Thought I'd call attention to a list of languages that lichtrausch came up with on the old forum that seems very much in the spirit of this thread.
The list includes 20 modern languages (if we continue to count Hindi and Urdu separately but assume you'd want to learn both) plus 5 ancient ones, and although this wasn't the original goal, the modern languages on the list would allow you to talk to around 53% of the people in the world in their native tongue.
The list appeared in a
thread about panglottery, or "polyglottery with an emphasis on achieving a more global perspective," and it's meant to include "the languages of vast numbers of people spoken over vast areas of the world and from vastly diverse language families."
To that end, it includes at least one language from each of eleven major language families: the seven families that Deinonysus highlighted and four more. The families and the languages that represent them were chosen based on
total speakers rather than native speakers. And the families don't need to include more than one language.
lichtrausch wrote:Only 11 language families in the world contain a language that has at least 20 million speakers (Indo-European, Sino-Tibetan, Afroasiatic, Japonic, Korean, Austronesian, Austroasiatic, Turkic, Niger-Congo, Dravidian, Tai-Kadai). All of these language families contain at least one language that has at least 50 million speakers.
Building on Deinonysus's chart, here are the eleven families. All the numbers are still from the
Nationalencyklopedin list of the 100 languages with the most native speakers (so the second column shows the number of languages from each family in the top 100, and the third column shows the percent of the world's population that speaks those languages natively). But there's a new column at the far right showing the language in each family with the most total speakers. In a few cases, this is different from the language in that family with the most native speakers. And the languages in this rightmost column are the ones in lichtrausch's list.
In the Dravidian family, the language with the most total speakers is Telugu, but Tamil isn't far behind, so the list leaves the choice between the two open.
Following the
Nationalencyklopedin, the chart below counts Malaysian and Indonesian together as Malay.
Family | # of Languages | Percentage of world population | Most spoken language (native) | Most spoken language (total) |
Indo-European | 44 | 41.05% | Spanish | English |
Sino-Tibetan | 11 | 19.79% | Mandarin | Mandarin |
Afro-Asiatic | 5 | 5.70% | Arabic | Arabic |
Austronesian | 9 | 4.49% | Javanese | Malay (inc. Malaysian and Indonesian) |
Dravidian | 4 | 3.36% | Telugu | Telugu |
Niger-Congo | 11 | 2.28% | Yoruba | Swahili |
Turkic | 6 | 2.21% | Turkish | Turkish |
Japonic | 1 | 1.92% | Japanese | Japanese |
Austroasiatic | 2 | 1.38% | Vietnamese | Vietnamese |
Koreanic | 1 | 1.14% | Korean | Korean |
Tai-Kadai | 2 | 1.09% | Thai | Thai |
In addition to the most widely spoken language from each of those families, the list includes languages from two other categories, plus a few languages that are special cases. Here's lichtrausch explaining the choice of languages, from post 19 on
this page of the thread...
My thinking on panglottery has evolved a little bit and these are the languages I now think are worth taking a closer look at: First of all the top language from each of the 11 premier languages families in the world*. Then the huge, continent spanning languages of the colonial powers**. Then the great classical languages*** which have left behind a plethora of classical literature and provide great insight into some ancient civilizations and the daughter civilizations they spawned.
And now I want to consider Persian and Hindi a little more. They aren't the top language of any macrofamily, they don't extend over an entire continent, but they are the top languages of their respective subfamilies which broke off a long time ago from PIE and spawned dozens (hundreds?) of languages. These subfamilies are so large and diverse, not to mention distinct from their distant PIE cousins, that there is a lot of value in learning a language from each of them. Persian and Hindi are the obvious choices because of number of speakers, abundance of learning resources, geographic area covered, and wealth of media available in them.
German and Italian don't fit neatly into any of the above categories, but if you total up their past cultural contributions plus the weight they carry today (especially German as the central economic and demographic power of Western Europe), they are hard to neglect. I also think it would be beneficial to learn languages from any of the other dozens of language families in the world. However there are usually problems with this such as lack of learning materials, lack of places to speak it, lack of media to enjoy in the language, and so on. If you can overcome these challenges, by all means have a go at one of them.
[…]
*English, Turkish, Arabic, Mandarin, Korean, Japanese, Vietnamese, Thai, Indonesian, Swahili, Tamil or Telugu
**English, Spanish, Portuguese, French, Russian
***Latin, Ancient Greek, Arabic, Sanskrit, Classical Chinese
Others: German, Italian, Persian, Hindi
Once again building on Deinonysus's work, here are the modern languages on the list ranked by number of native speakers. I considered adding a column showing each language's total speakers, but that would require mixing in data from a different source, and that source probably wouldn't make all the same choices about which languages to count together, so it soon gets messy. So this sticks to the
Nationalencyklopedin data. (For a ranking based on total speakers, see
here.)
The rightmost column shows the cumulative percent of the world population but leaves out Tamil and Telugu, since the list leaves that choice open. Either can be added at the end.
And we don't have
Nationalencyklopedin data for Swahili, since it doesn't make their list of the 100 languages with the most native speakers. (According to
Wikipedia, estimates of the number of native speakers of Swahili "range from 2 million (2003) to 15 million (2012)," whereas total speakers is much higher, perhaps around 100 million.)
Rank | Language | Family | Native speakers (million) | % of world population | Cumulative % of world population |
1 | Mandarin (entire branch) | Sino-Tibetan | 935 | 14.10% | 14.10% |
2 | Spanish | Indo-European | 390 | 5.85% | 19.95% |
3 | English | Indo-European | 365 | 5.52% | 25.47% |
4 | Hindi | Indo-European | 295 | 4.46% | 29.93% |
5 | Arabic | Afro-Asiatic | 280 | 4.23% | 34.16% |
6 | Portuguese | Indo-European | 205 | 3.08% | 37.24% |
8 | Russian | Indo-European | 160 | 2.42% | 39.66% |
9 | Japanese | Japonic | 125 | 1.92% | 41.58% |
11 | German | Indo-European | 92 | 1.39% | 42.97% |
14 | Malay (inc. Malaysian and Indonesian) | Austronesian | 77 | 1.16% | 44.13% |
15 | Telugu | Dravidian | 76 | 1.15% | – |
16 | Vietnamese | Austroasiatic | 76 | 1.14% | 45.27% |
17 | Korean | Koreanic | 76 | 1.14% | 46.41% |
18 | French | Indo-European | 75 | 1.12% | 47.53% |
20 | Tamil | Dravidian | 70 | 1.06% | – |
21 | Urdu | Indo-European | 66 | 0.99% | 48.52% |
22 | Turkish | Turkic | 63 | 0.95% | 49.47% |
23 | Italian | Indo-European | 59 | 0.90% | 50.37% |
25 | Thai | Tai-Kadai | 56 | 0.85% | 51.22% |
29 | Persian | Indo-European | 45 | 0.68% | 51.90% |
– | Swahili | Niger-Congo | ? | ? | – |
So that's it. This wouldn't be the quickest or easiest path to being able to speak with half the world in their native tongue, but it would achieve that goal. And I find it to be a very balanced and cohesive list with excellent coverage in terms of geography, culture, and linguistic diversity. By the way, it includes the thirteen
supercentral languages that Hashimi mentioned.
I find it interesting to think about the different approaches one could take if they wanted to pursue the goal explored in this thread of being able to speak to lots of people around the world. For example, if you could only choose one, given limited time and energy, would you rather learn a language with more native speakers (more people you could talk to and have it "go to the heart") such as Yoruba or Javanese, or one with more total speakers and greater geographic reach, such as Swahili or Malay? Would you rather sample as many different language families as possible, or sample fewer families but learn more languages from each? And so on...