This section looks at how children develop the various components of mental grammar: phonetics, phonology, morphology, syntax and semantics. As we’ll see, children’s minds are equipped to acquire grammar very quickly as long as they have other language users to observe. Children’s rapid language development can look quite magical from the outside, but it results from immense brain activity in response to the language environment, also known as the ambient language, the language used by the family members, caregivers, adults and older children that a child spends time with. The baby brain tracks the patterns in the language environment and uses them to build up the child’s own mental grammar. If the language environment is English, the child develops a mental grammar for English. And obviously, children in a Tamil-speaking environment develop a mental grammar for Tamil, children in a Mandarin-speaking environment develop a mental grammar for Mandarin, and so on. If the ambient language is Auslan, then the child will develop a mental grammar for Auslan, regardless of whether they’re deaf or hearing. If a child is in an environment where two or more languages are used, they’ll develop mental grammars for both those languages.

2.9.1 When does language learning start?

Early exposure to language input is super-important because the mind starts building a mental grammar from a very young age. Just how early does language learning start? For hearing babies, it can start even before birth! If we measure the heart rate of a fetus in the uterus, we find that the heart rate increases in response to external sounds at about the seven-month point in pregnancy, so we know that a fetus can hear outside noises even while it’s still inside the uterus. Not only can fetuses hear, but they’re also remembering some of what they hear: by eight months of pregnancy, a fetus’s heart rate increases more in response to their pregnant parent’s voice than to a stranger’s voice (Kisilevsky et al., 2003). This means they’ve stored some memory of what their parent’s voice sounds like.

Ok, so by looking at fetal heart rates we can conclude that hearing babies have already stored some auditory memories by the time they’re born. But that doesn’t necessarily mean they have any mental grammar, does it? How can we tell what newborns know about their language? After birth, there are so many more interesting stimuli in their world that measuring heart rate isn’t as informative, but they certainly can’t tell us what they know. What can we observe that would tell us something about mental grammar?

Using this high-amplitude sucking habituation method, researchers have found that newborns whose parents speak French notice the difference between French and Russian sentences spoken by the same person (Mehler et al., 1988). The fact that these newborn infants are sensitive to this difference tells us that they must have some memory of the patterns of French, to be able to tell it apart from Russian. What could these babies with French parents have learned about French before they were born? A lot of the sound information they receive in the uterus is quite muffled, but what they do have access to is the prosody. The rhythmic pattern of French depends on syllables of similar length, while that of Russian depends on syllable stress. That prosodic rhythm is audible to a fetus in the uterus, so by the time they’re born, babies can tell the difference between the rhythm they’ve heard before and an unfamiliar rhythm.

In fact, if the pregnant parent speaks more than one language, a newborn can even tell the difference between those two languages! In a high-amplitude sucking study in British Columbia, babies born to Thai-English bilingual parents noticed when the spoken language switched from English to Thai (Byers-Heinlein et al., 2010). This suggests that their minds are already starting to set up two different mental grammars for the two languages they’ll be acquiring!

So even when they’re just born, hearing babies have not only learned what their parent’s voice sounds like, they’ve also already learned some of the prosody of the language (or languages!) spoken in their environment.

I also want us to remember, though, that language acquisition is not a race. In some cultures, parents like to believe that their child is “advanced” in some way: smarter or stronger than other babies. Even though this chapter talks about general patterns of acquisition, remember that there is huge variation across individual children, and learning something earlier isn’t necessarily any advantage. The reason I mention that here is there’s no evidence that a fetus that doesn’t have access to speech sounds in utero, either because the fetus is deaf or the parent uses signed language, is disadvantaged in any way. As soon as they’re exposed to language in a modality they can access, they start building their mental grammar.

2.9.2 Phonemic contrast

There’s another part of the mental grammar that hearing babies start to learn well before they can speak. Remember that the phonology of each language is specific to that language: the patterns of which features and segments contrast with each other and which are simply allophones is different in each language of the world. So, for example, we know from that in English, aspirated [pʰ] and unaspirated [p] are both allophones of a single phoneme. But in Thai, these two segments contrast with each other and are two different phonemes. The phonetic difference is the same, but how that difference is organized in the mental grammar is different in the two languages. Phoneme contrasts are a classic example of unconscious linguistic knowledge: in all likelihood, nobody ever had to teach you that [k] and [b] are different sounds and that the words cat and bat refer to two different animals! But if no one ever taught it to you, how did you learn it? How did the phonology of your first language come to be in your mental grammar? Researchers use a different habituation method for slightly older babies.

Using this technique, linguists and psychologists have learned that babies are very good at noticing phonetic differences, and they can tell the difference between all kinds of different sounds from many different languages.

The research on babies acquiring signed language came later, but it showed exactly the same pattern (Baker Palmer et al., 2012). Babies younger than six months noticed the difference between phonemically contrastive American Sign Language (ASL) handshapes even without any exposure to ASL, just like the babies without Hindi experience noticed the Hindi spoken contrast. By age 14 months, ASL-acquiring babies had retained their ability to recognize phonemically contrastive handshapes, but the English-learning babies without ASL experience had lost that sensitivity.

So for either sign language or vocal language, a child’s mind has built up phoneme categories in their mental grammar by about age one, according to the contrasts that they have experience with in their language environment. One thing to note here is that this doesn’t mean that it’s impossible to learn new phonemic contrasts in a new language, but that later learning will be shaped by the learning that has happened in the first year.

2.9.3 Early language production

We learned previously that babies are already learning a lot about their language even before they can talk. If they’re learning so much so fast, why don’t they talk right away? When babies are just born, their bodies just aren’t ready yet. Newborns can certainly cry, which uses the lungs and the vocal tract. But a newborn’s larynx is higher in the vocal tract than an adult’s: it starts lowering around age six months. And newborns just aren’t very good at controlling their bodies yet! It takes them a few months to begin to learn how to move their articulators. They begin to gain control of their fingers, hands and forearms first, and their ability to control their jaw, tongue and lips comes a bit later. So in those first few months after birth, their hand movements and vocal productions are kind of random: mostly wiggling and gurgling.

Babbling

Once babies start to get their bodies organized, their productions tell us that their mental grammars are starting to get organized, too. Starting around age six months, the sounds babies make begin to have some syllable structure. They start to produce reduplicated consonant-vowel syllables. Usually the vowel in these syllables is the low [a] made with the jaw fully open, and the consonants are usually voiced stops. Why voiced stops? They’re still learning to control their vocal tract, so it’s easier to alternate between fully open for vowels and fully obstructed for stops than to produce the consonants with partial obstruction like fricatives or approximants. Same thing for the voicing: if they just keep the vocal folds vibrating for vowels and voiced consonants, it’s easier than alternating between voiced vowels and voiceless consonants. So the classic baby babbling sound, which linguists call canonical babbling, is [babababa] or [dadadada]. As they continue to grow and gain control of their muscles, they start to vary the vowels and consonants, and their babbles might sound more like [badibadi] or [ɡudaɡuda].

Vocal babbling starts at about age six months for both hearing and deaf babies, but deaf babies gradually produce less vocal babbling as they continue to grow. When the language environment is a signed language, babies start to babble using their hands. Both hearing and deaf babies who grow up with sign as an ambient language produce reduplicated syllables using handshapes from the language environment (Petitto et al., 2004; Petitto & Marentette, 1991). Sign-acquiring babies’ first productions more often use the proximal articulators (the ones closer to the torso, like the elbows and wrists) than the distal ones (the articulators farther from the torso like fingers and knuckles) (Chen Pichler, 2012).

So we have some evidence that babbling isn’t just random sounds: whether the language environment is vocal or sign, babies start to produce forms that are organized similarly to the language in the environment:

• Their babbles are made up of repeated patterns that have the structure of syllables.
• They alternate between handshape and path movement or between a closed and open vocal tract.
• Their babbles use a subset of the segments/handshapes that appear in the language environment.

So when you see or hear a baby babbling, it might look or sound like random nonsense, but what they’re really doing is exercising their mental grammar!

First words

When babies babble, they’re practicing making the forms, that is, the signs or sounds of the language they’re acquiring. Remember  that a word links a form with a meaning. So how can we tell if a child who produces a form has a meaning linked to that form? In other words, how can we tell if they’re babbling or producing words? We have to look at the context. If an English-acquiring child says [baba] when they’re reaching for their bottle or for a ball then they’re probably using that form to refer, so it counts as a word even though it doesn’t have the same form that the adult word [bɑɾəl] or [bɑl] does. But if they’re producing [baba] just for the fun of making the sounds, then their utterance is non-referential, – it doesn’t have meaning, so it counts as babbling.

Babies acquiring sign languages often sign their first words at about age 0;8 or 0;9, while it’s usually later than that for babies acquiring a spoken language, closer to age 1;0. This difference might be because babies develop muscle control of their hands and arms earlier than of their tongues and lips. It could also be because sign words more often have iconic forms than spoken words.

When children start to produce and understand words, the first words in their vocabulary are quite similar even when we compare across languages and modalities. Usually the first word meanings they acquire are for referring to things that are common and observable in the immediate environment, like names for their family members and pets, the word baby, and words for common objects like milk, ball, shoe. It’s also common for their first words to include greetings like hi and bye, and other expressives like uh-oh and no. If there are verbs in their set of first words, the verbs are likely to refer to actions like cry or eat.

Of course, because one-year-olds don’t have a lot of experience with the world yet, they often haven’t got adult-like meanings in their mental grammar. For example, I know a toddler who saw a pumpkin for the first time and declared, “Apple!”. If the child hasn’t yet got a mental representation for pumpkins, gourds and squashes, they might well overextend the meaning of apple to include many other roundish fruits. It’s also common for children’s word meanings to be underextended, so, for example, the word elephant might refer to a particular stuffed animal but not to any other elephants, real ones or toy ones. In short, children develop meanings for words based on their experience of encountering the word in their environment.

Landau and Gleitman (1985) provide some very interesting examples of how children’s word meanings are shaped by their experience of the environment, from their research comparing blind children to sighted children. You might guess that a blind child doesn’t really have a concept for the verb look, since they can’t see, but that’s not actually the case. In one experiment, the researchers asked children to “look up!”. Sighted kids tilted their heads to face the ceiling, even if they were wearing a blindfold. But when they asked a blind child, whom they called Kelli, to “look up”, she kept facing forward and put her hands up toward the ceiling! So does that mean that Kelli’s meaning for the verb look is the same as for the verb touch? In the next experiment, the researchers put an object in front of Kelli and said, “You can touch this but don’t look at it.” She tapped or stroked the object, and then once they told her, “Now you can look at it”, she ran her hands all over the object to explore it. So just like sighted children, Kelli’s mental grammar had two distinct meanings for the two verbs look and touch. It’s just that her meaning for the verb look was different from that of sighted children, since her experience of the world was different.

Understanding word combinations

We’ve seen that babies start to learn the phonology of their first language very early — as soon as they’re born and maybe even earlier than that! And they’re starting to learn some lexical semantics before their first birthday. What about syntax? What do babies and young children know about the syntax components of their grammar? And how can we tell? We know that children start to speak or sign their first word around twelve months, and they start to combine two or more words around eighteen to twenty-four months. But if you’ve ever spent any time with young children, you know that they can understand a lot more than they can say! Their comprehension is often much more advanced than their ability to produce spoken or signed words. But comprehension is much harder to observe. How can we tell what babies and toddlers understand about language?

There’s so much learning happening in that second year. Kids are learning new words very rapidly, and learning how words pattern in the morphological and syntactic behaviour. In fact, by the time they turn two, kids are sensitive to verb arguments and subcategories too! In another split-screen experiment (Arunachalam & Waxman, 2010), when experimenters presented the novel verb mooping in a transitive frame, like “The lady is mooping my brother,” then two-year-olds looked more often to the scene where one participant is doing something to the other, like pushing. But when the novel verb appeared in an intransitive frame, like “The lady and my brother are mooping,” then the children looked more often to the scene where the two participants are doing the same activity together, like waving. This suggests that, by age two, children are sensitive not only to syntactic categories, but also to subcategories! So to go back to the question we started with, “What do kids know about syntax?” It turns out the answer is that, even before children start combining words to make phrases in their own speech or signing, they already know quite a lot about how words combine in the grammar.

Syntax in early utterances

Beginning at about age 1;6 to 2;0, most children start to combine words into phrases of two or even three words. At this early stage, their utterances are usually telegraphic, containing mostly content words like nouns and verbs, with few function words and few inflectional morphemes. It can be hard to figure out whether these short utterances have any syntactic structure to them. Let’s look at some of the evidence.

Corpus data (StromswoId, 1995) show that English-acquiring children start to ask questions not long after they begin combining words in their utterances. Simple questions like, “Where kitty?” and “Who crying?” are common in children’s speech, with the wh-word moved to the beginning of the sentence just like in adult English. Why questions show up a little later, but as any parent of a preschooler will tell you, the why stage feels like it lasts forever!

Developing word meanings

We saw previously that children’s early meanings for nouns are often over- or under-extended. Their concepts become more adult-like as they get more experience of things and categories of things in the world. something something noun classes. There is a difference between count nouns like pencils and cookies and mass nouns like rice and money and different languages encode this meaning distinction using different morphosyntactic strategies. In English, singular count nouns must have a determiner, while plural count nouns can appear without one. Mass nouns, don’t really allow a plural form, but are grammatical without a determiner in their singular form.

These morphosyntactic properties are salient enough that English-acquiring children distinguish between count nouns and mass nouns very early in their productions. Researcher Peter Gordon looked at a corpus of speech from two English-acquiring children between ages 1;9 and 3;6 (Gordon, 1988). He found that even before age 2;0, these children rarely produce pluralized mass nouns. In other words, their speech included many pluralized count nouns like dogs and balls, but almost no ungrammatical mass plurals like milks or sands. They did, on the other hand, produce some singular count nouns without determiners (like “I want cookie”), because of the telegraphic nature of their speech. Their production of obligatory determiners for singular count nouns became reliable only after age 3;1. Children’s semantics for gradable adjectives (e.g. big, little) includes some measure of degree along a scale, and the scale they choose depends on the context in which the adjective is uttered.

2.9.4 Morpheme development