Speciation and extinction

From one ancestral language (e.g. latin), several languages are born (e.g. spanish, portuguese, french, italian, romanian, ...). Languages therefore speciate. Such speciation could eventually happen via hybridization of existing languages. Languages also get extinct.

When the rate of speciation is higher than the rate of extinction, then the number of languages is increasing and when the opposite is true, then the number of languages is decreasing.

Atlas of endangered languages

UNESCO, in its Atlas of the World’s Languages in Danger (also listed on wikipedia) recognizes a large number of endangered languages. Increase world-wide communication and globalization makes me think that the number of languages should be currently decreasing.


Do we have estimates of the speciation rates and extinction rates of languages (or estimates of the rate of decrease/increase) today? A rate would be in unit of number of languages lost or gain per amount of time.

  • 1
    Yes, it's decreasing. But most of those going extinct are very small anyway, that's where all the churn is. Sep 11, 2017 at 19:08

3 Answers 3


If language-name plays a central role in determining what "language" one speaks (I speak English, as do hundreds of millions of others), then the number of languages is decreasing. Many languages are completely disappearing very rapidly, so that the bottom 80% of languages with only a couple hundred thousand speakers or fewer are likely to be mostly gone in another 100 or so years.

At the same time, if you look at "distinctive speech forms", the number of languages may be increasing. For example, the number of distinct speech forms known as English has increased over the past 100 years (and it's not just English where this happens). We tend to talk about those differences as being "dialects".

There are no meaningful estimates of rate of decrease (in the first sense of "language"), because there no predicting what facts cause language attrition. 40 years ago, people would not have predicted the current decline in linguistic diversity, because they were not aware of the impact that cell phones, the internet, and general urbanization would have. Social isolation encourages language diversity (you generally only interact with people near you who speak the same language, and have no idea that you "ought to" speak English or one of the other major languages, to get ahead).

  • Thanks. Just to correct, according to this site, there are 1.5 billions (not just hundreds of millions) who speak english (375 millions of which speak it natively).
    – Remi.b
    Sep 11, 2017 at 19:18

I would like to add to what user6726 said, in that even "distinctive speech forms" is a vague idea. There's a case to be made that idiolects (one's personal language patterns) are languages, at which point the number of languages in the world is equal to the number of people in the world, more or less.

It's not very satisfying, but the definition of "language" in linguistics has never been very concise at all. To someone working on language planning, and even many working in generative grammar, sociopolitical entities such as "English" may be taken for granted as being languages, but for a sociolinguist interested in variation or a psycholinguist interested in idiolects, that definition doesn't fly at all.


Languages often disappear when they coexist with an other language.

Official languages are often institutionalized (national, supranational), whereas dialects are in general more volatile (currently Catalan vs. Spanish). Frisian in the Netherlands one official language, in Germany quite lost.

In general in an ever smaller world dialects, regional/ethnic languages may fear extinction.

Languages came into (stronger) existance by the creation of Israel in the previous millenium, the balkan fragmentation (Croatian becoming more "Croatian" as opposed to Serbian). Separate evolution of a language like American English from British English is hard to judge.

So total disappearance might be measured, but other numbers are (1) hard to collect numerically and objectively (language like Catalan being part of ethnic pride). One might for instance count languages in the internet... - a minority.

Or are there sufficient linguistic studies?

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.