9.5: Child Language Acquisition Bilingualism

Child Language Acquisition Bilingualism, from Sarah Harmon

Video Script

I have referenced several times that most children do not grow up in a monolingual society, that they grew up in at least a multilingual society. In many cases, they grow up in a multilingual household where at least two, maybe more, languages are spoken, and that's not even counting the different dialects. I thought it would be important to have a section talking about bilingualism in children.

When we're talking about what is considered to be native or quasi-native bilingualism, we’re describing a situation where a child is multilingual from birth, that would be native, or within the first 2-3 years for a quasi-native situation. Some of you are native bilinguals; you grew up in households where more than one language was spoken, or in your society more than one language was spoken. Many of you are quasi-native bilinguals, meaning you spoke one language in the house, and then at a certain point—preschool or kindergarten is usually the start—things change, and then you have to speak a different language.

Let's talk about that process, a little bit. Historically, in the last 40 years, there have been two main hypotheses with respect to childhood bilingualism. These hypotheses are meant to describe how the child is constructing these grammars in their head and their mental lexicon. Unitary System Hypothesis states that the child initially only constructs one lexicon and one grammar, and then has a word associated with a term from one language or another, but not necessarily both. The evidence for that suggests that young children will only have one term for a given item, so one reference for a given meaning. For example, if a child is a Spanish-English bilingual, and they see this thing (🍼), they're only going to either say ‘bottle’ or ‘botella’; they're not going to say both. There's some evidence for this. The other hypothesis is Separate Systems Hypothesis, which states that the child is constructing one lexicon and one grammar for each language that they're trying to acquire. In this hypothesis, children make the connections along the way. While at times you could argue for either of these hypotheses to explain what happens with native bilingual children, most of the time Separate Systems Hypothesis has more data supporting it.

The real example of this has to do with children when they are producing language, and specifically as they mix their languages. I'll use English and Spanish as the examples. If you take a Spanish-English bilingual child, and you listen to them talk, you’ll hear them mix, their languages. There's a system to it; they never use the incorrect syntax. They never try to use morphology from one language on lexicon from a different language. You would hear a child say, “I want to go to la playa.” (La playa in Spanish means ‘the beach’.) They do not say, “I want to voy a la playa.” They don't try to mix the morphology and syntax from Spanish in English; they will just use lexicon in the appropriate spaces. If they're talking about ‘the brown dog’, they will say ‘the brown dog’; they will not say ‘the dog brown’, the order of the noun and the adjective in English is the opposite of what it is in Spanish. The child won't mix the order of the of the noun phrase, they will stay in whatever structure is set up for the given language and they do this early on. Whether it is Unitary or Separate Systems Hypothesis that best describes native childhood bilingualism, that we're still not sure of. We tend to lean more on Separate Systems Hypothesis, but there is evidence for both.

Let's talk about code-switching. In many immigrant societies, code-switching is considered horrible and something to be avoided. Prescriptively, when a child or an adult comes into a formal education or formal setting., they're told they can't code-switch; they have to say in one language or the other but they cannot mix. Here's the reality: code-switching is the best thing that a language learner can do, whether we're talking about a child or an adult. This is particularly true for child language acquisition in multilingual situations. Code-switching shows the amount that the child has acquired in both languages.

I'll give you a great example. We have a neighbor growing up that, while the parents spoke English, the parents were teaching their children Mandarin because they were Taiwanese. The mother was also a teacher of Cantonese, so she was fluent in Cantonese as well as Mandarin. She was teaching the children Cantonese and predominantly Mandarin as their first language; English was going to come later and they just accepted that. When their daughter was 2-4 years old, she would come toddling down the hill with her family for visit; she would start talking to us in her language. I say ‘her language’ because it was a code-switch of Mandarin and Cantonese, with an occasional English word thrown in there for fun. Of course, my parents and I didn't understand the child, because we speak neither Mandarin or Cantonese, so the mother would have to translate. It was so cute to see how she was putting things together, and how the mother, in particular, would have this question mark on her face, when she could not put the pattern together initially.

Children will subconsciously code-switch; they're doing as part of spitting out their thoughts. There's no regard for audience; they haven't learned the contextual clues about not speaking a language in front of somebody if they don't understand. As a result, they just blurt stuff out. There are combinations, but those combinations tell us how much they have learned about the respective languages. Adults are different; adults will consciously code-switch, meaning that they do it with intention. They will only preferably do this with other bilingual because they know the stigma of speaking a language that the other person you're talking to does not understand. it's really interesting to see how code-switching develops in a bilingual child; around the age of eight is when they start understanding when and why to code-switch. If you think about a child's development, that is about that time they're starting to learn social mores and norms. Politeness starts really factoring in around that time, so they're starting to put a lot of pieces together.

The other important aspect to code-switching is consistent input in both languages. If a child has consistent input in both or all of the languages that they speak, they will continue to code-switch and become more proficient at it. But if the input of one of the languages stops, the code-switching also stops, and they start become more monolingual. It can come back, if they go to learn their heritage language. This gets back to contextual knowledge, and this is where semantics and pragmatics come in; a young child isn't going to have that knowledge, because they haven't developed that mentality, nor have they developed neurologically to that point. As they age, things improve. This is all led to the Contextual Knowledge Hypothesis, which is really becoming more plausible when we're talking about bilingualism both in children and adults. It states that as a child reaches that age of about eight, they start intentionally code-switching. It explains a bit about how adults switch between languages—because they're using the contextual knowledge, the environmental cues to know when they can switch. It also explains how listeners go between languages; if you are an English-Spanish bilingual, and I say [si], depending on the context and which language I’m using, that will lead you to understand that syllable, that lexicon differently. If I’m speaking English, and/or if I’m referring to the ocean or the sea, then if I say [si], you're going to think of body of water. If I'm not, and especially if I’m speaking Spanish, then you're going to think of it as ‘yes’, because that is how you say ‘yes’ in Spanish. (It's also how you say ‘if’, but that's another story.) You're using the contextual clues to understand when someone is code-switching; this is both passive and active in its processing, both in receiving/listening to language and when you're producing language. We use parsing to help us with this; we’ll come back to parsing in neurolinguistics in the next chapter.

It's really important to understand that this is part of what a child acquires if they're in a multilingual environment. It's not just learning the actual lexicon and the morphology and the syntax and everything else. It also ties into the context clues.