Saturday 23 March 2013

Children speak child-speak

            Me: Children speak child-speak, you know?
            Possibly you: Duh!?
            Possibly you: Of course they do!
            Possibly you: Hellooo (singsong), they’re *children*... (ditto)

If your reaction to the title of this post matches one of these (possible) responses to it, I’m with you all the way. But do read on, by all means. I’m not going to tell you that children speak child-speak, I’m going to tell you about those of us who appear to believe that children don’t. And I won’t be telling you only about your average, averagely misinformed layperson, either.

The first time I realised that something must be very wrong with academic treatments of child multilingualism was back in the 1980s, when the literature about “translation equivalents” was making headlines. Briefly, proponents of multilingual equivalence had it that children being raised multilingually should either provide evidence that they had acquired words for the same things in all of their languages, or forfeit their right to be (called) multilingual. Being multilingual was then viewed as being what I call multi-monolingual: multilingual children are just like monolingual children, only several times over. Implicitly, the claim was that multilingual adults must also be multi-monolingual, on the understanding that children don’t typically remain children all their lives.

If you’re shaking your head in benevolent disbelief at how this kind of nonsense could ever have made headlines, then don’t. Not just because nonsense about multilingualism keeps making headlines, but principally because this claim appears to have resurfaced lately: one announcement of a meeting on multilingualism, dating from only a few weeks ago, states that “early bilingualism often result[s] in perfect parallel proficiency”. Note: perfect, parallel, proficiency. I doubt it that the convenors’ choice of words was a simple tribute to alliteration. But I do wonder what these words mean, in isolation and in collocation, and I would love to know about the evidence supporting this statement.

My first publication explaining that perfect parallel translation equivalence proficiency can’t make multilingual sense also explained why. The data came from my own children, then only two of them, and then users of only Portuguese and Swedish. At the one-word stage, the children started using the “wrong” language with us, parents. Or rather, they started using the same language with both parents, in words like, for example, Swedish där (‘there’) or Portuguese (‘gimme’), and I wanted to know why.

On preliminary inspection, the observation was that “translation equivalents” of these words in their other language were way beyond babies’ articulatory abilities, e.g. Portuguese ali or Swedish får jag. I concluded that the children’s “vocal tracts were, at the time, not mature enough to pronounce the respective translations in each language, which are phonologically more elaborate.” Child speech doesn’t reflect adult articulatory sophistication.

On closer inspection, the children turned out to use different intonations with each of the words that they so “mixed”. In fact, they used different prosodies altogether with each parent, and they did so in their babbling. I understood their use of prosody as a means of differentiating between the otherwise very similar syllables/words that they were able to pronounce, like Swedish där or Portuguese . The children were signalling, through intonation, rhythm, and stress, that they were speaking different languages, and that they were doing so with the “right” parent. So I wondered: should we also look for “equivalents” of intonation across languages, as evidence of “perfect parallel proficiency”? Where can we find such equivalents? And if we can’t find them (we can’t), does this mean that multilingual children who do this (my children are not the only ones) are not multilingual after all, and are instead “confused”, because they’re using the “wrong/right” words with the “right/wrong” parent?

We may want to rethink what we mean when we talk about using “words”, and we may want to rethink what being multilingual means. It’s not the children’s fault that adult expectations shape the way we see and hear our little ones. My children were using both of their languages, and they were differentiating their use of both. Not by replicating adult uses of these languages, not in ways that adults believe languages “should” be used, and not in baby ways of doing adult things: they were doing baby things the baby way. If you’re keen to read a detailed account of (all three) children’s strategies to sort out their (three) languages, have a look in my book Three is a Crowd?

In contrast to many adults, small children do and say things that make sense. What they do and say teaches us about how we all learn, including how we learn to be multilingual, and teaches us how to give evidence of what we’ve learned: by using the means that are available to us. If, that is, we choose to *see* what’s going on, instead of attempting to fence in facts within the theory of the day.

OK, so untrained vocal tracts explain child mixes, and productions which don’t match uses of language in the children’s surroundings. Could this also be the case for adult language learners? My next post has something to say about this.
Cruz-Ferreira, M. (1990). Karin and Sofia in Bilingual-Land In J. Leather & A. James (Eds.), New Sounds 90. Proceedings of the 1990 Amsterdam Symposium on the Acquisition of Second Language Speech (pp. 248-254). Amsterdam: University of Amsterdam Press.

© MCF 2013

Next post: Shibboleths & Co. Saturday 6th April 2013.


  1. It is certainly true that the capacity to articulate certain sounds (e.g. the rolled 'R' in Spanish or Scottish English) develops later in childhood, and there is a face validity in your assertion that bilingual children would choose to express themselves in the language that offers the phonetically simpler option.
    But I'm not sure I believe it. My impression from observing our boy (En/Fr parents in Fr environment) is that he preferred to express himself in French because that worked *everywhere,* while English only worked at home. So early on, his language was grammatically French with many English words embedded, notably for food items, which he'd learned at home.
    He did not learn to express himself in English until confronted with monoglot English speakers on a trip to the UK, whereupon his English expression leapt forward dramatically.
    The less that I draw from this is that, even at a very early age, the child's expression is driven by a desire for simplicity and a pragmatic need to be understood. Perhaps children choose the more easily pronounced words in the language the parent does not speak to try to help the adults around them who don't speak that language to learn it?

  2. Anonymous: Thank you for these observations! Please note that what I report here concerns the one-word stage, as I wrote. So your child’s “early on”, with embedding of “many” words of one language into another, is rather later than my children’s “early on”.

    And please believe it :-) You can read a full discussion of this issue in the book that I mention,Three is a Crowd?

    The data that I discuss in this post do not support the additional/alternative interpretation that you propose, of children’s deliberate choice of more easily pronounced so-called translation equivalents to teach a language to adults. The children had evidence that both parents spoke both home languages. Data from your family might suggest otherwise?

    Your remark on the dramatic effect of input on language development matches my children’s linguistic progress, also reported in the same book. You may have seen these related posts:

    The trick is in the input

    Languages lost and languages regained

    Come back any time


  3. From what I understand of children being taught two languages, I feel like you are right. Like you said, children do things that make sense, and they will speak what language makes sense to them at the time. I really like the points you support in here as well. Great article!

  4. I was brought up in an English environment by an English father and a German mother, both of whom speak each other's language pretty fluently. I spoke English with my father and German with my mother. My guess is that early on in my life I thought everyone spoke differently and I probably considered that normal.

  5. Thank you, Chris and Robin! I’ll have to agree with you both: what makes sense to us when we first start making sense of our surroundings becomes our “normal”, not least the uses of language that we observe. The next step, the trickier one, is to realise that our normal is not necessarily everyone else’s...



Note: only a member of this blog may post a comment.


Related Posts Plugin for WordPress, Blogger...