Lost in Translation

One Nation divided by Three Languages

Published in

On Politics

6 min readJan 20, 2020

Suppose a random Sri Lankan met another random Sri Lankan. Would they be able to talk to each other? And understand each other?

The 2012 Census has some interesting data about what languages Sri Lankans speak. In this article, I analyze this data. And also speculate on some of the consequences of these statistics.

What do Sri Lankans Speak?

Here’s a diagram summary of the census data.

[See Appendix for how these numbers were derived from Table A19]

The circles represent people who speak some set of languages. For example, the red circle represents the 56% of the population who speak only Sinhala. The lines between the circles represent the set of common languages that the people across circles can use to communicate.

For example, the people in the Sinhala + Tamil circle (orange), communicate with the Tamil + English (green) in Tamil (yellow line).

[For clarity, I have excluded “tiny circles”. For example, a small proportion of Sri Lankans speak only English, and a few also speak other languages. I have excluded groups that have under 100,000 speakers.]

“Weak” Understanding

There is a ~82% chance that the two random Sri Lankans will speak a common language.

Is 82% good enough? Should we do better? Should it be 100%? The “shortest path” the 100% would be for the 13% who don’t speak Sinhala to learn Sinhala.

But would that mean that Sri Lankans would understand each other? Would that be the end to all tension around language and related ethnic, social, economic, political and religious tensions?

Intuitively, it scarcely seems the case. And were this intuition true, the 82% would be at best a “weak” metric. And at worst meaningless.

Dividing Lines

Our 82% metric measures “languages shared”. But what about “languages not shared”?

What are the consequences of one person speaking a language that the other does not speak?

English

Consider two Sri Lankans. One only speaks Sinhala (red 56%). The other speaks both Sinhala and English (purple 15%).

On the one hand, this is an example of the 82%. The two can “talk to each other”. On the other hand, there is an asymmetry. The English Speaker has access to all the advantages afforded to English Speakers.

For example, English Speakers can speak to other English Speakers. Consume all manner of content produced in English. And perhaps most significant to Sri Lanka, they are members of the “English-Speaking” club. Which, for most of the last two centuries, has dominated economic, political and social power in the country.

Sinhala

Since independence, Sinhala as steadily caught-up with English as the “language of privilege”.

While English still (probably) affords more economic power, Sinhala is top-dog in many broader settings. It is almost impossible to work with a government ministry or department without Sinhala. Most official documents, notices and signage are published exclusively in Sinhala. At some security checkpoints, merely speaking in Sinhala reduces one’s chances of being suspected of being a “terrorist”.

Tied to “English-privilege”, in some circles, Sinhala continues to have adverse effects. For example, in some social circles, “Sinhala-Speaking” is considered the opposite of “English-Speaking”. Rural and unsophisticated. A few even equate Sinhala-Speaking to fascism or worse.

Tamil

Just as English and Sinhala divide, mostly favouring speakers over non-speakers, Tamil also divides, though usually disfavouring speakers. In many cases, the disfavour is compounded by irrational paranoia, often fuelled by politicians and religionists. To many Sinhalese, Tamil is the language of “LTTE Tiger Terrorists”, “Wahabists”, and more recently “ISIS Terrorists”.

Compound Divisions

Multiple dividing lines could compound.

For example, a Sinhala + English (purple 15%) speaker, might discriminate a Tamil Only (yellow 12%) as being (say) both “Non-English-Speaking” and an “LTTE sympathiser”.

“Strong” Understanding

Suppose a random Sri Lankan met another random Sri Lankan. What is the probability that each speaks all the languages the other speaks?

About 36%. Hence, ~64% of the time, there’s some possibility of “language division”.

Happily, there’s a solution. The human brain is more than capable of learning three languages. And often even more.

What should you do if you’re a Sinhala only speaker? Again, learning Tamil gives you the best opportunity of “expanding” your ability to understand other Sri Lankans. Similarly, if you are a Tamil only speaker, then learn Sinhala.

Of course, many Sinhala only and Tamil only speakers will try to learn English as their “next language”, particularly since it has disproportionate advantages in Sri Lanka, and globally. But even these groups, could learn Tamil and Sinhala, respectively, parallel to English.

Concluding Caveats

Much needs to be done besides people learning languages. For example, official documents, notices and signage need to appear in all three languages. Not just one or two. Deeper discriminations based of language, beyond mere documents, need analysis and solutions.

You might also claim that I’ve reduced a complex problem, to a single factor: language. This is a fair critique. In the end, my “language model” is only a model. And all models are wrong. But some are useful. Hopefully, this is.

Finally, you might claim that I made unequal things sound equal. For example, that I imply that the English, Sinhala and Tamil “divides” are equally problematic. Again, fair critique. Take my “sounds” as more qualitative, than quantitative.

On the one hand, acquiring a “strong” understanding of each other’s languages won’t necessarily lead to strong understanding as people. Many Sri Lankans speak many languages. But in their thesauri, “Tamil” is a synonym for “Terrorist”, and “Sinhala” is a synonym for “Nazi” (for too many). As long as these deeper prejudices exist, learning languages alone won’t foster real understanding.

On the other hand, learning others’ languages will be a start. And will have a multitude of other benefits. Even beyond understanding.

Appendix: Deriving “What do Sri Lankans Speak?” from Table A19 (added 2022–11–03)

You will note that the percentages in the Total row of Table A19, are different from those in this diagram.

For example,

In Table A19,
“Sinhala & Tamil” = 16.1% ,In “What do Sri Lankans Speak?”,
“Sinhala + Tamil” = 9%

This is because the former refers to people who speak “Sinhala and Tamil, and possibly English”, while the latter refer to people who speak “Sinhala and Tamil, but not English”.

Hence,

“Sinhala + Tamil” 
= “Sinhala & Tamil” — “All 3 Languages” 
= 16.1% — 7.1% = 9%

Similarly,

“Sinhala + English” 
= 22.5% — 7.1% 
= 15.4%“Tamil + English” 
= 8.4% — 7.1% 
= 1.3%

(Note, the chart also rounds percentages to 1% for clarity.)

Finally, for each ethnic group, we compute the proportion of people mono-lingual by subtracting population of multi-lingual speaks from the total population.

We make the assumption that if you are Sinhala and mono-lingual, you speak Sinhala only. For Tamils, Moors, Malays and Burghers we assume that you speak Tamil only. This assumption might not be completely accurate.