Where is language located in the brain? Is the human brain specialized for language? Are there sensitive periods for acquisition? Is any aspect of language innate? Are there learning mechanisms dedicated to language? Wernicke and Broca identified language areas for comprehension and production in the left hemisphere, and modern studies rely on PET, fMRI, and MEG for tracking just where information is processed. It is unclear whether there is a sensitive period for language acquisition. Evidence from brain injuries and feral children is problematic. Evidence from second language learning is rarely comparable in amount of experience, feedback, and practice to first language. As children acquire more language, they process it faster, with greater left-hemisphere specialization. With bilingualism comes greater density in the left hemisphere. Sign languages are also processed in the left hemisphere. But some aspects of language are processed in the right hemisphere. Language is part of a more general system of communication, with affect, facial expression, gesture, and stance, so storage in the brain occurs in both hemispheres.

Children learn to distinguish registers for different roles: talk as child versus as adult, as girl versus boy, as parent versus child, as teacher, as doctor, marking each “voice” with intonation, vocabulary, and speech acts. They learn to mark gender and status with each role; what counts as polite, how to address different people, how to mark membership in a speech community (e.g., family, school, tennis players, chess players), and how to convey specific goals in conversation. They reply on experts for new word meanings and identify some adults as reliable sources of such information. They mark information as reliable or as second-hand, through use of evidentials. They adapt their speech to each addressee and take into account the common ground relevant to each from as young as 1;6 on. They keep track of what is given and what new, making use of articles (a versus the), and moving from definite noun phrases (new) to pronouns (given). They learn to be persuasive, and persistent, bargaining in their negotiations. They give stage directions in pretend play. And they start to use figurative language. They learn how questions work at school. And they learn how to tell stories.

More than half the world is bilingual or multilingual. So when growing up exposed to two (or more) languages at once, children have two systems to learn, and they must also learn when to speak each language. The choices here depend on who the addressee is, and on the setting. Exposure to the two languages may be uneven, and also vary over time, depending on who the child spends time with. Choice of language depends on common ground, on the topic, and on the language common to the child’s conversational partners. The early stages of acquisition are very similar, from perception of sounds and sound sequences to early babbling; from comprehension of words to attempts to produce them. Early vocabularies contain many doublets, freely accumulated as children learn more of each language. (This is consistent with contrast, but not with mutual exclusivity.) Language mixing tends to mirror adult usage and so varies across languages. Children attend not only to differences in the sound systems but also to structural differences of all kinds. Conversational skills develop in similar ways across languages, depending on exposure and practice, with language dominance fluctuating over one’s lifetime. Acquiring two dialects involves similar skills.

As children learn more about language, they use it more effectively to achieve their conversational goals. They choose appropriate speech acts, establish joint attention, contribute new information, take up information from others, and take turns. They learn how to enter an exchange among others from as young as age two. Their intrusions in ongoing exchanges typically contain new information. Planning an utterance takes time, and children learn to plan what to say so as to take turns on time. This can be tracked in their answers to yes/no and wh- questions, where they get faster with age. They plan pretend play, assigning roles, assigning actions, and also utterances for each character enacted. They track common ground and design referring expressions for their addressees, and they repeat new words to mark uptake. They distinguish requests from offers, and, on occasion, persist in making repeated requests themselves. They clarify what they mean when asked and offer spontaneous repairs as well. In all this, they track what the others in the exchange say and choose when to enter the exchange themselves.

Children add further complexity by combining two or more clauses. They can link them with coordination or subordination. In subordinate constructions, one clause is embedded in the main or matrix clause. The embedded clause can fill a grammatical role, as subject, say, in complement constructions, or it can modify parts of the main clause, adding to a noun phrase with a relative clause, or to a verb phrase with a temporal clause. These constructions allow for more options in the flow of information as well as in the expression of more complex events. Among the first constructions here are coordinations of different elements in a clause, as well as of different clauses. Among subordinate clause constructions, because, what, when, and so were the most frequent up to 2;9, followed by if, that, and where. They produced relative clauses to specify referents; complements with verbs like think and know. And they produced temporal, causal, and conditional constructions to describe sequences of events. Children treat clause order first as reflecting the actual order of events, only later assigning the appropriate meanings to connectives like before and after. And they take time to master the meanings of because and if.

Infants attend to speech, even before birth, preferring what they heard in utero over another language. In the first months, they can identify certain sounds as the same. (Other species also display categorical perception of sounds.) By 7 to 8 months, infants can identify recurring sequencies of both nonsense syllables and real words. Measures used include high-amplitude sucking, head-turn preferences, and visual fixation. In segmenting speech, infants face two problems: variation in speakers’ word uses, and which sounds to group as the same. They can also recognize a few words. By 9 to 10 months, infants attend more to distinctions in the ambient language than distinctions outside that language, and appear sensitive to legal clusters of sounds that begin and end syllables. They also prefer the prosodic stress patterns of the ambient language and so can use multiple cues to possible word-boundaries. By age three and later, children manage to recognize words even when spoken in another dialect or with a foreign accent. Their recognition of words at age three shows their representations in memory are adultlike, and not based on their own productions. They store representations of words in memory for recognizing words from other speakers, and as targets for their own production.

This chapter looks at early interactions between parent and infant, from joint looking at faces, gaze following, and attention to hands and gestures, to later interactions where infants and adults readily capture each other’s attention. It examines the ways adults modify their speech to infants and young children, e.g., with short, grammatical utterances, formulaic routines, repetitions in variation sets, higher pitch, slower rate, and pausing at the ends of utterances. Adults adjust their speech to what their children understand and provide feedback on children’s errors, checking up to make sure they have understood them and so offering them a conventional way to say what they appear to intend. Adults establish joint attention and engage with infant and child activities, anchoring their conversational contributions to what is physically present and visible, and talking about the child’s current activities. And infants become adept at attracting adult attention and enlisting their help in different activities. In child-directed speech, adults focus on what is physically and conversationally present, and respond to the topics children introduce. They choose short, high-frequency words, with high neighborhood density, many with concrete referents present in the here and now. Conversational interactions provide the setting for acquisition.

Grammatical morphemes are used to modulate word meanings and to link words in constructions. They consist of inflections, usually suffixes, added to words, and free-standing function words (prepositions and articles). Different language-types make different uses of these, including cases added to each noun, tense and aspect markers added to each verb, and agreement markers linking nouns, adjectives, and determiners. Children have to identify each inflection, its meaning, and where it is used on each word class. They start to add modulations to their words as soon as they start to combine words (and may understand some of them before this). They use regular forms as their starting point and over-regularize irregular forms. They show some consistency in the order of acquisition for different modulations, depending on the semantic complexity of each grammatical morpheme. Semantic complexity, formal complexity, and frequency all play a role here. Children may initially rely on filler-syllables, and only later produce the relevant form. Word class plays a role here, since the choice of grammatical morphemes depends on this. Initial use of grammatical morphemes may be limited to specific words and only later extended. The same holds for agreement in gender and number.

Language use is a skill that requires exposure to language, feedback on usage, and practice. So children need exposure from expert speakers, feedback on the language being acquired, and on any errors children produce, and practice along the way. Languages differ, so the paths children follow within and across languages may vary, and some constructions may be harder to acquire in one language, easier in another. The goal is to learn to use language for communication. Language is essentially social, relying on common ground. Part I (Chapters 2-6) focusses on how adults talk with children; children’s analysis of the speech stream; their first production of words; and how they assign meanings to words. Part II (Chapters 7-11) focusses on children’s acquisition of structure: elaborations of information inside clauses, and combinations of clauses. They also rely on structure when coining new words. Part III (Chapters 12-14) looks at turn-taking, learning to be polite, persuasive, and informative, and how to tell stories. Children who hear two languages have two such systems to learn. Part IV (Chapters 15-16) summarizes evidence for biological specialization for language and considers how continuity and change are reflected in language processing.

Children’s first words may bear little resemblance to adult forms (/ga/ for squirrel), and they may at first produce multiple versions of words. Mastering adult pronunciations takes time and practice. Babbling allows practice with some sounds, but those sounds may only appear in words years later. Early vocalizations, at times paired with gestures, often precede recognizable words. And children take time before they can produce versions that match adult productions. Some 70% of children’s words up to age four are inaccurate and variable in form. Children may select some word forms over others in production because they are easier targets. They simplify words by substituting sounds they can manage; they assimilate some sounds to neighboring ones; they omit some sounds, especially in clusters. And they may set up templates for multisyllabic words, focussing on stressed syllables over unstressed ones. As they get older, they make spontaneous repairs to words that don’t match their stored representations, and also repair when asked by others. Finally, they practice words, to themselves, and in their daily language use. Single words are easier than word combinations, so longer utterances take still more practice.