How Language Develops

Learnings from neuroscience about language.

Posted Nov 01, 2019

When my daughter was approaching her first birthday, she said what we would proudly describe to everyone, probably for decades to come, as her "first word." It was "mammam," which is not even a real word in my native tongue, Tamil, but a cutesy baby-talk way of saying "food." Would this count as her first word then? I'd say yes because it was the first time she was putting what she wanted to say into something resembling language.

If there is one faculty that is unique to the human species, it would have to be that of language. Other animals indeed communicate with each other for survival, but they do so in ways that are distinct from human language. In addition to using primitive modes of communication, such as chemical signals (ants release hormone-like chemicals called pheromones to direct others in the colony to a food source) and facial gestures, some animals also show a certain degree of sophistication to their communication systems, albeit to a much lesser degree than that of human language.

The more complex ways in which animals communicate are vocalizations that are used, for instance, by monkeys, to warn others in their group of a predator (a species of monkey has even been found to have distinct calls to indicate different predators). Bees use a pretty sophisticated dancing pattern to inform other bees of a food source they have found—the longer the bee waggles, the farther the nectar source is from the hive. They can even indicate the direction of the flower patch by modifying the angle of the waggle.

Could the communication systems of animals be described as language too? This, of course, is simply a matter of definition. The Principles of Neural Science, the Bible for neuroscientists, describes language as "the remarkable system that allows people to communicate an unlimited combination of ideas using a highly structured stream of sounds (or, in signed languages, of manual and facial gestures)." What seems to make the human language unique, then, is that the number of ideas that can be communicated using a limited number of words and a set of rules, such as grammar, is infinite.

There have been the odd cases of research animals (primates, parrots, and even dolphins) having been taught to associate words with their meanings, but none of these animals have been able to communicate very meaningfully, or master that essential component of language, grammar, the way human babies seem to without much effort at all. In 1959, Noam Chomsky proposed the game-changing hypothesis that the ability to acquire language is governed by innate neural circuitry in humans. Various kinds of evidence have been put forth to support this still-controversial hypothesis (opponents believe that the ability to learn a language is merely a facet of general intelligence, and is not distinct neurobiologically from the capacity to learn general patterns).

One piece of evidence in support of Chomsky's hypothesis is that the use of language and the mastery of grammar develops without any formal training in children, and this ability cuts across cultures and backgrounds. This is not to say that language acquisition is entirely innate, because, as with any other cognitive abilities, learning does play a role. Children raised in mute environments during the critical phase for language acquisition usually end up not learning to communicate using language at all.

I am by no means trying to provide additional support to Chomsky's hypothesis by anecdotal evidence, but it's interesting to me that my daughter Sam, who is now approaching her second birthday, seems to have already grasped a rudimentary understanding of the parts of speech. She's currently at the one to two words in a sentence stage, and will usually describe something as, say, "funny" or "nice." She always follows this description up with a clarification to us as to what it was that she was describing.

For instance, she'll say "funny" followed by "song," or "nice" followed by "mammam" (that baby-talk word for food, remember?). She's slowly but surely learning that adjectives are used to describe a noun, and this is mighty exciting. Of course, she's learned this because whenever she would describe something, my husband and I would ask her what it was she was describing; but it is a vastly exciting achievement, nonetheless, for that little brain to have observed this, and to have learned some of the rules of grammar without being explicitly taught them.

Another indication that children perform complex grammatical analysis of their parents' speech, rather than just mimicking them, is that the errors that children make while talking are highly systematic—which is to say that even their nonsensical sentences have a thread of grammatical sensibility running through them. Many children, once they reach the age of 3 or so and know that rules of grammar exist and have to be followed, occasionally fall into the trap of hyper-regulating their speech. Having a rudimentary knowledge of the past tense, they might say, "I broked the vase," for instance. Sam knows that anything plural has an "s" attached to it in the end. So she says, "So many waters," while having a bath. It's fascinating.

Wikimedia Commons. Polygon data were generated by Database Center for Life Science (DBCLS)
Location of the Broca's area in the human brain
Source: Wikimedia Commons. Polygon data were generated by the Database Center for Life Science (DBCLS)

Language is more than just words and grammar, though. In order to understand a sentence, the sounds that we hear must be mapped to the grammar and the vocabulary, and information about the resulting comprehension must be sent to the systems in the brain responsible for faculties such as reasoning and memory. Similarly, for producing a sentence, motor areas of the brain must receive input about how to articulate the sentence correctly. Language production and comprehension, thus, require the concerted effort of many different parts of the brain. The current neurobiological framework for language suggests that the following three systems are involved in a complex interplay:

A) The implementation system which involves areas such as Broca's area in the frontal lobe of the brain (which is responsible for the short-term working memory that is essential for speech production) and Wernicke's area in the temporal lobe (which associates sounds with concepts)

B) The conceptual system which comprises regions of the brain that support conceptual knowledge

C) The mediational system, which, as its name indicates, mediates between the implementational and the conceptual systems.

Wikimedia Commons. Polygon data were generated by Database Center for Life Science(DBCLS)
Location of Wernicke's Area in the human brain.
Source: Wikimedia Commons. Polygon data were generated by Database Center for Life Science(DBCLS)

It has been clear for a while now that Sam thinks almost entirely in concepts, and definitely more so than we as adults seem to do. This is perhaps because her language faculties have not yet developed to the stage where she can just name everything in her head robotically.

When she sees the number "9," she sometimes identifies it as "P." Not having got around to learning lowercase letters yet, she misidentifies "a" as the uppercase letter "P" as well. It is clear that to her mind, "P" is anything with a long stalk and a curve attached to it.

Of course, it could also be that she has memorized the label for P, but just hasn't got round to learning the concepts of symmetry and mirror images yet. If the former explanation is true, though, learning by concepts is a very nice way to learn, and is so much quicker than the alternative of getting the names of objects by heart.

Sam is also able to get abstract concepts. We had been to an Indian restaurant when she was perhaps 18 months old and were extremely surprised when she correctly identified a really weird "modern art" statue as an elephant. She knows that anything with a long trunk and big, flapping ears is probably what people call an elephant. Just the other day, she looked at the following picture and said, "Eyes." 

Aditi Subramaniam
The abstract picture that was identified by my daughter as "eyes."
Source: Aditi Subramaniam

The rate at which children learn new words increases dramatically when they are 18 months to 2 years old. By the time they are 6 years old, children know the meanings of about 13,000 words. One way in which children learn new words so quickly is through a technique called fast-mapping.

Here's how fast-mapping works—Let's say there are five objects in front of Sam at the moment, and she knows the names of four of these. If I say an unfamiliar word and ask her to fetch one of the five objects in front of her, she will assume that the unfamiliar word is the name of the unfamiliar object. She just needs to hear the word once and doesn't need further feedback to learn this. Further understanding of the word and its conceptual meaning takes place over time, but fast-mapping helps explain how children pick up such a huge vocabulary in such a short period of time.

It appears that even before they start associating words with their meanings, infants as young as 10 months old discriminate speech sounds in much the same way as their parents do in their native tongues. For example, Japanese babies at the age of 10 months fail to discriminate between r and l sounds. This indicates, once again, that babies perform acoustic analyses of speech that are much more complex than previously imagined. I enjoy speaking in different accents to make my husband laugh (yes, I'm funny like that), and my daughter instantly knows when I'm talking in an accent different from my own. She laughs and makes hilarious attempts to imitate me.

One could write a book about all the myths of the past that have been debunked by modern neuroscience. One of the more recent theories that teaching a child multiple languages would confuse them has been almost completely discredited, with study after study showing that bilingual children perform better on multiple tasks, most of which involve cognitive control. Further studies need to be done, but bilingualism also seems to be associated with greater levels of empathy—when a child has two parents speaking different languages to her, she has to have the intelligence to respond to each of them in the appropriate language, and this translates to enhanced empathy overall.

Far from being just a means with which to communicate with other people, our language could also shape the way we think. This is an idea known as linguistic determinism. An illustration of this concept lies in a study that examined languages, such as Japanese, where dropping a pronoun such as "I" or "you" from a sentence is common, versus English, where a sentence without a pronoun, such as "went to the temple," would not make sense at all. It was found that people living in countries where "pronoun drop" was common tended to have less individualistic and more collectivistic values (such as greater loyalty towards their employers).

In another study, Mandarin speakers were compared with English speakers for the way they think about time. In Mandarin, time is usually described as a "vertical" concept. Where in English, we would say "the past is behind us," or "the future is ahead," Mandarin speakers tend to refer to the past as being "up" and the future "down." Mandarin speakers were found to be quicker to recognize that March comes before April if they had just been presented with a vertical array of objects. The opposite was true of English speakers. Also, the earlier in life the Mandarin speakers were exposed to a second language (English), the less pronounced was the tendency to think about time vertically. In the words of the scientist who conducted this study, "one's native language could play an important role in shaping habitual thought."

As the field of neuroscience continues to grow and become more interdisciplinary, more insights into this uniquely human faculty are to be expected. I, for one, cannot wait. And in the meantime, my daughter and I have plenty of learning and growing to do together.


Kandel, E. R., Schwartz, J. H. 1., & Jessell, T. M. (2000). Principles of neural science (4th ed.). New York: McGraw-Hill, Health Professions Division.

Fan SP, Liberman Z, Keysar B, Kinzler KD. The Exposure Advantage: Early Exposure to a Multilingual Environment Promotes Effective Communication. Psychol Sci. 2015;26(7):1090–1097. doi:10.1177/0956797615574699

Crivello, C., Kuzyk, O., Rodrigues, M., Friend, M., Zesiger, P. & Poulin-Dubois, D. (2016). The cognitive benefits of growing up bilingual: A longitudinal study. Journal of Experimental Child Psychology. doi: 10.1016/j.jecp.2010.10.009.

Boroditsky, L. (2001). Does language shape thought? Mandarin and English speakers' conceptions of time. Cognitive Psychology, 43(1), 1-22.

Kashima, Emiko & Kashima, Yoshihisa. (1998). Culture and Language The Case of Cultural Dimensions and Personal Pronoun Use. Journal of Cross-Cultural Psychology. 29. 461-486. 10.1177/0022022198293005.