Daniel Tammet invents his own Siwu ideophone

I loved Daniel Tammet’s second book Embracing The Wide Sky (2009). In his own words, Embracing The Wide Sky is “a personal and scientific exploration of how the brain works and the differences and similarities between savant and non-savant minds”. It surveys work from psychology and linguistics and even indirectly (okay, very indirectly) features my work on Siwu ideophones.

One of the chapters of the book is on language, covering topics from Greenberg’s universals to psycholinguistic research on language acquisition and from aphasia to sound symbolism. The section on sound symbolism, onomatopoeia and phonestesia contains a survey “to test your own intuitive sense for word meanings”. That little survey is proving quite popular online: copies have appeared on dozens of websites (just search for siwu pambalaa and you’ll see what I mean). Here’s how it starts:

Test your own intuitive sense for word meanings from a range of languages with the following multiple-choice questions:

1. Does the adjective ‘pambalaa’ in the Siwu language of Africa describe (a) a round, fat person or (b) an angular, thin person?

That’s right! The first question features “the Siwu language of Africa” — the Ghanaian language that I have been studying —and writing about— as a field linguist since 2007. The question was probably inspired by my online writings, since very little has been published on Siwu apart from my own work. (This post may be the source.)

No pambalaa in Siwu…

Now here’s the funny thing: there is no Siwu word pambalaa. The word just doesn’t exist. But rather than taking Daniel to task for spreading misinformation about Siwu, I want to argue that his ‘misremembering’ illustrates exactly what is so interesting about this kind of words, known in linguistics as ideophones.

First things first though. How do I know that this word is not actually an existing word in Siwu? The answer is that I checked. I have a list of hundreds of ideophones in Siwu, and it isn’t on there. It’s also in none of my publications, on or offline. Additionally, I asked several speakers of Siwu, and they tell me that the word doesn’t exist, though there are some words that are like it (that were indeed on my list, and in my publications): pimbilii, pɔmbɔlɔɔ, and pumbuluu.

Now pimbilii, pɔmbɔlɔɔ, and pumbuluu all have to do with something round and protruding, though varying in magnitude. This kind of pattern is familiar to ideophone aficionado’s. In my thesis, I have called it “relative iconicity”; it has also been referred to as “vowel symbolism” or “magnitude symbolism”.

What about the non-existent pambalaa? Daniel Tammett proposes the meaning “someone round and fat” (as opposed to someone angular and thin). His form is clearly a variation on a theme. And it is a sensible one. I’m pretty sure almost anyone would answer (a) to the quiz question above. Heck, even Siwu speakers say it’s (a) and not (b). That is, they can make sense of the word even though it doesn’t exist — just like us.

…whereby Tammet proves his own point

Daniel’s point in the section on sound-symbolism is that we have an intuitive sense for the meaning of some words. By misremembering this Siwu word, he inadvertently proves his point in a powerful way. For he got the sound-symbolic pattern right. Many of the world’s languages feature words like these, in which the vowel quality is used in a meaningful way (more on that in this recent post on lɛkɛrɛɛ and lukuruu, and an old post on some Japanese ideophones). Cross-linguistically, the vowel a tends to be used for things of greater magnitude; or more precisely, the relation between that vowel and other vowels is often used iconically to map onto a relation between bigger and smaller things. This is why we go for choice (a) in Daniel’s quiz.

Tammet thus has unwittingly demonstrated that sound-symbolism makes sense, and that memory is not just about storing hard and dry facts, but also (perhaps even moreso) about storing relational structures that allow us to creatively reconstruct stuff when needed. To me, this is one of the things that make human language and the human mind so endlessly fascinating.

Rembrandt and Van Gogh

Some interesting features of ideophone systems can be illustrated using this case. For one thing, we can often at least partly make sense of ideophones even if they’re not our own language. This is because they often tap into the general depictive potential of speech sounds and articulatory gestures. But that is not the whole story. Notice how Daniel’s quiz focuses on only one dimension of the ideophone’s meaning: size. Notice, too, that we use only one cue to decide on our answer: vowel quality. But pambalaa and more precisely the existing forms pimbilii and pumbuluu are not just about magnitude; they depict something quite specific, namely the bulging roundness of a belly. I’m pretty sure you couldn’t have predicted that part. Here we are moving beyond generic, possibly universal cues; we are getting into the realm of convention.

Ideophone systems always show this interesting combination of iconicity and convention. This is why they don’t always look the same across languages. We can tell Siwu ideophones from Semai ideophones just like we can tell a Rembrandt from a Van Gogh. Different languages represent slightly different depictive traditions, and this is what gives ideophone systems their language-specific signature.

Siwu speakers can tell whether some form is an existing word or not. Pimbilii is, pambalaa is not. Ideophones tend to be actual words — existing items in a language, not just expressive outcries or spontaneous sound-paintings. But pambalaa illustrates a good way to make new ideophones: use widely shared iconic principles and build on existing words. I occassionally see this happening in my video-recordings of every conversation in Siwu: people may create new ideophones, and when they do so, they use the toolkit for depiction provided by existing ideophones. Like Daniel Tammet’s pambalaa, the cases of ideophone creation I recorded in the video corpus show that creative depictions never occur in a vacuum, but always in the context of a broader linguistic system, using existing depictive practices.

That’s it for today — I’m off to get something to eat. If I tell you that my belly is pimbilii now, you can tell what it should be like in a couple of hours!


  1. Tammet, Daniel. 2007. Born on a Blue Day: Inside the Extraordinary Mind of an Autistic Savant: A Memoir. New York: Free Press.
  2. Tammet, Daniel. 2009. Embracing the Wide Sky: A Tour Across the Horizons of the Mind. New York: Free Press.

1 thought on “Daniel Tammet invents his own Siwu ideophone”

Leave a Reply

Your email address will not be published. Required fields are marked *