Which of the following provides evidence against the behavioral view of language development

31.10 Conclusion

Language development is inherently a process of change. Exploring the multiple and varied trajectories of language can provide us with insights into the development of more general cognitive processes. Studies of language development have been particularly useful in helping us to understand the emergence of specialization of function and the scale and flexibility of cognitive processes during learning. Novel approaches and technologies for capturing the linguistic environment that the developing child grows up in (Greenwood, Thiemeann-Bourque, Walker, Buzhardt, & Gilkerson, 2011)—and for capturing what the child is saying (Oller et al., 2010)—should allow for more fleshed out theories and models of how language development actually works. Correspondingly, new tools for understanding brain structure (Dick et al., 2012; Glasser & Van Essen, 2011; Sereno, Lutti, Weiskopf, & Dick, 2013), development (Dosenbach et al., 2010), representation (Huth, Nishimoto, Vu, & Gallant, 2012) and learning (Wiestler & Diedrichsen, 2013) should allow us to make much finer-grained predictions about when, where, and how language development changes the brain.

Assumptions and ideas about language learning have changed dramatically in the last two decades. Nativist assumptions have been dispelled by neuroscience, modeling, and psycholinguistic data. First-language learning begins before birth, and its foundations are laid down during infancy. During the second year toddlers begin to interpret and produce multiword utterances, refer to absent or abstract topics, and use social context to interpret language. During the third and fourth years children master most of the structure and uses of various linguistic elements, and acquire a great deal of critical lexical knowledge. Finally, during middle childhood the nuances of complex structure and dependencies are refined as children learn to use language more flexibly and consciously.

Introduction

The ability of human children to acquire a language is one of the hallmarks of the species. Within 24 h of being born, infants already show evidence of having learned aspects of the broad rhythmic structure of their mother tongue, most likely from hearing speech in utero during gestation. Infants continue to attend to the patterns in their language and start to learn a considerable amount about its structure well before they start combining words in their own speech. After just a few years, children will have mastered many of the complex grammatical structures in their language, and persistent systematic errors become rare as children approach their tenth birthday. Children exposed early on to additional languages will learn them without any significant added difficulty compared with monolingual acquisition, provided they have sufficient exposure to each language.

Language development in children is remarkable for its regularity across individuals and different languages, even when the languages are quite diverse. Children universally progress through many of the same stages, mastering certain components of the grammar before others and making characteristic errors in production and comprehension at each stage. Despite individual differences in intelligence and other abilities, and despite normal differences in language exposure, all children end up able to speak and understand their native language(s) fluently. This is in marked contrast to adult learners, who show much greater variability in learning processes and ultimate outcomes. Indeed, differences between children and adult learners in ultimate outcome, in the kinds of errors made, and in the stages of acquisition have led many researchers to conclude that there is a critical or sensitive period for language development. As in other developmental domains with critical periods, learning that takes place outside this period is more variable and less robust, leading to a greater range of outcomes and more idiosyncratic patterns of acquisition.

In addition to the uniform patterns of acquisition across languages, there are some aspects of language development that are tied more closely to structural properties of the particular language being learned. Comparing components that are more or less language dependent can lead to insights about the learning mechanisms involved in language development and the properties of the linguistic input to which learners attend.

This article gives an overview of early language development, discussing some key phenomena in acquisition and some major theoretical issues. Other articles offer more-detailed discussions of specific areas within language development.

Abnormal language development used to define autism, but it no longer does. Language development no longer even figures into contemporary diagnostic criteria, although early delays in language often lead to parents’ concerns. In this chapter, we review recent empirical research on language development in autism. To paint a contemporary picture, we restrict our review to studies published in the twenty-first century. We conclude the following: language development in autism is often delayed, but not deviant; a delay in language development is not unique to autism; and language development in autism is remarkably heterogeneous.

Conclusion

The CHILDES system has led to significant improvements in research in child language acquisition over the last 20 years, and will continue to do so with the new developments taking place. It is a huge benefit to the research community that there are now large amounts of data for English, and increasing amounts of data from a wide range of other languages. However, it is necessary for researchers to exercise caution. First, over-reliance on the CHILDES database can lead to the extensive analysis of data from just a few children, which may bias our findings in various ways. Second, when comparing children using data from CHILDES, it is important to carefully control for the age and linguistic stage of the children concerned, and to ensure that the corpora are comparable in terms of transcription procedures and coding before carrying out any analyses. Third, the increasing trend toward analyses at the lexical level means that although the CHILDES system provides data from many children, relatively few of these children have enough data in a dense enough time period to support such analyses. These issues mean that although the CHILDES database is a valuable resource, it is essential that researchers continue to donate new data to CHILDES to promote rigorous cross-linguistic research into language acquisition.

The Child Language Data Exchange System (CHILDES) promotes sharing of transcript data and shared standards for transcription and documenting using computerization. It built on a history of sharing child language transcript data established by Roger Brown, adding automation of analysis and procedures for compiling data across sets of transcripts. The CHILDES database now encompasses corpora in 26 languages, corpora of bi- and tri-lingual children, of children and adults with language disorders, and elicited narrative. A successor system, TalkBank, offers digitized audio and video records, web-enabled access, and richer coding capacities to language researchers.

Resilience in the face of internal variation

Language learning is also resilient in the face of many organic variations from the norm, variations that alter the way children process whatever input they receive. For example, intermittent conductive hearing losses from repeated middle ear infections can cause a child’s intake of linguistic input to vary over time in amount and pattern. Despite this variability, spoken language development for the most part proceeds normally in children with this type of hearing loss. As a second example, blind children live in a nonvisual world that is obviously different from the sighted child’s world, and that offers a different spectrum of contextual cues to meaning. However, this difference has little impact on language learning in the blind child.

Organic variation can be much more severe and still result in relatively intact language learning. For example, grammar learning in the earliest stages can proceed in a relatively normal manner and at a normal rate even in the face of unilateral ischemic brain injury. As a second example, children with Down syndrome have numerous intrinsic deficiencies that complicate the process of language acquisition. Nevertheless, most Down syndrome children acquire some basic language reflecting the fundamental grammatical organization of the language they are exposed to (the amount of language that is acquired is in general proportion to their cognitive capabilities). Finally, and strikingly, given the social impairments that are at the core of the syndrome, autistic children who are able to learn language are not impaired in their grammatical development, either in syntax or in morphology, although they do often have deficits in the communicative, pragmatic, and functional aspects of their language.

Interestingly, even when children do have trouble learning language, some properties of language (the resilient ones) are spared. For example, a basic understanding of the organization that underlies predicates appears to be intact in children with specific language impairment (children who have neither hearing impairment, cognitive deficit, nor neurological damage yet fail to develop language normally). However, these children have difficulty with morphological constructions. As another example, children who are not exposed to a usable language until adolescence have no trouble mastering word order when learning language late in life, but do have difficulty with morphology. Some properties of language appear to be robust, and some fragile, across a variety of circumstances and internal states.

There may be no greater testament to the resilience of language than the fact that children can invent language in the absence of a model for language. A combination of internal factors (the fact that the children are profoundly deaf and cannot acquire a spoken language) and external factors (the fact that the children have not been exposed to a conventional sign language) together create the unusual language-learning circumstances in which the deaf children described earlier find themselves. Despite their lack of a model for language, these children still communicate in language-like ways.

In sum, language development can proceed in humans over a wide range of environments and a wide range of organic states, suggesting that the process of language development may be buffered against a large number of both environmental and organic variations. No one factor seems to be ultimately responsible for the course and outcome of language development in humans, a not-so-surprising result given the complexity and importance of human language.

Infancy

Language development is a fascinating and complex behavior. One of the reasons language development is so remarkable is because it is never explicitly taught to young children. Language is learned primarily through immersion in a linguistic environment. Even before children are born, they are exposed to linguistic information and are learning in utero. As young as four-days-old, infants prefer listening to their own language compared to a foreign language. Young infants also show the ability to distinguish between phonetic elements of speech. By 6 months, infants evidence canonical babbling represented by repeated consonant/vowel combinations.

Even before children say their first word, they are communicating intentions through gestures, smiling, eye contact, and vocalizations. At around 12 months of age, most children can produce 10 spoken words and can comprehend as many as 50. By their second birthday, children move from the use of one word utterances to communicate through the use of telegraphic speech, which is characterized by two word phrases that only contain important words.

1.2 Individual variations in adult language learning success

Language learning during adulthood compared to childhood is notoriously more difficult (Newport, 1990). The challenges appear to be universal across all aspects of language, including speech sounds, vocabulary, grammar, and ultimately the overall language proficiency. The capacity to discriminate non-native speech sounds fades during early development. Infants go through a “perceptual narrowing” phase before their first birthday, when they become increasingly more tuned to native speech sounds and less tuned to foreign speech sounds (Kuhl et al., 2008; Werker & Hensch, 2015). Adults seem to preserve some degree of sensitivity to non-native speech sounds, as evidenced by laboratory-based training studies finding widely distributed learning outcomes across individuals after intensive training (Chandrasekaran, Sampath, & Wong, 2010; Myers, 2014; Wong & Perrachione, 2007).

Vocabulary learning during adulthood has also been shown to be markedly difficult, potentially owing to many contributing factors, including the lack of translation equivalents across certain languages (Malt & Sloman, 2003) and differences in phonological memory (Baddeley, Gathercole, & Papagno, 1998; Service & Kohonen, 1995), and language training context (Collentine & Freed, 2004; Lan, Fang, Legault, & Li, 2014; Milton & Meara, 1995). The most consistent findings were that there were large variances in L2 learning success in each of these learning contexts, with no single language learning context or methodology equally benefiting all individuals. Due to the vocabulary learning difficulties seen across adult learners, an increasing number of studies have been geared toward providing additional scaffolding and developing new technologies to aid the mass of students struggling to learn L2 vocabulary (see Lan, Chen, Li, & Grant, 2015 for a review).

Regarding grammar learning, a seminal study by Johnson and Newport (1989) provided the earliest evidence for a sensitive period for attaining second language grammatical abilities. English grammar abilities of native Chinese or Korean speakers were highly related to their age of arrival in the United States, but the individual differences in English grammatical skills were no longer explained by the age of arrival after puberty (Johnson & Newport, 1989). Using computational modeling and web-based grammatical assessments collected from a large population of native and non-native English speakers around the world, a recent study suggested a relatively stable grammatical learning ability until about 17.4 years of age, which then declines steadily afterward (Hartshorne, Tenenbaum, & Pinker, 2018). Both studies highlight the increasing difficulty of learning with age. These findings raise more questions regarding what factors beyond the AoA contribute to the individual variation in grammar learning success.

Decades of behavioral research have also examined the cognitive, motivational, and environmental influences on holistic language learning outcomes (Dekeyser, 2012; Linck et al., 2013; Sparks, 2012; Sparks, Patton, & Ganschow, 2012). The set of skills documented as potential building blocks for adult language learning aptitude ranges from linguistic-specific to domain-general skills, such as perceptual sensitivity (Chandrasekaran et al., 2010; Wong & Perrachione, 2007), executive functions (Linck et al., 2013), memory capacity (Ettlinger, Morgan-Short, Faretta-Stutenberg, & Wong, 2015; Wong & Ettlinger, 2011), and first language skills (Melby-Lervåg & Lervåg, 2011; Sparks, 2012). These behavioral findings imply a much wider brain network associated with adult language learning success that is beyond the traditional language regions associated with native language processing. The remaining questions are whether the left hemisphere also dominates the adult language learning process and whether neural diversity in the left hemisphere underlies the variations across learning outcomes.

LANGUAGE AND COGNITION

Language development is intricately entangled with cognitive development. Much discussion and research have been devoted to the question of whether language is necessary for cognition (i.e., whether we need words and semantic structure to think) or whether language is merely an outcome or reflection of growth in mental capacity. This chicken-and-egg question aside, the clinician will find that language assessment becomes increasingly enmeshed in other areas of functioning as a child grows because the usual evaluation of language, using standard instruments, is really one of cognitive growth and vice versa. Most often, language and cognitive development are linked in an individual child, but not necessarily so. Language development, particularly in a population of disabled children, is highly predictive of overall functioning, and language remains the best predictor of cognition in general. School systems are heavily biased toward verbal skills, so it is not surprising that school performance and linguistic abilities are linked. Early language disorders are highly predictive of learning difficulties in school, and most studies show that greater than 40% of children with early language difficulties will have learning difficulties in school. The language concern may be the chief complaint, but a comprehensive look at the child and the environment is needed when this problem becomes evident.

Both directions are possible. Emotional and cognitive deficits may be secondary to primary language disorders, particularly in a society that places so much emphasis on verbal competency and formal schooling performance in its children. Conversely, emotional disorders may have language delay symptoms. In any case, a language-delayed child must be carefully and comprehensively assessed for these associated difficulties.