On Language and Communication

In the context of the Cairo speech, I had asked the question whether President Obama ‘got’ his audience right. The question was prompted by a conviction that speakers of different languages had subtle differences in how they saw and understood the world.

It is quite a coincidence that just a week later I found a fascinating study that has empirically tested this hypothesis.

Here are some (unconnected) excerpts from the article describing the study:

Do the languages we speak shape the way we see the world, the way we think, and the way we live our lives? Do people who speak different languages think differently simply because they speak different languages?

These questions touch on nearly all of the major controversies in the study of mind. They have engaged scores of philosophers, anthropologists, linguists, and psychologists, and they have important implications for politics, law, and religion. Yet despite nearly constant attention and debate, very little empirical work was done on these questions until recently. For a long time, the idea that language might shape thought was considered at best untestable and more often simply wrong. Research in my labs at Stanford University and at MIT has helped reopen this question. We have collected data around the world: from China, Greece, Chile, Indonesia, Russia, and Aboriginal Australia. What we have learned is that people who speak different languages do indeed think differently and that even flukes of grammar can profoundly affect how we see the world.

The fact that even quirks of grammar, such as grammatical gender, can affect our thinking is profound. Such quirks are pervasive in language; gender, for example, applies to all nouns, which means that it is affecting how people think about anything that can be designated by a noun. That’s a lot of stuff!

I have described how languages shape the way we think about space, time, colors, and objects. Other studies have found effects of language on how people construe events, reason about causality, keep track of number, understand material substance, perceive and experience emotion, reason about other people’s minds, choose to take risks, and even in the way they choose professions and spouses. Taken together, these results show that linguistic processes are pervasive in most fundamental domains of thought, unconsciously shaping us from the nuts and bolts of cognition and perception to our loftiest abstract notions and major life decisions. Language is central to our experience of being human, and the languages we speak profoundly shape the way we think, the way we see the world, the way we live our lives.

Let us go back to what I had written in the post on the Obama speech:

Knowing one’s audience is an issue that has significance beyond the Cairo speech and beyond politics – it is at the heart of all communications, all attempts at persuasion, and all efforts at marketing.

Some socioeconomic situations make it easier to realize the nature of this phenomenon. Take a country like Pakistan with a colonial history where there are two broad groups in society – the English-educated, English-speaking elite and the others who communicate in languages other than English. The ways of persuading these two groups to a point of view require very different approaches. Again, if the constraint was to generalize with a one-word characterization, it could be argued that the first group is swayed more by deduction, the second more by precedent.

This might be difficult to accept as a first reaction. Try an experiment. Take an English language op-ed that you find particularly convincing. Translate it into a local language and give it to a local language speaker not familiar with English. Ask him or her the degree to which the message was found to be convincing. It is not that the local language speaker cannot be convinced or is impervious to logic; it is just that he or she has to be convinced in a different way.

This would be a great experiment for a journalism class.

In the post I had not mentioned how I came to feel that language had an impact on our thought process but in the context of the study mentioned above it is worthwhile doing so.

Many years back I happened to meet an editor of a local language newspaper at a time when he was planning an English language edition. My question was how he intended to deal with the scarcity of good opinion writers in English. He answered it would not be a problem because he would have the local language op-eds translated into English.

We met some years later and he told me the experiment was a failure. An op-ed translated literally from the local language to English seemed to lack something critical. It had to be fixed by someone who thought in English.

So I had empirical evidence of a limited nature.  Subsequently my own experiences of translating content from English into local languages for study groups strengthened the belief that it was necessary to understand how a person from a different language group thought before one could succeed in convincing him or her.

It is very exciting to see a systematic study attempting to deepen our knowledge of this complex phenomenon. It also tells us how we can go about understanding complex issues better by testing our hypotheses after we have argued and speculated about them.

The article by Professor Lera Boroditsky is also archived in The Best From Elsewhere section of the blog.


  • Arun Pillai
    Posted at 00:33h, 15 June Reply

    I think one has to be a bit careful whenever one puts forward a claim of (linguistic or any other) relativism because neither extreme – all-out objectivism or all-out relativism – can be right.

    I read Boroditsky’s article and I was a little troubled by the large conclusions (e.g. language shapes how people choose their spouses) she appears to draw from relatively simple experiments. An American anthropologist in the early part of the twentieth century wanted to get rid of native American languages because he deemed them inferior.

    The example of translation you give also appears to have more to do with the different concerns that different groups might have, not so much with the particular language they speak. A thought experiment is to ask what would change if all social conditions were kept the same but the language was changed from A to B (and vice versa). It seems to me that all that is being claimed is that different social groups have different concerns and practices.

    Thinking about relativism is rather tricky. Some of your readers may enjoy reading about the Sokal hoax at http://www.physics.nyu.edu/sokal/.

  • Vinod
    Posted at 02:38h, 15 June Reply

    I saw a talk by the anthropologist Daniel Everett who presented on the South American tribe – the Pirahas who spoke the language Piraha. Their language was influenced by the worldview they had. Their words for counting stopped having unique words for anything greater than 3. Additionally, being a group that lived in the present their language did not have the grammar for future tenses. They did not need it!! Same goes for the past tense. They were only concerned with what happened in the immediate past and the immediate future. Their notion of time was wound tightly around the present. Here’s more about it –


  • Vinod
    Posted at 02:50h, 15 June Reply

    The Chinese have a lot of respect for the position that a person holds in the network of relationships. They have separate words for mother’s brother, father’s brother, mother’s mother, father’s mother, male cousin from father’s side, male cousing from mother’s side, same for female cousins and different words for each sibling depending on where you stand in the line of entering the world.

    Compared to that is the English language which has few such distinctions. ‘You’ goes regardless of whether you are one year older, younger or many years senior. ‘Cousin’ works regardless of whether you are from mother’s side or father’s side. What does that say about the culture from which the language developed? Is it that the hierarchy in the network of relationships is not significant?

    What about the changes in the understanding of the Quran that came about when the Quran spread outside of Arabia and into Iran? How did phrases such as ‘Allah rose above the throne’ start to get new dimensions to it? Why were they not originally debated by the Arabs themselves? I think that the Iranians who thought through a different mode of language that had long dabbled in philosophy were bound to see amorphous and multi-layered meanings to verses which didn’t mean much for an Arab whose language developed to address the basic suvival needs of a desert environment.

    There is a lot of potential to research in this topic.

  • Vinod
    Posted at 02:55h, 15 June Reply

    I think there is a lot of values illustrated in the oaths of a language. Oaths are typically taken on something important and sacred. A study of the oaths will reveal what the people in that culture in which the language developed thought were sacred.

  • Vinod
    Posted at 02:59h, 15 June Reply

    In Turkish you’d have to include in the verb how you acquired this information: if you had witnessed this unlikely event with your own two eyes, you’d use one verb form, but if you had simply read or heard about it, or inferred it from something Bush said, you’d use a different verb form.

    I got the above from the article you referenced. Interestingly this is also a characteristic of the Pirahas. According to Daniel Everett, the language is so structured because the Pirahas were a very evidenced based community. The language had to indicate the quality of the evidence. In other words, they were very empirical in approach. Quite remarkably, the Pirahas are also atheistic!!

  • Vinod
    Posted at 03:23h, 15 June Reply

    Another example I can think of is the different pronouns and verb suffixes in Hindi for men and women and the absence of this info in English verbs. When I talk Hindi I am forced to be aware that I am addressing a female or a male. I don’t need that awareness when talking English. Does this affect gender relations in some way?

  • Vinod
    Posted at 06:45h, 15 June Reply

    Here’s another example that I’ve encountered. I’ve been wanting to let my mother know that I love her. I talk to my mother only in tamil. And in the language there is no way of saying ‘I love you’ in an unromantic context. Is it because the relationship between parent and child is viewed dominantly as one of obedience and authority, regardless of the affections one may feel? On enquiry with my Chinese colleague, I got to know that Mandarin too has the same limitation. I have to resort to English to say that to my mom. And that simply is not an option. ‘I love you’ cannot be viewed unromantically in Tamil. Fortunately for me, in explaining this difficulty to my mother, she got the message. 🙂

    Yet another instance is that of ending a phone call with my parents. There is no equivalent of ‘Bye’ in Tamil. The Tamil phrases are ‘Let’s meet’ or ‘take care’ and they feel very inappropriate to end a call. They lack the feature of finality. I have to resort to saying it in English and I am still trying to get used to that with my mother. Tamil did not develop to adapt with technology. The language is more suited to face to face conversation.

  • Arun Pillai
    Posted at 16:17h, 15 June Reply


    The fact that you didn’t have the words in Tamil does not mean that you couldn’t think the thought. I am reasonably certain that this must have been true even for monolingual Tamil speakers in earlier times. One can say his father’s sister’s son in English even though there is no single word for that kind of cousin. Many linguists are skeptical about Everett’s findings. The connection between the words in a language and what can be thought by users of that language is difficult to make. All I am saying is that arguments for linguistic and other forms of relativism should not be made lightly.

    • SouthAsian
      Posted at 16:32h, 15 June

      Arun, I agree that non-existence of a word doesn’t mean that one couldn’t think the thought. The point might be that the thinking process might become different, i.e., labored, roundabout or artificial, which might have a bearing on actions. So the real question might be why do words for some things exist in one language and not in another (leaving aside obvious things, like snow, that themselves vary across societies).

      After all the time I wasted trying to understand Foucault, I came to the (self-serving) conclusion that perhaps it was not that I was that obtuse; it could well be that the French understood the world in somewhat different ways.

    • SouthAsian
      Posted at 19:16h, 15 June

      Arun, Some more thinking makes me focus on the following:
      1. Language has to be derivative phenomenon. It could not have been prior to but must have emerged to articulate the ways of understanding and social relations of a group. So, if the former varied across groups, the languages must have reflected these differences.
      2. In turn, the languages might have reinforced the particular ways of thinking and specific social relationships.
      3. Languages must have evolved to reflect new phenomena, external influences, and interactions with other groups but because of (2) this evolution would have been slow.
      4. The adaptation of vocabulary to new things and objects would have been much faster than changes in the ways of understanding the world. So Indians assimilated motor car, cell phone, etc. but their ways of thinking (which would have also varied across India) need not have adapted at the same pace. This goes back to our discussion of the meaning of modernity in India.
      5. An interesting thought experiment would be to imagine an India that was never exposed to English. Would we have been thinking today the same way we do now?
      6. So, a new language is like a tool which gives us a new vocabulary and also some ability to think in ways we might not have done earlier.
      7. This might also mean that if we learn a new language we start getting an insight into how the people of foreign language group think about the world.
      8. And this might imply that a mechanical translation (say of Foucault) would miss a lot of nuances because of these non-overlapping worldviews.

      I now recall reading a fascinating book (Alain Peyrefitte, The collision of two civilizations: the British expedition to China in 1792-4) that describes the issues that arise when two thought systems interact for the first time and how words are continually misinterpreted when filtered through different worldviews.

  • Vinod
    Posted at 16:31h, 15 June Reply

    As you suggest, I could perhaps think of the thought. But what would the dominant character of my relationship with my parents be like if I only knew Tamil – I think it would be one of obedience regardless of whether there is enough affection to go along in the relationship. Often in my community sons find themselves unable to stand up for what is right against their parents. Parents have an irresistable controlling hold on the minds of sons. This may even take a toll on the son’s relationship with his wife. Is it a valid hypothesis to make that the Tamil language, to some extent, moulds this undying obedience to parents in sons?

  • Vinod
    Posted at 00:32h, 16 June Reply

    Languages must have evolved to reflect new phenomena, external influences, and interactions with other groups but because of (2) this evolution would have been slow

    SA, in corroboration of your point, I think the fact that muslims refer to non-muslims with the epithet ‘kafir’, which does not have a positive or even a neutral connotation, is going to prevent true respect of non-muslims in the mind of muslims and thereby affect relations with non-muslims.

    • SouthAsian
      Posted at 03:17h, 16 June

      Vinod, I agree with you. Two thoughts come to mind. First, how does one group initially refer to another; and, second, how does this usage evolve over time?

      I looked up the Wikipedia entry on ‘barbarian‘ and it makes for very interesting reading. Initially every group tended to refer to the ‘Other’ in derogatory terms – thus the Greeks called them barbarians and Indians called them Mlechch. Reportedly the Aztec used the term Chichimeca which meant ‘dog people.’

      However, the Greek no longer label non-Greeks as barbarians and I doubt the Aztecs would be employing Chichimeca. So for Muslims to still call non-Muslims as Unbelievers and for some Hindus to call other Hindus as Untouchables suggests a case of arrested development. If one continues to use these terms one cannot get away from prejudice and arrive at a state of accepting the other as a social equal.

      In the Wikipedia entry I was really impressed with the reported position of Plato that “dividing the world into Greeks and non-Greeks told one nothing about the second group.” It seems there were always some people who could think things through. No surprise that we still talk of Plato today.

  • Arun Pillai
    Posted at 17:05h, 16 June Reply

    This has been a fascinating exchange. I do agree with much of what has been said above. I was only trying to voice two concerns. One is that arguments of the sort above for one or other kinds of relativism should not be black or white but nuanced. So I can agree that language makes it difficult to think certain thoughts but not impossible to think those thoughts. It is always a moot point how strongly one factor x (e.g. language or culture) influences another factor y (e.g. thought or reality). The second concern is that (linguistic) relativism should not be allowed to completely close off one’s access to the world. Otherwise it becomes difficult to explain how worldviews, languages, cultures etc. change. This second point relates, for example, to what I see as a potential problem for Balgangadharan’s view that colonialism completely wiped out Indians’ access to reality. These effects were partial and not as dramatic as he makes them out to be. That is all. This may sound like I am splitting hairs but I think the caution I am expressing is important; otherwise one ends up with absurd views.

  • Arun Pillai
    Posted at 18:36h, 16 June Reply

    A very interesting – and, on reflection, obvious – example of how certain types of language can enable thought comes to mind. Think of mathematics – those parts of it that have become languages like set theory or linear algebra or the differential and integral calculus. Without these languages, it becomes very difficult to think certain thoughts (e.g. how can one think of something like De Morgan’s laws? how can we think of high dimensional spaces? how can we think about computing complex rates of change and areas?). Certain complex ideas become completely unthinkable without the relevant language (e.g. the idea of an open set is unthinkable without the idea of a set). Likewise with the sciences: Einstein’s relativity becomes unthinkable without Newton’s laws of motion or molecular biology becomes unthinkable without physics or quantitative finance becomes unthinkable without economics.

    What is the relation between such specialized languages and the natural languages we speak as human beings? The key notion is that of expressive power – the range of meanings that can be expressed by any language, whether specialized or not. The question then becomes: what is the expressive power of different natural languages? Here, too, I now agree that certain types of ideas are probably just not expressible by certain natural languages (e.g. certain types of feelings that one cannot point to in a shared physical world) and also perhaps cannot be thought without considerable effort.

    The more one thinks about this issue the more complex it appears.

  • Arun Pillai
    Posted at 19:18h, 16 June Reply

    The question that needs to be posed carefully is: what kinds of things can be thought that lie outside the expressive power of a natural language? Certainly, everything that lies within the expressive power of a language can be thought even if with more or less difficulty depending on the specific resources (e.g. vocabulary) of the language. So, the question is what can we think that lies outside the expressive power of our language? If the answer is nothing, then we get full-blown linguistic relativism. But this is not convincing because without being able to think outside the language, the expressive power of a language would have to remain constant which cannot be right. The expressive power of English is certainly different today from earlier times.

    • SouthAsian
      Posted at 02:35h, 17 June

      Arun, I am reaching the limit of my capability but let me pursue this for a bit. I like the extension to specialized languages. I can imagine that when some new instrumental need arises, individuals struggle to address it with the tools they have and when the latter prove inadequate to think it through they painstakingly invent new ones – specialized languages in this case.

      The interesting thing about a specialized language like mathematics is that with training individuals from any linguistic group can understand and manipulate it. You can sense the relief in Pervez Hoodbhoy’s conclusion to his review of Mathematics in India: “But a quadratic equation solved by whoever, by whatever means and for whatever purpose must give exactly the same solutions. Ultimately, mathematics is mathematics.” Perhaps, this is because compared to natural languages, specialized ones are much simpler with rigorously defined terms and operations. The question that does come to mind is whether individuals from some linguistic groups find it relatively easier to master the language of mathematics than others?

      At the same time, Pervez does mention that “the early development of Indian maths was influenced by religion.” If this was the case then is it reasonable to speculate that because Indian religion had embedded in it such a strong sense of cyclicality, it made it easier to think of the notion of zero? In contrast, did the linear concepts of the Europe prolong the struggle with the Roman numerals? This would be a plausible link of culture and language to thinking in particular and different ways.

      [Pervez notes that the author of the book “is not ready to certify that the concept of zero was an Indian invention; it could well have been conveyed by Chinese Buddhist pilgrims.” But the Chinese also had a tradition of complex cyclical reckonings.]

      However, I wish to go back to the simpler question posed by Vinod that pertains not to instrumental needs of measurement etc., but to expressions of emotions. As Vinod mentioned, many South Asian languages do not provide a natural way to communicate feelings of love for parents whereas respect and gratitude can be very easily conveyed. I am sure there are many who regret that they cannot communicate their feelings of love to their parents because the language interjects a kind of awkwardness in the expression. Are we aware of this only because we now know English and can think in a different way? If we had never known English, would we have even realized that there was a feeling we could not express? Did English always have the equivalent of ‘hey mom, I love you?’ Or is this American and not English?

      This is such an obvious human emotion that it is hard to understand why South Asian languages didn’t come up with a mechanism to express it naturally. Could we not have evolved a ritual for such needs? Is it possible that we have drifted away and are unaware of something that might exist for such a purpose?

  • Vinod
    Posted at 01:30h, 17 June Reply

    Arun, your cautionary point is well made and taken. Your last post also has a valid point. We are capable of thinking beyond the expressive power of language. But in my view, the likelihood of that happening is far greater only if there is an exposure to a diversity of views that emanate from a diversity of backgrounds. If everybody around us speak one kind of language it conditions the speakers to think in a certain mode. Breaking out of this conditioning is not easy, although possible.

    • SouthAsian
      Posted at 02:53h, 17 June

      Arun, Apropos of Vinod’s last comment, can we link this to Richard Florida’s recent research that shows very strong links between diversity and innovation? Florida is talking about technology but there is no reason to exclude linguistic and cultural innovation from the discourse. The ways of thinking have to be stimulated before any kind of innovation can occur. Whenever I read descriptions of pre-partition Lahore as one of the most culturally vibrant cities in india, I think of this phenomenon. It is such a wasteland now, over-run with people who are all alike and think increasingly alike in the same fundamentalist terms. Those who disagree do so at considerable risk.

      Vinod, I know Richard Florida (the guru of Creative Cities) was engaged as a consultant by the Government of Singapore worried about losing out in the competition for globally mobile creative workers. There was reluctant permission to allow a Bohemian Village in Singapore where gay bars would be tolerated. Any truth in this?

  • Vinod
    Posted at 10:19h, 17 June Reply

    SA, I’m sorry I have not heard of that issue. I cannot give any more info than what can be found through google.

    • SouthAsian
      Posted at 14:21h, 17 June

      Arun/Vinod, Now that we have started discussing mathematics as a specialized language and its relationship to natural languages, could we revisit a very old stereotype – that South Indians have a greater facility in mathematics than North Indians. This had never made much sense earlier (much anecdotal evidence notwithstanding) but perhaps we have a plausible line of thinking to pursue now. Any thoughts?

    • Vinod
      Posted at 02:02h, 18 June

      SA, I’m inclined to hold onto Arun’s caution and resist speculative hypothesising.

      This is the first time I’m hearing about that stereotype.

  • Vinod
    Posted at 03:39h, 09 August Reply

    SA, you may be interested to know that research has shown that there is a direct correlation between the extent of the presence of genes ASPM and Microcephalin in a population, genes that determine brain size, and whether the population speaks a tonal or non-tonal language. Where these genes are common the language spoken is non-tonal and vice versa.

    Added by South Asian: This would be of interest to Aakar Patel. Research also indicates that a larger proportion of individuals speaking tonal languages is likely to have perfect pitch.

Post A Comment