Baby Babble & Early-Life Language Acquisition
If you’ve ever spent time with an infant, you know that, even though they can’t speak, they can most definitely make a lot of noise. In addition to crying and screaming (with or without reason), babies engage in vocal experimentation, pronouncing different sound sequences that usually involve a lot of repetition. Whether it be cooing, the use of tone and vowels while crying or yelling, or babbling, bursts of consonants in succession, babies love to test out their articulatory capabilities. It is both a process of mimicry and exploration; they might try to copy the sounds they hear around them or they might just see what their mouths can do.
You’ve probably heard of a pregnant person putting headphones around their belly and speaking to their child through a microphone. While this is a silly image, it draws on something very real. The ears of babies start to develop at the beginning of the third trimester and before long, they can actually begin to “hear” things. It probably sounds strange, similar to trying to listen to something underwater, but it is sound perception nonetheless. For this reason, babies actually start to “learn” the languages spoken by their parents before they are even born. Of course, no child comes out speaking fluently; an infant’s oral cavity and brain are much less developed than a toddler’s. But, they begin to recognize what feels familiar to them and they exhibit behavior that gives them access to the familiar elements of the world. Because of this, if you put a minute-old newborn in a room with their mother and ten other people, they will be able to recognize their mother just by the sound of her voice. Even crazier, older infants are often savvy enough to recognize which language is their native one before they can even speak it. Babies are human machines of input in a sense, taking in and processing whatever is given to them.
For many years, linguists have sought to understand how and when infants start to categorize the languages that they hear and favor certain sounds over others. Most languages make use of allophones, when the sound that is physically pronounced does not match the way it is mentally understood. This sounds complicated, but English is full of them. The t-sound in English is pronounced differently in many different words. Pronounce the following words out loud and pay close attention to the way you say the t…
true
tape
water
kitten
If you paid too much attention, you might have tricked yourself into saying one of them in a way that isn’t actually natural for you, but for most English speakers, the first word would start with the ch-sound as in “chair.” The second word begins with what would be considered the normal t-sound that starts the letter name itself. The “t” in water actually sounds closer to a “d” in American English. The double “t” in kitten turns into a glottal stop, the same sound found in the middle of the phrase “uh-oh.” Some of these sounds vary between dialects. Round your lips on the “a,” pronounce the glottal stop for the “t” in water, and drop the “r” and now you’ve got a London accent. Allophones don’t exist simply because we’ve memorized the pronunciations of these words either. If I told you to say out loud the made up word trokey, you would know to use the ch-sound, even if you might not be able to tell me why. Regardless, it’s clear that we can hear sounds that don’t match physically and still mentally categorize them as the same. It’s one of the reasons why we understand people who speak with an accent different from our own and why we tend to think of our own selves as accent-less.
While most languages have allophones, they don’t always make use of the same ones. In Spanish, they have different allophones that use sounds that we don’t even have in English. Take the word for cat: gato/gata. Putting the respective gendered article in front of this word would result in el gato and la gata. Even though these are essentially the same word, the g’s in them don’t match. A native Spanish speaker would pronounce el gato with a “g" like we have in English. But, when g’s are in between vowels (intervocalic) in Spanish, they turn into a different sound which is pronounced in the same place as a “g” but is instead a fricative sound, like “v” or “z.” A fricative is a sound which constricts airflow without cutting it off completely. So, while “g” in el gato stops the air in the mouth and then releases it, the “g” in la gata sends the air through a tight opening without complete closure. Try to pronounce “f” without letting any air out of your mouth. It isn’t possible because this is a fricative sound. The symbol for the fricative form of “g” is ɣ, the voiced velar fricative. You can hear it pronounced here. This whole concept might be confusing, and it definitely is a complicated mental process that we are still trying to study, but it is something that comes naturally to the native speakers of a given language. Screwing it up doesn’t render a message completely meaningless, but it is a mark of accent.
Place of articulation for the voiced velar fricative, ɣ
What does this have to do with babies? Well, a lot. As young children hear languages spoken around them, they pay attention and try to replicate the sounds that they perceive. Babies make many sounds that are not very useful from a linguistic standpoint, such as blowing a raspberry, which is not used in any language as a consonant. The older a baby gets, the more exposure they have to the way the languages of their family and community members speak. They start slowly and in a way that is not very complex, usually with nasal sounds (m, n, ng) and plosive sounds (t, d, b, p). It’s no coincidence that the words for mother and father in most languages are made up of simple syllables that involve repetition, such as mama, dada, papa, or baba. This is also the reason why so few toddlers can say words such as grandma or grandpa and instead favor something like mimi or nana or pop-pop. Words that involve consonant clusters are much more difficult for a young child to perceive and replicate because they are more complex. Which foreign word do you feel more comfortable pronouncing: aloha kakahiaka (good morning in Hawaiian) or zdravstvuytye (hello in Russian). Chances are you favor the one without difficult clusters. A toddler would feel the same.
There are several processes that children cycle through in order to simplify the language that they hear and desire to communicate with. Think about a time that you were trying to speak to a young child. Did they ever say something that you didn’t understand and you asked them to repeat it a few times before you got what they said? Maybe you never understood on your own and you had to ask their parents for clarification. This is normal, and you might think afterwards,”how did they come up with that?”Maybe the word they said sounds even more complicated to you than the word they were trying to say. Well, think of it this way. If you have never picked up a basketball, even if you’ve watched the sport on TV for years, you aren’t likely to have a perfect free throw the first time you try. Similarly, little kids aren’t likely to know how to perfectly articulate the words in their native languages, no matter how much they hear it in their daily life. They have to practice and be given pointers from older people in order to develop their speech. Even though everyone has their own idiolect, language only works if it is intelligible between individuals.
Ways that children simplify their languages include: reduplication (wawa for water), coalescence (feak for speak), gliding (woad for road), devoicing (dok for dog), reduction (nana for banana), and more. Most of these we can understand readily; we probably employ processes such as gliding when engaging in baby-speak or Motherese, as it’s been called by linguists. But, take the coalescence example, feak for speak. How does a child come up with this? Speak begins with a consonant cluster, sp-. Coalescence is a process where two sounds converge into one. “S” is a fricative sound that is pronouced behind the teeth and “p” is a plosive sound that is pronounced with the lips. These two different sounds become one, taking qualities from both sounds and presenting as something in between. Thus, “f” is a fricative sound that uses both the upper teeth and lower lip. It is a simple sound for a child to perform, even if causes difficulty for a listener.
Some examples of phonological processes
As a child grows older, they begin to fall away from these simplification processes. In English, there are certain speech processes that children go through that really inhibit understanding. These are the ones that parents and family members will instruct the child on and help them work through. A good example of this would be denasalization: the transformation of a nasal consonant into a non-nasal one. A child might simplify the word maybe by changing the “m” to a “b.” Unfortunately, the resulting word is already a word in English, baby. Once a child’s caretaker understands that they actually mean maybe while saying baby, they will likely help the child pronounce the word correctly in order to limit this misunderstanding in the future. If the child says bore while meaning more, the parent will reinforce the lesson, and gradually the child will learn that their simplification process actually complicates communication for them. A toddler will stop denasalization very early, but not all simplifications cause problems of the same severity. Adding vowels to break-up consonant clusters doesn’t usually matter or change the message, so you might hear a child saying turue for true or pulace for place until they are eight or nine years old.
Children are for the most part very perceptive of the world around them and the ways that they are instructed. Of course, kids can be stubborn and opinionated as well, but they can also form new habits with greater ease. While it is very hard to change one’s speech habits well into adulthood, at a younger age humans can adapt to language naturally. It is one of the reasons why kids can learn a second language much faster and better than a thirty or forty year old person. Children are also very attuned to the sounds that their languages deem useful and the differences between them. Back to what we were talking about before, languages make use of allophones, sounds that are different but which are mentally recognized as the same. One of the earliest things that a baby learns is what categories of sounds their language uses.
Babies are born as blank slates. They learn something from every situation that they encounter. Studies have been done on infants which test whether they contrast certain sounds. How is this possible if they can’t communicate? Linguists have to get creative. Babies naturally turn their heads towards a sound when they perceive it as new. They also will suck with a greater intensity if something is in their mouth. There is a special pacifier which calculates how hard a baby is sucking it, i.e. high-amplitude sucking. Using these methods, several people have studied how sensitive babies are to subtle sound changes. These experiments have found that very young infants (0-3 months) can hear the differences very well, while older infants (9-12 months) can only hear the differences that are relevant and contrastive in their languages. In Russian, there are two different types of “sh” sounds, one further forward in the mouth and one further back. Both an English speaking and Russian speaking baby can hear this difference at 3 months, but by 9 months, the English speaking baby loses this ability. They simply don’t need to know the difference because it doesn’t help them when speaking English. The same happens for Mandarin speaking babies at a similar age. In English, we distinguish between tense and lax vowels (eat vs it) but these sounds are not contrastive in Mandarin. A year-old child growing up in China probably would not be able to hear this difference, but a child who was only just born would likely be able to.
An infant with a pacifier that tests for high-amplitude sucking
This might seem astounding that children can be so intuitive, but it is something that we all did at one point in our lives, relatively unconsciously. Many more studies will be done in the future to understand the brain activity and the processes that young kids go through to create mental sound categories; it is hard to fully comprehend a human activity when those humans can’t yet talk! Child language acquisition is a fascinating field with lots left to learn. I’m sorry to my future kids if I treat you like test subjects every once in a while. I think many linguists have probably made a similar apology.
