Only 4% of languages are used online | Hacker News Reader

Only 4% of languages are used online

priceonomics.com

67 pointstaofu55 comments

And what percentage of people in the world uses one of the languages in that 4% as their primary language? I'm guessing it's a really large percentage, probably a majority of the world's population. It's not really useful to look at the percent of languages when they vary in the number of speakers by so much (and there's a long tail of languages with a vanishingly small number of speakers).

Indeed. To get a rough idea I added up the first table in https://en.wikipedia.org/wiki/List_of_languages_by_number_of... , it seems that about 85% of the world population are native speakers of one of the top-100 (top 1.6%) languages.

It's not clear from the Wikipedia list if the language categories are mutually exclusive. For example, it seems quite likely that nearly everyone who speaks the Wu, Jin, and Min Nian Chinese family languages also speaks Mandarin.

I'm not even sure what it means to say that Min Nian is the native language for anyone in Taiwan. Sure, lots of people learn it at home but anyone under 60 is going to be just as fluent in Mandarin.

The first language I learned is Min Nan, though my Min Nan is not as fluent as my Mandarin now, I still consider it my mother tongue.

The cool thing about the Internet is that if you're curious about some data you can analyze it on your own. Perhaps you would show us what you've found? And what impact that's had on how you conduct business?

We parsed a lot of the data from App reviews (about 100 million) in different countries. Languages used are predominantly skewered toward english or other european languages. For example, in Israel, 51% of all reviews are in english while only 46% are in Hebrew.

Writeup on major countries: http://blog.sensortower.com/blog/2013/11/27/what-apple-app-s...

That's pretty cool analysis. Do you think the high proportion of English relative to other non-native languages could be related to the general affluence of Apple users? It would be interesting to see how these ratios stack up relative to the languages used in reviews on some of the Android app stores in their respective countries. Thanks for sharing.

I suspect the affluence of smartphone users in general is more selected towards those that can speak/write english. Also, most apps are not localized to native languages, and so the people that use apps are exposed to english more often. It's much harder to get reviews on Android, especially for different countries.

I guess the title "Only 240 languages are used online" doesn't sound like a problem? That probably represents 5 or 6 of 7 billion people?

I think the world should move towards even fewer languages than 240. However, it would be very sad if we didn't capture the complete grammar, vocabulary, etc. of those other languages before they die, along with lots of audio, to preserve them.

You're promoting both the killing and preserving of languages in the same paragraph, seems a bit misguided. Surely if they're worth documenting, then its worth trying to preserve their use as well.

Surely the world should be moving to a place where it doesn't matter what language the source is, at least online. A fully competent Google translate would seamlessly allow everyone to communicate without destroying culture in the process.

Why promote the use of obsolete technology (including language), when it can be satisfactorily preserved in museums?

If it doesn't happen the same as it happen to Alexandria's Library, that is.

Because "satisfactorily preserved" and "museum" are contradictory for a language.

A spoken language is a living thing -- and it carries with it the patina of all those that have used it, their stories and how they shaped the language in return etc.

Relegating to some museum excibits (that noone will care about, let's be honest) is like being satisfied by a polaroid of your significant other, instead of him/her.

We're not talking about artifacts though; we're talking about people.

Yes it would be better to be able to observe and use a language within a native, living body of people in order to be able to well comprehend and thoroughly appreciate the nuance, breadth and depth of a language. But in order to do that you also must require that population not to use some other language which may well benefit them (in terms of education, trade, international relations, access to information, etc.).

In Wales for example the historic Welsh language is being propped up at vast expense to a relatively poor region of the UK. All lessons¹ in [state run] primary schools (4-11 year olds) are used to promote the historic Welsh language and children are taught that their own language - predominantly English - is to be deprecated in preference to Welsh.

None of that serves the children. Without government interference, if trends continued, barely a handful of people of any age would speak historic Welsh. As it is now - made compulsory for all children in state schools - use of historic Welsh language is still declining.

That would indeed be a loss to those interested in Brythonic languages or who mourn for the simpler times of their youth when they spoke historic Welsh or who despise the English people so fervently that using "their" language is anathema [I'm not even joking], or whatever. But, would it be any loss for the people living in Wales as a whole? I warrant not. Indeed it seems it would be a gain.

---

¹ PISA results for 2012, http://priceonomics.com/only-4-of-languages-are-used-online/, are just in and show Wales is falling behind the rest of the UK and that the UK is slipping relative to the other countries surveyed. I think an insidious over-emphasis on historic Welsh is at least in part to blame.

Why promote the preservation of obsolete species (including animals), when they can be satisfactorily preserved in museums?

When not having those animals around improves the lives of millions of people?

I think the majority of the 6000 languages are spoken by very few people. Look at the number of languages spoken in India, for example.

http://en.wikipedia.org/wiki/List_of_languages_by_number_of_...

Is there a benefit to having all those languages where less than 100,000 people speak the language?

If it were possible to translate every concept or expression, transparently from any language into any other, there wouldn't be any real reason to have more than one language. Or, in other words, if the only differences between languages are cosmetic, different words and slightly different grammar, then there would be more to gain from being able to communicate without using a machine as an intermediary than from keeping the old phonemes (which will eventually be replaced anyway) in use.

That reminds me of William Temple Hornaday's, who killed a whole herd of Buffalos to preserve them.

http://www.loe.org/shows/segments.html?programID=12-P13-0002...

I'd bet a large portion of these languages don't even have official written languages. Some of the Chinese languages have tens of millions of native speakers and even they have no standardized way of writing their languages. When you start getting into languages that are specific to a tribe or very local area, the odds seem pretty low.

The www.jw.org website (Jehovah's Witnesses website) is translated to 316 languages atm, which is on par with the article's claim (170 actively used + 140 borderline cases). You can disagree with the site's message, but it tries to do a good job into sending the message to people all around the world in different languages. In comparison though, JW publications are printed in 600+ languages, so there's still lot to do for the website.

Edit: Going into "publications" section, there's a picker which lets you select a language from total of 538.

I hope the research cited is better, because the article makes a complete hash of mixing interface language, script and language actually used to communicate. All of these are thrown together as "use of language on digital stuff".

It's quite common around the world to use an interface in language X (because the options are limited) but communicate in their own language. (I personally do it all the time, so much so that I prefer an English interface over one in my own language for the sake of consistency.)

Lack of support for certain kinds of script may be an issue, but that never becomes clear from the article. If the data is heavily based on the formal support for languages (instead of actual usage) then it's seriously skewed. And even if it isn't, the article doesn't tell us where the actual problem is.

I can't decide whether this is a good thing or a bad thing.

That depends on whether you are globalist or nationalist.

Globalist - good, less languages - less barriers. Good for international commerce and knowledge sharing.

Nationalist - bad, loss of language means some part of cultural heritage lost.

Truth is always somewhere in between.

Many regional languages and dialects have been lost due to the efforts of nationalists, not globalists. Nationalists favour the use of a single language for all citizens of a country.

"Globalist - good, less languages - less barriers. Good for international commerce and knowledge sharing."

This is wrong, as there being more languages does not mean everyone is monolingual.

http://en.wikipedia.org/wiki/Lingua_franca

"Truth is always somewhere in between." What if I claim it isn't?

>"Truth is always somewhere in between." What if I claim it isn't?

Then, axiomatically, truth is somewhere in between "possibly being near some extreme" and being "somewhere in between".

So it's not in between anymore...

>Truth is always somewhere in between.

Either that's totally wrong (implies equal weighting), or it's a near-tautology (only false if one position is less than 100% accurate).

>Either that's totally wrong (implies equal weighting)

Equal weighting would be "Truth is always in the middle".

>it's a near-tautology

Another way to say near-tautology is "something that is true".

> Another way to say near-tautology is "something that is true".

Something that is true, but has a completely different meaning from what a naive reading implies.

Your statement is only true if we include the extrema at the "somewhere in the middle" interval.

> Truth is always somewhere in between.

Good thing we agree about globalization otherwise I'd grab one of your goalposts and run with it ;)

196 countries and 240 languages are represented online. The other 5000+ languages being used aren't national languages.

Loss of a language also often means a slightly unique way of seeing the world and thinking/talking about it is lost. Many languages have cool or weird features [1], and I care about that even though I can't communicate with speakers of those languages and therefore will never know the difference. I don't care about the national heritage of others, it's theirs - I care about diversity for the sake of it existing, not for the sake of me experiencing it. If we all thought the same stuff in the same ways, without any barriers, that would be much worse than having disconnected pockets of diversity, IMHO.

[1] https://www.youtube.com/watch?v=QYlVJlmjLEc

That way lies the Sapir-Whorf hypothesis, and madness. The value isn't about (slightly) unique ways of thinking/talking about things, but (slightly) unique ways of expressing those thoughts. That is, without at least a good description of as many of the various grammars that exist, we lose a lot of data that can help us to understand language as a whole. F'rinstance, if languages like Hixkaryana had died out before being described, then the prevailing belief that OVS was an "impossible" regular order (as opposed to a special construction) would likely have become "truth", and the possibilities only explored in artificial languages like Klingon. Losing that data means that we are losing the ability to make inferences as to how language works at the most fundamental level in the human brain.

On the other hand, it is rather inconvenient (from the perspective of a great global society) to have as many "pure" languages on the ground as we have (and have had). But while linguae francae bridge gaps across culture, full-time bilingualism and multilingualism tend to cross-fertilize grammars (and, of course, vocabularies), which has the effect of eroding the more interesting (and the most difficult) aspects of grammars in the smaller languages.

> That way lies the Sapir-Whorf hypothesis

What's wrong with it? I'm not familiar with it, but Wikipedia says "Current researchers such as Lera Boroditsky, John A. Lucy and Stephen C. Levinson believe that language influences thought, but in more limited ways than the broadest early claims".

> The value isn't about (slightly) unique ways of thinking/talking about things, but (slightly) unique ways of expressing those thoughts.

How is talking not expressing thoughts? Anyway.. Then you go on to make good points, but I don't see what makes you think they're in contradiction to anything I said. I was not talking about "the value", someone else claimed that everybody who cares about language diversity must be a nationalist, and that being a "globalist" means welcoming putting all eggs in one basket. I completely disagree with that, but for my nilly-willy reasons. If you have other reasons that's fine, they're good reasons, but to say that my view is vaguely wrong, because there lies madness for reasons you don't care to go into, well.

I personally do not care as much about what is accessible to "us" as about what is accessible to itself, wether I am aware of it or not. If there was a button to exterminate all parts of the universe we will never have access to and then forget that happened, I wouldn't press it, even if you bribed me, I hope. And again, remember this is in response to someone saying the only reason someone would care is X; I'm not saying my opinion is correct or perfect in any way, I'm just saying it's not X, cuz it's not :P

Just by way of explanation, B L Whorf's hypothesis (a considerable expansion of Sapir's original opinion) was that vocabulary, and more importantly, grammar, guided (and was guided by) modes of thought. A particularly famous example was that Hopi, having no past or future tense, which led its speakers to conceive of time as cyclical in nature, as opposed to the almost-neurotical need for Europeans to keep historical records and plan everything out. Quite apart from the fact that his analysis of the Hopi language was hideously inadequate (it is as tense-filled as most European languages, lacking only the explicit present progressive that is a peculiar feature of English and Celtic languages), he seems to have managed to miss the fact that the culture had developed a sophisticated set of timekeeping technologies as well. Much of the follow-on research ("language as culture" is a tasty morsel for anyone bent on believing that language is nothing more than learned behaviour) can be shown to use specific meanings gathered from informants for words that can be used to express much more general concepts. The problem is that the researchers failed to ask if the words could be used in another way.

The provable aspects revolve mostly around concepts like the colour "grue". That is, people whose languages lack separate words for blue and green, using instead a single word for both that could be rendered in English as "grue", take longer to distinguish between blue and green. It's not that they can't see the difference, but they initially file blue things and green things under the same heading, and only go on to distinguish between them when it's necessary. We have the same problem in English with yellow -- only people who've spent enough time in the world of colour would be likely to immediately distinguish warmer/redder yellows (like cadmium yellow) from colder/greener ones (like lemon yellow). And they learn that (usually) by figuring out that the two are not interchangeable when mixing colours -- both yellow and blue done wrongly and yellow and red done wrongly result in a muddy brown rather than the expected orange and green. It doesn't affect our perception, just our initial means of filing things, and it means that some circumlocution is needed when describing a particular yellow to somebody who doesn't have the technical vocabulary.

We also don't have a single word for "someone who always feels cold" as the French do, but we all know a few frileux even if we don't have a word for it, and are aware of that knowledge. We have dropped hither, thither and whither, but are still aware of the difference between direction and location when we use here, there and where. A large number of the world's languages don't have sex-linked gender, but their speakers are perfectly capable of telling the difference between men and women.

The only place where the Sapir-Whorf hypothesis really holds any water is in the realm of computer languages (and similar ordering-machines-around environments). Sometimes it is impossible to express a thought/instruction even when the machine is perfectly capable of carrying it out. Sometimes it's just so far away from idiomatic use of the language that it's better from any perspective to find a suitable paraphrase that is idiomatic. But these are toy languages designed to accomplish specific goals, not general-purpose languages that anybody has to live in. (Try having a relationship in Lisp and see how far you get.) In the human world, the ability or inability to express a thought has a lot more to do with experience and culture than with any limitations that the language you speak may impose on you.

> the provable aspects revolve mostly around concepts like the colour "grue". That is, people whose languages lack separate words for blue and green, using instead a single word for both that could be rendered in English as "grue", take longer to distinguish between blue and green.

What about languages that require the speaker to keep track of their spatial orientation?

And how come thinking and talking in English and German are both slightly different for me? Those languages aren't even far apart, but in my mind one can't replace the other.

> In the human world, the ability or inability to express a thought has a lot more to do with experience and culture than with any limitations that the language you speak may impose on you.

And again with the strawmen. Where did I claim such radical difference as being completely unable to express a specific thought, or seeing the world completely differently? Nowhere.

There's one big difference though, jokes. You can make some jokes in some languages, but they just don't work in others.

That is, though, precisely what the Sapir-Whorf hypothesis claims -- I never said you were claiming anything of the sort, merely that the hypothesis did. Linguistics, in general, is a fascinating area of study (though if you are good at spotting straw men where none exist at all, you'll find yourself twitching a lot).

That sounds very liberal and open minded, but it really isn't. Us native speakers of minor languages speak that minor language because it is the only communications tool that allows us to fully understand and to be fully understood, not just to provide you normal people a little extra insight.

Many people think of language as part of their group identity, but in a very real way it is just a tool that makes it possible for multiple people to be a group. A minor language confines you to a smaller group, and is therefor a less useful tool.

This is why minority children don't want to preserve the language of their grandparents. It confines you to a lower level of understanding and being understood, to a less successful education and career, and worst of all, to being a quaint object of study and marvel for the likes of you.

Why does it have to be either? At best it's convenient, at worst it doesn't effect you. No need to make a value judgment. :)

As a native speaker of a geographically confined language, I've come to the conclusion that a small language can be a prison. Here, in the small corner of the world where I just happened to be born, I understand everything. Every nuance, every idiom, even speech with a heavy regional accent.

As soon as I move 100 miles in any direction, this clarity disappears. Despite my decent grasp of English, I miss things. It is like my verbal IQ is lower, and I have become hard of hearing. I don't understand slang, and regional accents are difficult to make out. Every mispronunciation or malapropism is a risk of ridicule or misunderstanding.

Linguistic diversity may seem nice in theory, but not for the sake of performing a of quaint tribal dance for tourist dollars. It is simply undeniably more useful to be a native speaker of a major language. I'm trying to raise my kids as native speakers of English, to prevent them from being trapped the way their old man was. As far as I'm concerned, minority languages are better off going the way of Latin or Etruscan.

"A language’s Wikipedia presence was one of the most important indicators of its ability to leap into the digital age."

Which made me curios to search for and find Vicipaedia http://la.wikipedia.org/wiki/Vicipaedia:Pagina_prima, it's in the 10k+ article category.

I would rather focus on the many opportunities technology and the internet offers to preserve those 6000+ languages that are currently not being used online. It's not very bold to say that technology will prevent them from disappearing completely unlike many languages in history that were not persevered.

That statistic is somewhat misleading. Or, at least, not very meaningful. A more meaningful statistic would acknowledge the different "weights" that certain languages carry (i.e., "how many people use it").

These types of discussion about language always reminds me of http://www.economist.com/node/15108609 which I find a fascinating read.

I don't think wikipedia is such a great way to decide on a language. A better way is things like chat messages, but they are harder to analyze.

As someone who lives in an area where language is extremely political, and government is already involved in language-related legislation in day-to-day life, I worry that this is a potential vector for more government intrusion on the Internet. The excuses are practically built-in, from the perspective of the nationalists.

La oss håpe dine landsmenn ikke tar til internett med sitt eget språk der det ikke passer.

And here is a map of where they are used on Twitter

http://www.ccs.neu.edu/home/qianz/MapTwitterLanguage/v1/inde...

Useless things die out - it's a natural process.

Things which aren't useless also die out.

Well, for languages the main way for dying out is when ~2 generations of their population decide that this language is useless for them and the vast majority choose not to use it anymore.

There are exceptions like genocide and forced assimilation of conquered nations; but these refer to extermination of languages/cultures, not the much more widespread issue of them simply dying out through disuse and voluntary assimilation.

I wouldn't wonder why there is only 4%. There are thousands and thousands of languages allover the world.

Are most languages really a big loss? I suppose there are some interesting variations in grammar, but what else? What can we learn from some obscure languages?

I really agree on this..

Reply on news.ycombinator.com