the second, parents' responsiveness was examined in relation to the developmental timing of specific language milestones in early productive language.

In several investigations, mothers' responsiveness related to children's productive vocabulary sizes concurrently and predictively. For example, in one study, mothers who responded more to their 5-month-old infants had babies who displayed larger flexible vocabularies at 13 months (Bornstein & Tamis-LeMonda, 1989). In other investigations, mothers' responsiveness to their 13-month-old toddlers' exploratory and communicative initiatives was concurrently related to children's productive vocabulary sizes (Tamis-LeMonda et al., 2001).

Perhaps more revealing than the predictive validity of responsiveness for vocabulary size at the start of the second year is the finding that mothers' responsiveness affects the timing of productive milestones in early language. In one study, we asked whether mothers' responsiveness would relate to the developmental onset of infants' first imitations and first words—that is, when in development children first imitated a word and when in development children first used words spontaneously and flexibly.

To address this question, we utilized the statistical technique of Events History Analysis (also referred to as Survival Analysis; Willett & Singer, 1991, 1993) which is suited to modeling whether and by how much specific predictors (here responsiveness) affect the timing of target events (here imitations and first words). Time is considered along its continuum, and the conditional as well as cumulative probabilities of an event occurring can be modeled and plotted across successive ages. As a hypothetical example, if a group of children were assessed on their ability to "walk" from 6 through 18 months, the cumulative probability of children walking independently might be .00 at 6 months (as no children have yet achieved the milestone), .10 by 9 months, .15 at 10 months, .50 at 12 months, and so forth, until all children have achieved the milestone of walking, at which point the cumulative probability reaches 1.00. The median lifetime represents the point when half of the population experiences the target event (i.e., when the cumulative probability reaches .50).

We assessed the contributions of maternal verbal responsiveness at 9 months (before children had begun producing conventional words) to the subsequent onset of children's first imitations and first words by comparing baseline conditional and cumulative probability functions (i.e., those in the absence of predictors) with "fitted" functions (i.e., those that included predictors) in nested, hierarchical chi-square analyses (Tamis-LeMonda et al., 2001). As anticipated, responsiveness was a robust predictor of the two milestones, as indicated by highly significant changes to chi-square statistics when responsiveness was added to models. To illustrate, when the language development of children at extremes of responsiveness were modeled (i.e., the top and bottom 10th percentiles), nearly all the children whose mothers demonstrated high levels of responding with descriptions at 9 months were estimated to imitate words by 13 months. In contrast, only 60% of the children with low responsive mothers were estimated to imitate words at 13 months (see Figure 4.1). Similarly, nearly all the children whose mothers displayed high levels of maternal responding with affirmations, descriptions, and play prompts at 9 months were estimated to produce their first words by 13 months, whereas only 20% of the children with low responsive mothers produced their first words at this age (see Figure 4.2).

THE VOCABULARY SPURT

The achievement of 50 words in children's production, which typically occurs midway through the second year, is often considered a benchmark in early language development (e.g., Bates, Dale, & Thai, 1995; Bloom, 1973, 1993, 1998a; Bloom, Tinker, & Margulis, 1993; Nelson, 1973). Around the time

Age in Months

Figure 4.1 Modeling fitted survivor functions for the timing of first imitations: Estimated function for high levels of maternal responses with descriptions at 9 months (upper 10th percentile) and estimated function for low levels of maternal responses with descriptions at 9 months (lowest 10th percentile).

15 17

Age in Months

Figure 4.2 Modeling fitted survivor functions for the timing of first words: Estimated function for high levels of maternal responses with affirmations, descriptions, and play prompts at 9 months (upper 10th percentile) and estimated function for low levels of responding with affirmations, descriptions, and play prompts at 9 months (lowest 10th percentile).

15 17

Age in Months

Figure 4.2 Modeling fitted survivor functions for the timing of first words: Estimated function for high levels of maternal responses with affirmations, descriptions, and play prompts at 9 months (upper 10th percentile) and estimated function for low levels of responding with affirmations, descriptions, and play prompts at 9 months (lowest 10th percentile).

children accumulate 50 words in their expressive vocabularies, they experience a sudden acceleration in production, referred to as the vocabulary "spurt" or vocabulary "explosion" (see Bates et al., 1988; Bloom, 1973, 1993, 1998a; Gershkoff-Stowe & Smith, 2004; Reznick & Goldfield, 1992). For example, Bloom (1993) showed that children averaged 51 different words in their productive vocabularies at the time of a vocabulary spurt and that all children in her research reached the 50-word vocabulary mark within 1 month of showing a substantial acceleration in their productive lexicons (Bloom et al., 1996).

In our research, we have also tracked children's language development from first words through the vocabulary spurt, and have documented a sharp incline in word growth around the 50-word mark. Specifically, children produced an average of 5.9 words per month prior to the 50-word mark, as

Age in Months

Figure 4.3 The net growth in the number of new words children produced each month between the ages of 9 and 21 months.

Age in Months

Figure 4.3 The net growth in the number of new words children produced each month between the ages of 9 and 21 months.

9 11 13 15 17 19 21

Age in Months

Figure 4.4 The cumulative growth of the number of total words in children's productive vocabularies at each month between the ages of 9 and 21 months.

compared to an average of 39.3 words per month subsequent to the 50-word mark (Tamis-LeMonda et al., 1998). Of course, children varied enormously in both when they achieved the 50-word milestone as well as in their growth rates both before and after this period. Therefore, our focus has been on both average gains across children as well as language growth at the individual level.

To illustrate the phenomenon of the vocabulary spurt, Figure 4.3 presents the net growth in children's language, which referred to the number of new words children produced each month between the ages of 9 and 21 months. Figure 4.4 presents the cumulative plot of the number of total words in children's productive vocabularies each month. As shown in Figure 4.3, there is an overall upward trend in word production across all children (N = 107). Prior to 15 months of age, children added fewer than 10 words to their vocabularies each month; by 18-19 months they added 40-50 words per month to their productive vocabularies on average. However, these average developmental functions (both in terms of net growth, Figure 4.3, and cumulative growth, Figure 4.4) obscure the rapid growth in word production that exists in individual children, and also mask the fact that some children showed little to no improvement in their language over this time frame.

Figures 4.5 through 4.10 present data on individual children to illustrate both the phenomenon of the growth spurt as well as the dramatic individual differences among children in growth over time. Certain children added over 100 words per month to their lexicons (with one acquiring over 200 new words in a month; see Figures 4.5-4.8). Other children hardly added any new words to their productive vocabularies, as illustrated in the individual plots in Figures 4.9 and 4.10. Children such as those depicted in Figures 4.9 and 4.10 had therefore failed to achieve the 50-word milestone by the end of the study (i.e., at 21 months), and showed no acceleration in their word growth.

In addition to children exhibiting quantitative change in language around the middle of the second year, as indexed by impressive gains in the absolute number of words in their productive vocabularies, word growth may be qualitatively different from the more effortful process of language acquisition observed at the start of the second year. At around the 50-word mark, word production is uniformly spontaneous, new words are acquired with greater facility, and words are no longer as transitory as they had been during earlier stages of production. Moreover, once children acquire a substantial number of nouns and verbs in their lexicons, they demonstrate a remarkable capacity to learn novel words from just a few exposures or even just one exposure (e.g., Clark, 1995); this phenomenon has been referred to as "fast mapping" (Carey, 1978), and is demonstrated by children as young as 2 years

Child 1

Figure 4.5 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

Age in Months

Figure 4.5 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

(e.g., Bloom, 2000; Heibeck & Markman, 1987). Children's acquisition of a relatively substantial large number of "count nouns" in production promotes comparisons across objects and increased attention to object shape, both of which further accelerate the rate of word learning (e.g., Gershkoff-Stowe & Smith, 2004; Namy & Gentner, 2002; Smith, 2003).

The expression of words around the vocabulary spurt appears to require fewer cognitive resources or less "effort" than the expression of words at the start of the second year (Bloom & Tinker, 2001). Support for this notion derives from Bloom's detailed, microgenetic tracking of the language of 12 children who were seen monthly from 9 months to 2% years of age both in a laboratory playroom setting and in their families' homes (Bloom, 1993). The study consisted of hour-long videotaping of mother-child play sessions at each visit, and language diaries kept by mothers that chronicled the

Child 2

Figure 4.6 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

Age in Months

Figure 4.6 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

Child 3

Figure 4.7 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

Age in Months

Figure 4.7 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

words their infants understood and produced during the 1-month intervals between visits. This enabled Bloom to document children's achievement of 50 words, and to examine the consequences of this milestone for children's abilities to express emotions and play acts simultaneous with the production of words. From videotapes, children's expressions of words, emotions, and play actions were documented in real-time through a frame-by-frame coding system; this detailed approach permitted analyses at time sampled units of less than 1 second. From these data, the overlap of word production with other forms of expression was documented. Bloom found that children were least likely to display positive or negative emotions simultaneous with talking if they were in the early stage of word production, but were adept at the simultaneous expression of words and emotions once they had achieved the 50-word mark. These findings suggest that the expression of words at the vocabulary spurt draws upon

Child 4

Figure 4.8 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

Age in Months

Figure 4.8 Individual plots of children's productive vocabulary growth spurts at different points in development between 9 and 21 months.

Child 5

100 80 60 40

100 80 60 40

Age in Months

Figure 4.9 Individual plots of children's productive vocabulary between 9 and 21 months; trajectories of children who displayed little or no production of new words between 9 and 21 months..

Age in Months

Figure 4.9 Individual plots of children's productive vocabulary between 9 and 21 months; trajectories of children who displayed little or no production of new words between 9 and 21 months..

fewer cognitive resources than the expression of words at the start of productive language. Similar findings have been extended to children's abilities to simultaneously produce words while engaged in play activities (Bloom & Tinker, 2001), and our research has revealed that children's production of words is inhibited when they are attending to toys at the start of the second year when word production is newly emerging (Tamis-LeMonda & Bornstein, 1990). In short, Bloom's finding that children are better able to simultaneously "talk" and "act" or "talk" and "emote" during the vocabulary spurt, but not at earlier stages of language production, suggests that the expression of words mid-way into the second year does not tax children's cognitive resources as much as early word use.

Child 6

100 80 60 40 20

Age in Months

Figure 4.10 Individual plots of children's productive vocabulary between 9 and 21 months; trajectories of children who displayed little or no production of new words between 9 and 21 months..

FIRST SENTENCES: THE START OF GRAMMAR

Changes to the size of children's lexicon are paralleled by qualitative changes in the ways that words are used. We have found that even though most children were not yet combining words into sentences at the juncture of 50 words (and therefore are still in the "one-word-stage" of language), they used language to express a variety of functions and semantic categories. Children's earlier use of words to label objects (e.g., "ball"), participate in routinized exchanges (e.g., "bye-bye"), and label or request people (e.g., "mama," "dada") was replaced with word functions that referred to actors, actions, patients, objects of actions, locations, existence, possession, negation, objection and so forth (Tamis-LeMonda & Bornstein, 1994). For instance, younger infants most often said "mama" to label mother (nominal usage) or to plea for mother when distressed or upset (instrumental usage). In contrast, somewhat older toddlers used the word "mama" (or perhaps by this time, "mommy") to refer to mother as an actor, experiencer, recipient, patient, possessor, in anticipation, in memory, and so forth. By 18 months, for example, "mommy" might be used in mother's absence, while the child pointed at her shoes, thereby expressing the notion of possession (as in "those are mommy's shoes!"). This more flexible use of words at the vocabulary spurt suggests that words function as true "symbols" in that they are no longer integrally bound to the concrete here-and-now, but can be used in the service of different meanings across past, present, and future timeframes.

The expansion of children's vocabularies and growth in the functional use of words toward the middle of the second year closely precedes the combination of words into simple sentences. Children's first "sentences" express those categories of meaning that had already been in evidence in their vocabularies at the later one-word stage. Therefore, "mommy shoe" might be used to convey the notion of possession (as in "That's mommy's shoe.") or might alternatively be used to indicate actors and the objects of their actions (as in "Mommy's putting on her shoe."). Both constructions build on children's earlier ability to use the word "mommy" and "shoe" flexibly across different categories of meaning (e.g., to recognize that the symbolic word "mommy" can be used as a possessor as well as actor depending on context; and that "shoe" refers to objects owned and objects that can be acted upon). Consequently, the emergence of combinatorial speech reflects not only a child's ability to specify different symbolic categories of meaning individually (e.g., "actor", "object of action"), but to also encode the roles and relationships between these separate concepts by placing them into basic grammatical structures (Bloom, 1998; McCall, Eichorn, & Hogarty, 1977; Fenson et al., 1994).

THE ROLE OF PARENTS IN THE VOCABULARY SPURT AND EARLY GRAMMAR

What role does parents' responsiveness play in children's language development during the period of the vocabulary spurt and beyond? We speculated that verbal responsiveness would be associated with more rapid growth in children's productive language over the second year; would predict the diversity of meanings children use to express their verbal constructions; and would be associated with the developmental timing of both the vocabulary spurt and combinatorial speech (i.e., combining words into sentences). Empirical support has been obtained for each of these predictions.

In one study, we assessed the role of maternal responsiveness in two contexts, play and mealtime, for children's growing vocabularies across the second year (i.e., when toddlers were 13 and 20 months; Bornstein, Tamis-LeMonda, & Haynes, 1999). Here the focus was on changes to mothers' responsiveness over time in relation to changes to children's observed language production. In this investigation, measures of mothers' responsiveness and children's language at both ages and in both contexts derived from transcripts of their observed interactions.

Mothers who grew in their verbal responsiveness over time had children who demonstrated significant gains in their own vocabularies over the 7-month period, both in play as well as at mealtimes. These analyses covaried children's own 13-month contributions to language outcomes, which were also significant. Perhaps most notable was the finding that the absolute size of mothers' observed productive vocabulary did not predict children's language above verbal responsiveness, nor did children's language growth predict later maternal productive vocabulary size. In contrast, mothers' verbal responsiveness mattered above the amount of mothers' language.

In a second cohort, predictive associations between verbal responsiveness and the timing of the milestones 50 words in production and combinatorial speech were examined (Tamis-LeMonda et al., 1998), based on our biweekly interviews with mothers. Word counts on children's productive vocabularies offered data on when children reached 50 words in language production. During these same telephone surveys, each mother was asked whether her child was yet combining words into simple sentences, in order to document the date of this later milestone. Probing for combinatorial speech was highly specific and the crediting of achievement was conservative. For example, if a mother stated that her child was "putting two or more words together" she was asked to provide specific examples of the child's construction as well as information about the situation(s) in which her child expressed the phrase. The child was only credited with "combinatorial speech" if he or she (1) linked two or more words in a single phrase without pause; (2) each of the words in the phrase could be classified as independent words in the child's vocabulary; and (3) each of the words in the phrase could be classified into distinct categories of speech (e.g., actor, action, object of action, patient, possession; for additional details, see Tamis-LeMonda & Bornstein, 1994, Tamis-LeMonda et al., 1998).

For each outcome, the "date" when the child achieved each of the milestones was noted as the date of the telephone call, providing a window of error within 2 weeks. Again, using Events History Analyses, mothers' verbal responsiveness to their 13-month-olds (at the start of language production) was examined as a predictor of when in development children achieved the two milestones. Findings revealed that verbal responsiveness predicted the timing of the vocabulary spurt and combinatorial speech, and did so above the timing of children's first words in production, the timing of their achievement of 50 words in receptive language, and mothers' responsiveness at 9 months. Similarly, responsiveness at 13 months contributed unique variance to the timing of combinatorial speech over and above the timing of first words in production, children's earlier receptive language, and responsiveness at 9 months.

Moreover, specific forms of responsiveness related to specific targets of child behavior. In particular, children of mothers who responded contingently to children's vocalizations and play behaviors achieved 50 words in expressive language and engaged in combinatorial speech sooner than children of less responsive mothers. In addition, specific maternal responses at specific developmental periods facilitated the achievement of specific language milestones. For instance, mothers' affirmations and descriptions at 9 months, but not at 13 months, predicted children's language milestones. And, mothers' responses with vocal imitations and expansions at 13 months, but not at 9 months, influenced the timing of children's production of 50 words and combinatorial speech (Tamis-LeMonda et al., 2001).

To provide an example of these effects, Figures 4.11 and 4.12 illustrate the role of mothers' responsiveness in children's language spurt, as defined by 50 words in production. Specifically, Figure 4.11 contrasts the effects of high levels of maternal responsiveness to children's play and vocalizations at 13 months (upper 10th percentile) with low levels of maternal responsiveness to children's play and vocalizations at 13 months (lowest 10th percentile). Figure 4.12 contrasts the effects of high levels of maternal responding with imitations to children at 13 months (upper 10th percentile) with estimated functions for low levels of maternal responding with imitations at this age (lowest 10th percentile). When the language milestones of children at the extremes of these various forms of responsiveness were modeled (i.e., the top and bottom 10th percentiles), approximately half of all children whose mothers demonstrated high levels of responsiveness (either in terms of what they responded to—Figure 4.11—or how they responded—Figure 4.12) at 13 months were estimated to achieve 50 words in language production by 15 months of age. In contrast, only half of the children of low responsive mothers were estimated to achieve 50 words in production by 21 months of age, the end of the study. Not shown are the very similar patterns for the timing of combinatorial speech. Specifically, one-third

- High Resp

Age in Months

Figure 4.11 Modeling cumulative probability functions for the timing of 50 words in children's productive language: Estimated function for high levels of maternal responsiveness to children's play and vocalizations at 13 months (upper 10th percentile) and for low levels of maternal responsiveness to children's play and vocalization at 13 months (lowest 10th percentile).

Age in Months

- High Resp

Figure 4.11 Modeling cumulative probability functions for the timing of 50 words in children's productive language: Estimated function for high levels of maternal responsiveness to children's play and vocalizations at 13 months (upper 10th percentile) and for low levels of maternal responsiveness to children's play and vocalization at 13 months (lowest 10th percentile).

of all children with highly responsive mothers were estimated to engage in combinatorial speech by 15 months, as compared to 21 months of age for children of low responsive mothers.

EARLY LANGUAGE AND COGNITIVE DEVELOPMENT IN LOW-INCOME FAMILIES

Thus far, emphasis has been on children's early achievements in receptive and productive language, and the facilitative role of parents' verbal responsiveness for these emerging abilities. In addition,

-High Resp

Age in Months

Figure 4.12 Modeling cumulative probability functions for the timing of 50 words: Estimated function for high levels of maternal responding with imitations to children at 13 months (upper 10th percentile) and estimated function for low levels of maternal responding with imitations to children at 4-13 months (lowest 10th percentile).

Age in Months

-High Resp

Figure 4.12 Modeling cumulative probability functions for the timing of 50 words: Estimated function for high levels of maternal responding with imitations to children at 13 months (upper 10th percentile) and estimated function for low levels of maternal responding with imitations to children at 4-13 months (lowest 10th percentile).

most of this work has been conducted with Anglo American children and mothers from middle- to high-socioeconomic status households (as determined by composite scores of parents' income and education). More recently, we have extended work on environmental correlates of children's language development to low-income, ethnically diverse families. Research on the language development of children from the low-income families is particularly important in light of studies that document the adverse consequences of poverty for children's cognitive and educational achievements (e.g., Burns, Griffin, & Snow, 1999; Brooks-Gunn & Duncan, 1997; Hoff, Laursen, & Tardif, 2002). In general, children living in economically disadvantaged households tend to have smaller vocabularies and less-than-optimal language competencies when compared to their more advantaged peers. As one example, children in poverty enter kindergarten with an average lexicon of about 5,000 words in comparison to vocabularies of approximately 20,000 words for children from higher-income families (e.g., Hart & Risley, 1995). These differences are thought to partly be explained by differences in the cognitive stimulation observed in low-income parents (e.g., Evans, 2004) and children's diminished participation in learning activities (e.g., bookreading; Anderson, Teale, & Estrada, 1980; Hart & Risley, 1995; Whitehurst, Arnold, Epstein, Angell, Smith, & Fischel et al., 1994).

Nonetheless there exists great variation in the language of children from low-income families. For example, 18-, 24-, and 36-month-old children living in poverty have MacArthur CDI scores that range from below the 10th to above the 90th percentile, even though their overall scores are lower than those of the normative population (Roberts, Burchinal, & Durham, 1999). Such findings challenge deficit models that emphasize group means at the expense of individual variation. Unsurprisingly, in light of this within-group variation, questions about the language environments of children living in poverty have been at the foreground of applied developmental research.

In response, and as an outgrowth of our earlier work, we have pursued two research directions on the language development of children from low-income families. First, we have begun to document children's language interactions with both their mothers and fathers. What is the role of fathers' language on children's language achievements in low-income households? Do patterns observed in mothers also extend to fathers? Second, we have extended inquiry on parent-child language engagements to incorporate children's broader learning experiences. This latter emphasis takes into consideration the various learning activities that children participate in with their parents (e.g., bookreading, storytelling) as well as the learning materials that are available to children (e.g., books and toys) (Tamis-LeMonda et al., 2004).

Language Engagements with Mothers and Fathers

There has been burgeoning attention to fathers' role in children's development over the past 20 years, and a number of large-scale national efforts have been designed to examine the nature, antecedents, and consequences of father involvement in low-income families more specifically (e.g., see Cabrera et al., 2002; Lamb, 2004; Tamis-LeMonda & Cabrera, 2002, for reviews). This heightened interest in father involvement stems from a number of social and demographic trends, including women's increased participation in the workforce, the prevalence of single-headed households in low-income families, and the adverse consequences of father absence for children's school readiness, academic performance, and social-emotional regulatory competencies (Tamis-LeMonda & Cabrera, 1999).

Limited resources, unstable employment, and inadequate education often make it difficult for fathers to establish and maintain positive and emotionally supportive relationships with their children (Black et al., 1999; Brophy-Herb et al., 1999; Cochran, 1997; Furstenberg & Harris, 1993; Garfinkel, McLanahan, & Hanson, 1998; Lerman, 1993; Marsiglio, 1987; McAdoo, 1986, 1988; McLoyd, 1989, 1990; Perloff & Buckner, 1996). These same obstacles also pose practical challenges to researchers who seek to understand the nature and meaning of fathering in economically disadvantaged, ethnically diverse groups (Cabrera et al., 2004; Tamis-LeMonda &

Overcome Shyness 101

Overcome Shyness 101

You can find out step by step what you need to do to overcome the feeling of being shy. There are a vast number of ways that you can stop feeling shy all of the time and start enjoying your life. You can take these options one step at a time so that you gradually stop feeling shy and start feeling more confident in yourself, enjoying every aspect of your life. You can learn how to not be shy and start to become much more confident and outgoing with this book.

Get My Free Ebook


Post a comment