Thinking in tongues

One of the popular questions addressed to multilinguals is “In which language do you think?”

Image © Clipart from clipartheaven.com

Like other favoured questions, this one also assumes a singular answer, and therefore that it makes sense to ask such questions of multilinguals but not of monolinguals. Monolinguals are in turn assumed to think in one language because they have a single language – something whose rationale some of us might wish to question.

The other assumption is that we all think in some language. This is intriguing, in that a cursory look at the literature shows all but clarity in thinking (or talking) about thinking. We think individually, of course, but if we do think in tongues, then maybe our findings about thinking will vary depending on the language we’re using to think about these things. If we don’t think in tongues, what do we think in, and how do we convert our thoughts into some language that may make our findings known to fellow thinkers? And if we can’t convey our thinking in any language, is there a problem with the thinking or with the languages? And so on.

You can check out some of the players in this controversy by searching for “linguistic relativity”, or “Sapir-Whorf”, the surnames of the two linguists who most recently became associated with it. Their names usually come up in this order, though Whorf got there first and foremost, and in the context Sapir-Whorf Hypothesis, meaning that any claims must be proved or disproved through empirical evidence, which is what hypotheses are there for. Facts are, however, that we’re still wondering what to think about all this: like multilingualism and other topics where we’re faced with too few data and too many theories, Whorfianism has also had its fashionable woes and joys. It went from a must-have approach to thinking about (and in) languages in the early 20^th century, through academic tabooing, to desirable revival from the year 2000. Alison Gopnik was one forerunner of Whorfian rehabilitation: the title of her book chapter ‘Theories, language, and culture: Whorf without wincing’ says it all.

Against this rather disconcerting background, I would like to offer a few thoughts (no pun intended) on a firm correlation that we nevertheless feel free to establish between thought and languages. I’ll deal with other assumed correlations in my next post. Please bear in mind that my thoughts might strike you (and me) differently, were we to use a different common language from the one I’m typing this in.

Parenting guides insist on the need for caregivers to develop one “good” language in their multilingual children. This is the language which usually turns out to be the one (also singular, yes) which the family’s community acknowledges as mainstream, or school language, or as having an enviable tradition in print, or all of the above. The reasoning here seems to follow two convictions. One, that there are languages which are (more) suited to thinking in, and which should therefore be chosen, top-down, for multilinguals. And the other, that we don’t so much think in languages because we have those languages, but that we need to develop a language in order to be able to think in it – or, perhaps, in order to be able to think at all. I have quite a few things to say about this presumed “language of higher thought” in my book Multilinguals are ...?. For instance, that if you are doomed to thinking higher things in only one of your languages, then you must also be doomed to having only lower thoughts in your other languages.

We can of course think about anything regardless of language, as thinking users of the 6,000 to 7,000 languages that (we think) we’ve identified worldwide make clear. We can do with any of our languages whatever we need (or want) to do with them, provided we do it, bottom-up. I can explain what I mean: one question which never fails to draw peals of laughter from my Singaporean students is whether we can discuss nuclear physics in Hokkien. In Singapore, Hokkien is a dialect (and “dialect” is a derogatory term), fit for the army and rough goings-on. The next question I ask of my students is why can’t we do with Singapore Hokkien (or with Singlish, for that matter) what’s being done with, say, Kreyòl in Haiti. Why can’t children, like all speakers of the languages that people do speak, “build solid foundations in their own language”? This Linguist site has more information on this Haitian project.

To me, attempting to assign one language to (higher) thought makes as much sense as attempting to extract other “privileged” single languages from within a multilingual, their “first, main, best” or their “dominant” one. It is clear that multilinguals have different languages for different purposes, but I don’t see how this must mean that multilinguals are stuck with the purposes for which they use their languages. Languages are as flexible as we make them, because languages have no claims to superiority over other languages: people have such claims over other people. If we do think in languages, the issue isn’t “in which language do we think”, but in which languages can we think.

Next time, I’ll talk some more about thinking, namely, about another correlation between language and thought that we seem to take for granted: if we speak funny, does that mean we think funny, too?


Gopnik, A. (2001). Theories, language, and culture: Whorf without wincing. In M. Bowerman & S. C. Levinson (Eds.), Language acquisition and conceptual development (pp. 45-69). Cambridge, UK/ New York: Cambridge University Press. DOI: 10.1017/CBO9780511620669.004

© MCF 2013

Next post: You speak so, therefore you think so. Saturday 15^th June 2013.