The Sapir-Whorf Hypothesis with a frown

There’s a theory in linguistics called the Sapir-Whorf Hypothesis, which I’m afraid means I have to take a short break in order to imagine a large Klingon with a pair of trope-inspried wire-rimmed glasses perched on his nose looking thoughtful in a lecture theatre.


Image by blueprint2015 from Pixabay

(Yes, I know the spelling is different).

This is appropriate as the theory is to do with whether language affects thought. In the case of the Sapir-Whorf Hypothesis, the obsession seems to be about how speaking different languages produces different thought in different language speakers.

This turns out to be an absolute minefield. Both in terms of actually proving it, but also whether or not you want to.

The trick here is to find things that are the same for everybody.

The physical environment, for example.

That old story about Inuit languages having 700 000 words for snow, a feat unheard of in any other language? Part of the Sapir-Whorf Hypothesis*. The idea here is that if you name it, you notice it better. Nothing to do with just, I dunno, noticing it better than, say, someone who is not surrounded by snow for an appreciable part of the year because it’s more part of your life? But would someone who has not needed to consider the difference between dry powdery snow and really wet sticky snow and their relative merits for building snowmen simply not be able to tell the difference if they were suddenly dumped in that environment? Or would they pick it up actually fairly quickly when all their attempts at snowballs turned to dust after the temperature got down below a certain point?

Can’t imagine where I plucked that example from.

It’s not even particularly true about the surprising number of words for snow in comparison to other languages. It turns out**. I mean, take ‘powdery snow’. Not, it’s not one word. It’s not even a compound, where two words come together to make a new, distinct word. But it is a collocation, which means those words are found together with a greater frequency than chance in English. Even British English speakers, in fact, know about the concept of powdery snow, and have a phrase to describe it. Well, the ones that go ski-ing in the Alps, anyway. Or move to Russia.

Of course, this is one of those things that gets brought up a LOT. Such and such a language doesn’t have a word for… that our language has. Or, such and such a language has a word for… that our language has not. This often seems to be really an excuse to be smug about some aspect of national character. Or trotted out as an example of a national failing. Russian has a word that combines conscience, shame and morality in an untranslatable hodgepodge, as befits a country proud of its deep Russian soul. Russian doesn’t have a word for privacy, which…

Where this becomes the Sapir-Whorf Hypothesis directly, however, is when people start saying that Russian doesn’t have a separate word for science. And speculating that this means that the findings of researchers conducting experiments in chemistry are given the same weight as the findings of scholars engaged in historical investigation. THE VERY IDEA!

At this point it is traditional to bring up the Pirahã Amazonian tribe***, who survived quite well without counting in their lives up to the point they got studied by behavioural scientists. While exploring the way this worked, someone actually measured what happened when, for example, you showed these people groups of animals with slightly different numbers and asked them to say if the groups were equal or not. Which proved difficult. As did copying exact numbers in lines of berries with any accuracy above something like 4. Clearly an effect that a (lack of language) was having on the thoughts of the people (not) experiencing it.

Except, it also turned out to be hard to teach them counting. Despite their having perceived a need (they wanted to make sure they were not being cheated in trading with outsiders), and being given the words, the attempt didn’t translate to success. New language didn’t cause new thinking to happen.

So much for the film Arrival**** and its contention that merely learning an alien language causes you to start experience the universe, nay, the laws of physics, differently.

Let us consider that in another part of the world, people locate themselves not in relation to themselves or some other arbitrary point (in front of me, on the left of the table, behind the church or whatever) but in absolute directional terms. South-west of the book, to the north, my western arm. And so on.

Are these people are actually better at locating themselves, specially-wise, as a result? Yes, yes they are*****. Now that’s the Sapir-Whorf Hypothesis, irrefutably. No?

But what I want to know is, is it really as a result of it being embedded in language or is it embedded in language because knowing exactly where you are at all times is quite important in that particular geographical context, which happens to be northern Australia?

My favourite example of this kind, though, is the one about how Swedish and Finnish factories had a difference in the number of accidents because something something prepositions vs grammatical cases having an effect on the way the factories and the work flow are organised. The Finnish languages fosters a fatal individualism, apparently. The thinking is here that the languages are very different (Finnish is, unusually for a European language, not Indo European), but the countries are neighbours and have similar standards of living and so on. Environmental factors causing difference are lessened.

Apparently. Look, it’s not my theory, OK? I’m not the one lumping Swedes and Finns together as indistinguishable aside from their language.

Of course, this is the point. Rugged individualism, I do rather gather, is a defining Finnish characteristic*******. But is this caused by their language? Or merely facilitated by it? Your answer to this question depends on whether you believe in linguistic determinism or linguistic relativity.

The idea that linguistic peculiarities constrain you to think a particular way (linguistic determinism) is hard to swallow, but perhaps they do force you to contemplate certain aspects of the world more (linguistic relativity).

Although if you don’t believe there is any relationship at all, you are Noam Chomsky or Steven Pinker ********. Who think that language is governed by universal principles, and that universal concepts therefore crop up and are described across all languages, no matter how dissimilar, and that any differences are very minor small beer.

Yet it does seem as though gendered languages have people assigning stereotypically gendered attributes to different objects, depending on the gender they are given in their first language. Even when they are being asked to consider these objects in a second, non gendered language (English)*********.

And then there’s the colour blue. Colour perception has been a particular battle ground for this discussion because, well, (nearly) everyone sees the same range of colour, right? Differences in colour perception MUST be significant.

There’s a study which shows that speakers of Russian, which has two words for blue where English has one, react in a statistically significant different way when shown the two shades of blue compared to English speakers. They did brain scans and everything.**********

There have been counter experiments along those lines too, by the Sapir-Whorf Hypothesis sceptics. Did you know, for example, that while there are differences in colour description, generally speaking there tend to be eleven basic categories, and when there are fewer, they go: black and white; black, white and red; black, white, red, and green or yellow. It’s surprisingly predictable.***********

Not quite sure, if I am honest how that explains away the blue/ blue thing, but then it was a study published earlier. So perhaps it doesn’t.


I admit a distinct preference for thinking that linguistic relativity, or a soft version of the Sapir-Whorf Hypothesis, is a thing. In much the same way I think that all environmental factors have a bearing on how we behave (if you want to be cornered by me at a party and have my explanation of why free will doesn’t exist, which takes in quantum physics and everything, feel very free to invite me).

What I do not understand is why this always seems to be couched in terms of pitting one language against another.************

Surely it would be much easier to determine if language shaped thought if you tried to find out whether people whose backgrounds really are similar, and who therefore speak the same language, can be swayed by into one way of thinking or another by the power of words alone.

Or, to put it another way, advertising.

On the other hand, I have just found out (and this is the reason I am writing this post, in fact) that ‘frown’ means something different in British English, to US English, Canadian English and Australian English.*************

The (right thinking) Brits think it’s all in the eyebrows. The rest of them think it’s a down-turned mouth thing.

Russians and speakers of other European languages of my acquaintance agree with me. Except the Dane, who says that Danish doesn’t have a word for ‘frown’, but does associate different facial expressions with, variously, confusion, skepticism, surprise or disapproval.

I don’t know what this says about the Danes.

I don’t know what this proves in relation the the Sapir-Whorf Hypothesis either, except that we are back to 700 000 words for snow again because I am now absolutely gagging to find out if North Americans (and Australians) have a different emotion as well as a different emphasis on the muscles involved.

Provocatively, a (British) friend suggested sulking. You?

*According to Google.

** Also according to Google.

*** Or at least it is when follow links suggested by Google.

**** Which I haven’t seen.

***** According to Google.

****** Also according to Google. Look, at least I looked a bit further than Wikipedia, OK? Although this example is also in Wikipedia.

******* This is based on my subscription to the Facebook page, Very Finnish Problems.

******** Says Google.

********* I haven’t read this paper either.

********** I did actually read this paper, but it was a while ago and I cannot be bothered to look up the reference, and in any case, it is all over Google.

*********** Google it.

************ This is, in fact, why I have not taken my reading further than Googling.

************* In the Lingthuiasm podcast, episode 20. I gather I am a number of years behind everyone else in linguistics in finding this out, which seems about right.