On Learning a Language

Linguistics expert William Snyder discusses factors in learning a first, second, or third language.

<p>William Snyder, associate professor and department head of linguistics. Photo by Frank Dahlmeyer</p>
William Snyder, associate professor and head of the linguistics department. Photo by Frank Dahlmeyer

William Snyder is an associate professor and head of UConn’s top-ranked linguistics department in the College of Liberal Arts and Sciences. He says learning a language is not as simple as just hearing it.

Many of us think that children acquire language simply by hearing it. But I suspect it’s more complicated than that. Can you comment on what other factors are important?

Many people believe that children acquire language because their parents teach it to them, but the reality is that children acquire language because they JUST CAN’T HELP IT. Language is a biological characteristic of our species, and acquiring at least one language during childhood is almost inevitable.

I should hasten to add that simply HEARING a language spoken isn’t quite enough. The child needs enough contextual information to be able to guess – accurately – what the adults mean, at least some of the time. So children normally don’t acquire a language PURELY by hearing it spoken on a radio, or even on television (given that people on television rarely SAY the things that the viewer can already SEE).

There’s some good research conducted in The Netherlands on children who grow up watching German cartoons (from TV stations across the border), but otherwise don’t have exposure to German. It turns out that even after years of this type of exposure, the children understand almost nothing.

Often children of parents who have heavy accents don’t acquire the accent. Why is that?

This is one of the remarkable features of child language acquisition. We believe that human biology narrowly constrains what is a POSSIBLE language – in terms of how sounds can be combined to build words, and how words can be combined to build sentences. What children acquire is always one of these “biologically sanctioned” languages.

In contrast, when adults try to learn a new language, the result is usually something that isn’t EXACTLY any of the biologically sanctioned options. For example, the adult may combine elements of the sound patterns – the “phonology” – from his or her native language with elements of the target language, and create something that isn’t quite either.

When the input provided to children [by adult second-language learners] is close, but not identical, to one of the biologically sanctioned options, children “nativize” it – they adopt the closest thing that their biology allows.

A well-studied example of this phenomenon was a highly detailed case-study of a deaf boy born to deaf parents. The parents had both learned American Sign Language (ASL) as adults, and signed with a heavy “foreign accent.” Yet, the child – based only on input from his parents – acquired a language that was far more like “native” ASL than what his parents used.

Is it true that if you learn a language before the age of 11, you will sound like a native speaker?

Yes, that’s about right. The foreign accents that are obvious to non-linguists result when a child has already entered puberty before he or she is first exposed to the given language.

Yet there’s evidence that children who are first exposed to a given language after about age six also exhibit subtle grammatical differences from those who begin using the language by age five or younger. I should emphasize that these differences are very subtle indeed, which is why only linguists ever notice them!

Conversely, is it more difficult to learn a language after, say, age 50?

The evidence I’ve seen doesn’t show any clear “upper bound” of that kind. The main finding is that after puberty has begun, we see a great deal of individual variation in language learning strategies and outcomes. The main generalization is perhaps simply that after puberty, language learning becomes something that you have to work at consciously. Some people are better at it than others, and a study strategy that works brilliantly for one person may fail completely for another. Factors of motivation and dedication play a huge role, and people over 50 will sometimes have even more of these than younger people.

When a speaker is multi-lingual, does he or she have to make a conscious effort to speak one language or the other?

A bilingual adult – or even a bilingual child, by a surprisingly early age – usually has an excellent command of which language to speak when communicating with some other, familiar person. The choice of language can be relatively automatic, or it can be a conscious choice, depending on the situation.

When bilinguals talk with other bilinguals, we usually find “code-switching” – where a speaker will use words from both languages if he or she knows the other person will understand them. Interestingly, code-switching obeys complex, unconscious grammatical rules; bilinguals feel strongly that switching languages is permissible only at specific points in a given sentence. Linguists can fairly accurately predict where these permissible points will fall, for a given pair of languages.

How many languages do you speak?

Linguists hate this question! We hear it quite a bit, and the answer is usually complicated – it depends on what you mean by “speak.”

Like most linguists, I’ve conducted research on a sizable number of different languages, but always with the help of native-speaker consultants. No matter how much effort I put into learning a new language as an adult, my intuitions about what sounds “natural” in that language will probably never be as accurate as the native speaker’s. To do this type of work, I usually need to have a high degree of familiarity with the grammar of the language I’m investigating, but that’s different from carrying on a fluent conversation.

So, to answer your question – as a child I was trilingual in English, Spanish, and French. At present I still use all three of those languages to some extent, though my English is by far the strongest, and the one I use by far the most.

I can usually also read – with a certain amount of difficulty! – linguistics articles written in German, Dutch, Portuguese, or Italian. In my research, I’ve also done work on languages like Albanian, Japanese, Hungarian, Russian, Estonian, Arabic, Indonesian, Swahili, and lots of others – but please don’t expect me to speak any of them fluently!

Is there any difference in how humans acquire a second or third language compared with learning the first?

This is an extremely active area of research right now, and the facts are still coming in. Let me get back to you in a few years!

Your work has focused on language acquisition and you are the editor of a professional journal, Language Acquisition. What are some of the hot topics researchers are looking at now?

This may be a little on the technical side, but one of the topics that I find the most interesting is work relating language acquisition, on the one hand, to cross-linguistic variation in grammar. As linguists get better and better at characterizing the grammars of the world’s languages, we’re simultaneously getting a better understanding of the child’s “hypothesis space” during language acquisition: What options are available to the child, and how can he or she make the right choices, using the kinds of input that are reliably available?

I’m especially interested in work on how the child’s grammatical knowledge develops over time. We’re finding that features that “go together” when they occur in languages of the world are also “acquired together” in the time course of children’s development!