Child Directed Speech
Child Directed Speech
Every normal child acquires a natural language. A child hears sentences of the target language and uses this raw material to generate a system which enables the child to produce sentences of his/her own. A child might be exposed to a range of different human language involving a large amount of variations in phonology, lexis and grammar. Each of these distinctive inputs produces a distinctive output (Foster-Cohen, 2009). Therefore, the input to the child bears a remarkable resemblance to the final outcome of the child development. Any child can learn different languages with equal ease and therefore become polylingual, if the child is exposed to the languages (Foster-Cohen 2009). Environmental inputs are a fundamental key in child language development. However, a crucial question is how this input is sufficient to explain the childīs construction of grammar and if there are other factors that also contribute to the process of acquisition.
The behaviourist view focuses on language as a learnt behaviour by emphasising the childīs imitation of the language of others and the role of adult reinforcement. Adults, when talking to children, adapt their speech in different ways at every level of linguistic analysis; this type of input is called: child directed speech (CDS) (OīGrady, 2005). Its phonological features are characterised by exaggerated intonation and stress; higher pitch and a greater pitch range; slower speech with longer pauses between utterances and after content words. Features of vocabulary in CDS include a restricted range of vocabulary items, the topics that seem to recur frequently are kin, animals, food, clothing and parts of the body. Moreover, there seems to be three times as much paraphrasing and more reference to the here and now. Regarding syntax, CDS involves more commands, questions and repetitions; the utterances tend to be shorter and grammatically simpler than in adult directed speech, and also fewer broken sentences (OīGrady, 2005).
It can be argued that CDS is not syntactically simple. CDS involves the frequent use of questions and imperatives together with declaratives. In English, declaratives show the least deviation from deep structure, maintaining the underlying order of elements Subject NP-Aux-V-Object NP, and in which none of these elements is deleted. In imperatives, the subject NP is deleted and in questions the order of the Aux and the subject NP is inverted. It would be expected that the child might acquire the most frequently offered order and use that as his/her basic order of elements. However, English-speaking children do not start using verb-first orders in their declarative sentences, even though imperatives and questions which show a verb-first order are more frequently used than subject-first declaratives in CDS (Fletcher and Garman, 1986). This is because declaratives and almost all questions show the same order of subject and main verb. In English, it is only the auxiliaries which are preposed in questions, and furthermore there is evidence suggesting that children at the beginning may not attend to auxiliaries because they are unstressed and unfamiliar. The basic order in English is SVO, therefore this basic order of elements is modelled for children more often than the distribution of utterances across sentence types might suggest (Fletcher and Garman, 1986). In addition, frequency plays an important role in syntactic acquisition. It is only after several years of experience with language, that a child dare formulate the big generalisation that all subjects come before verbs in English. On the other hand, it is also essential to highlight that the simplest CDS may not be the best basis for learning language, as the child uses the input to formulate hypotheses which are modified only when disconfirming evidence becomes available (Fletcher and Garman, 1986).
First language acquisition is believed to proceed on positive evidence, the language to which the child is exposed. It can be argued that discourse features of CDS are of importance in vocabulary and syntax development. The isolation of objects and actions, together with the appropriate verbal labels and key gestures, facilitates vocabulary. Furthermore, some studies have demonstrated a connection between expansions/recasts and syntactic development (Snow and Ferguson, 1977).
Expansions refer to a situation when parents repeat the childīs utterance, expanding it to a normally constructed grammatical phrase. However, no positive effect has been found that expansions accelerate language acquisition. On the other hand, an experiment in which children received not only expansions but also recasts did demonstrate an effect on childrenīs language ability (Snow and Ferguson, 1977). Recasts refer to the situation when parents repeat the childīs sentence in a new syntactic form. Recasts seem to play a role in helping children to identify and extract specific grammatical morphemes, yes-no questions, copula, auxiliary verbs and syntactic structures in general (Ayoun, 2003). Furthermore, Saxon (1997) found that children reproduced the correct irregular past tense forms more often and with fewer mistakes following negative feedback as opposed to positive input (cited in Ayoun, 2003). This negative feedback is a weaker form of correction than negative evidence, since it does not model the correct alternative for the child instead the adult provides a prompt to previously learned forms. As Chomsky (1965) states, negative feedback assumes that the child has already been apprised of the adult form, but has produced a mistake which is identified as performance based instead of competence-based (cited in Foster-Cohen 2009). However, as OīGrady (2005) states, mothers leave the vast majority of their childrenīs ungrammatical sentences alone, and also one study reported that mothers repeat as many as one third of their childrenīs incorrect utterances without making any changes. However, these errors presented in the input do not seem to hinder language acquisition. Therefore, recasts do not always have an apparent effect, at least not a reliable one.
It remains to be explained how CDS provides information about syntactic structure. Children need to understand the sentences that they hear without having a lot of knowledge of the target language. According to Macnamara (1972), the acquisition of syntax could be only explained if it is recognised that children collect information about the relationship between syntactic forms and semantic structures (cited in Fletcher and Garman, 1986). In other words, children find out the syntactic rules by using the cues provided by the meaning of the adultīs utterance. As Macnamara (1972) states, children interpret adult utterances by relating the words used to aspects of the situation being described (cited in Fletcher and Garman, 1986). Therefore, a child could observe that the word referring to the agent precedes the word referring to the action, inducing a rule about the order of those semantic elements. However, CDS provides a very restricted set of semantic content. CDSīs utterances are limited to the present tense, to concrete nouns, to comments on what the child is doing and on what is happening around the child (Fletcher and Garman, 1986). In conversations between parents and children, the mother follows the childīs lead in deciding the topic of the conversation. Therefore, regarding the semantic level, the adultīs speech is shaped by the childīs linguistic abilities. It can be argued that semantically relevant speech is advantageous for language acquisition, but it is not crucial.
As mentioned previously, the behaviourist view sees the parental interaction with the infant as the basis of linguistic development. However, it can be observed that this view fails to explain the fact that children regularly produce ungrammatical structures, which are unlikely to be produced by adults. It also does not provide an argument explaining the fact that parents do not provide corrections all the time and even that sometimes parents repeat their childrenīs incorrect utterances without making any changes. Finally, it also fails to account for the fact that all children manage to learn to speak and understand a language at about the same time even when they all receive different input and are all raised in an environment that vary considerably. Moreover, Pinker (1994) states that 'in many communities of the world, parents do not indulge their children in motherese' (cited in Foster-Cohen 2009, p. 74). Therefore, if CDS is not universally available it cannot be recognised as a necessary or even important feature of language acquisition. However, according to Foster-Cohen (2009) indications that aspects of CDS are present in non-Western cultures have been ignored.
In some cultures some features of CDS are present whereas others are not. There are cases, such as the example of Kīicheī Maya where a number of classic features of CDS are absent such as phonological adaptations of cluster reduction or the use of special sounds. However, a number of other features of CDS are present such as repetitions and expansions. Moreover, in the Trackton and Western Samoan society, adults are also reported to provide repetitions of child utterances (Foster-Cohen, 2009). It is crucial to consider that no one can obtain evidence of how parents talk to their children in private. CDS can be seen as inevitable if parents attempt to have a conversation with their children; some of the characteristics of CDS could be seen as adjustments made in response to cues from the child (Foster-Cohen, 2009). CDS can be acknowledged as a way to communicate with children instead of a way to teach children language. Therefore, the use of CDS is not critical to the language learning process.
As explained previously children need some linguistic input in order to learn language. However, CDS does not provide children with all the possible structures in the target language. In other words, the evidence available to children is insufficient to determine the linguistic knowledge that they must acquire. The input does not tell the child the range of possible meanings that a sentence has, and also nothing in the input tells the child which sentences are ungrammatical (Lust, 2006). Language grammar is too complex to be learnt by children only through experience. Therefore, the fundamental question is how children come to know the important hypotheses about their language, and the way the language exposure works. According to Barrett (1999), if children did not have any innate grammatical knowledge, it would be impossible for them to acquire grammar form the language input which they receive. Moreover, as Crain and Martin (2001) state, what is innate does not have to be learned.
In nativist approaches, Gibson and Wexler (1994) state, the child simply needs to hear key linguistic forms which will act as triggers for setting the parameters of Universal Grammar (UG) ( cited in Foster-Cohen 2009). Chomsky (1966) argues that children are born with innate knowledge which is known as the Language Acquisition Device (LAD). This knowledge includes principles common to all human languages: Universal Grammar. The LAD is the means by which the child analyses the linguistic input which is called Primary Linguistic Data (PLD). This LAD hypothesises a series of grammars and the last grammar formulated is the adult grammar (cited in Crain and Martin, 2001).
Children learning the same language follow a similar pattern. They progress through the same stages of acquisition, and in the same order. However, the rate at which children pass through these stages may vary (cited in Crain and Martin, 2001). Parents do not teach their children language, but they may provide a language model by talking to and around their children. Furthermore, despite cultural variations all babies in the world learn to speak at roughly the same age. In cases where parents do not completely master the language of the new community, their children acquire the new language without any difficulties. The same happens when children are raised in an environment in which an impoverished language such as pidgin or creole is spoken; children tend to use forms that are more grammatical than those of their parents (Boysson-Bardies 2001). In other words, every normal child acquires a natural language without any special sequenced language input.
This examination of CDS shows that children learn to speak when they grow up in a linguistic environment. The linguistic input must be sufficient for the child to be able to categorize speech sounds and to specify the principal parameters of his/her language. Furthermore, CDS is not a crucial part of first language acquisition; however this does not eliminate the possibility that the Language Acquisition Device might find it helpful in certain respects. It is important to highlight that the linguistic input must be accessible within a framework of interactive communication involving the child and those around him. Therefore, CDS is not used to teach language to young children, instead the general goal is to maintain a live interactive communication with the child.
Essay by: Merche Gómez Vázquez
This resource was uploaded by: Merche