More thoughts on second language choice

Language Pop.(m)/FSI Ratio
English 611.4 / 00 =
Spanish 346.3 / 24 = 14.43
Mandarin 706.4 / 88 = 8.03
Portuguese 168.1 / 24 = 7.00
French 108.9 / 24 = 4.54
Russian 178.9 / 44 = 4.07
German 95.4 / 30 = 3.18
Malay 101.7 / 36 = 2.83
Italian 60.5 / 24 = 2.52
Turkish/Azeri 91.6 / 44 = 2.08
Arabic 136.5 / 88 = 1.55
Japanese 127.1 / 88 = 1.44
Dutch/Afrikaans 33.6 / 24 = 1.40
Hindi 53.2 / 44 = 1.21
Persian 47.7 / 44 = 1.08

This table displays one method of ranking languages by dividing the size of their middle class populations by their relative difficulty for native English speakers. You can find some justification for this method in my previous articles (2012, 2013). This year, I switched to using Foreign Service Institute difficulty ranges from aboutworldlanguages.com as my divisor. This gives me finer gradations among the easier languages, but it pushes the most difficult languages down the list. The Python code that produced the table is here.

When I first assembled this table in 2012, I was considering learning numerous languages. Part of my motivation was to justify some of the less common choices. After meditating on those choices for a couple of years, I now feel that was a waste of time. The "imperial" languages of Europe and China are much more interesting and useful than the alternatives.

In 2012, I also had high expectations of learning multiple languages. Since then, I have developed a greater respect for the difficulty of learning a language. As an adult learner, it is quite hard to break the bonds of categorical hearing. After a couple of years of seriously studying Spanish, I still find it very difficult to separate the components of an audio stream at full speed. It is possible that I could become adept at the written language but I might never reach the point of being able to listen to TV in the background the way that I can with English. I have also been studying Mandarin Chinese where the tones provide a different listening challenge. There are people who believe it is impossible for adults to learn tones. I do not believe that, but it has been harder than I expected. Perhaps German or Russian would be easier in this respect.

I enjoyed school, but traditional scholastic courses for languages bored me. I much prefer the 10,000 Sentences method for acquiring language structures through massive input, but even if you can find a ready cache of material, this process requires a large investment of time and effort. If I had started twenty years ago, I would feel better about that investment. Now I have to start thinking about the opportunity cost of lost time. The counterpoint to that argument is that the time does not matter if you enjoy the journey. This truly is the most critical question in learning a language: where do you want to spend your time? What culture is so attractive to you that you are willing to spend thousands of hours immersing yourself in it? Or, vice versa, which cultures are so terrible that you cannot bear to spend time in them, despite any numerical analysis?


Any of the top European languages are attractive. Spanish, French, and Russian are working languages of the United Nations due to their multi-national utility. In late 2014, PNAS produced an interesting study that weighed the flow of book translations between languages. The second page has a graphical depiction of those relationships which shows that English, French, and Russian are the strongest hub languages in books. These three languages are distinguished not only by volume of translations, but they also form lobes within the graph. French and Russian intelligentsia have at times actively sought to protect their cultural and academic spheres against English incursions. Learning French or Russian could give you access to significant works not available in English.

Personally, I think that French history, literature, and cinema makes the French language the most interesting language in Europe. France is also one of the few countries in Europe with a healthy birth rate. One speculative study projected that population growth in Francophone Africa could make French the largest language in the western world by 2050, but this report has provoked some disagreement. Another report by Jacques Attali (French) predicted a wider range of possibilities, but included that high number as well. It is hard to find good news for Russia.

East Asia

Mandarin is the compelling choice in East Asia. Japan has a more attractive political system and a higher average standard of living, but at a personal level, Japanese society seems somewhat broken. Japan is facing a demographic decline more immediate and severe than Russia's decline. On many levels, Chinese culture seems healthier. Literacy is the great obstacle in both languages, since a student must learn thousands of ideographic characters. I enjoy this activity, but the sheer number of hanzi/kanji are overwhelming, and I see now that this project could easily take a decade. I suspect most people just do not have the stamina for it. (Note that respected academics say you should learn to speak, then learn to read. Personally, I could not wait that long.)

South Asia

Arabic is an "imperial" language, but it does not belong in the top tier of desirable languages. All languages dominated by Islamic governments share problems: censorship, religious violence, and corruption. The culture of political Islam is deeply broken but no Muslim state seems willing or able to have a rational discussion about it. Arabic has the highest profile, but outside of religion, even Arabic supporters will admit there is little reason to learn it. One could hope that Arabic would be a bridge language to other cultures, but it is quite weak in that role. "There probably has been no imperialism like that of Islam and the Arabs....Islam seeks as an article of the faith to erase the past," -- V.S. Naipaul, Beyond Belief. Malaysia and Indonesia seem diverse, but the Muslim majorities there are very insecure. Moderate Malaysia has so many fatwas there’s a website to keep track of them and there is a political battle at the moment to impose Muslim criminal penalties in Kelantan state. Historically, Indonesia has done little to curb outbreaks of religious violence.

2020 Estimate
Language Pop.(m)/FSI Ratio Change
English 748.7 / 00 = +22%
Spanish 412.0 / 24 = 17.17 +19%
Mandarin 1448.2 / 88 = 16.46 +105%
Portuguese 215.5 / 24 = 8.98 +28%
Malay 207.1 / 36 = 5.75 +104%
French 123.8 / 24 = 5.16 +14%
Russian 188.1 / 44 = 4.28 +5%
Hindi 150.3 / 44 = 3.42 +183%
German 96.7 / 30 = 3.22 +1%
Italian 62.0 / 24 = 2.58 +2%
Arabic 224.2 / 88 = 2.55 +64%
Turkish/Azeri 103.6 / 44 = 2.35 +13%
Dutch/Afrikaans 37.2 / 24 = 1.55 +11%
Japanese 124.3 / 88 = 1.41 -2%
Persian 61.9 / 44 = 1.41 +30%

The bright spot in southern Asia is India, the most populous democracy in the world with a colorful, pluralistic culture. India is remarkable for the size of its cinema: it is the world's largest producer of films. Very few films are in English. Surprisingly, Hindi does not dominate cinema production. (In 2014, Hindi films were third by quantity.) Dancing Hindus attract me more than violent Muslims. India is a weak publishing market for modern literature however. Perhaps the literati are all writing in English. Perhaps the growing middle class will provoke a renaissance, but for now, the student must rely on older material.


One could travel across most of Africa with a working knowledge of English, French, and Portuguese. Local languages have little literate history so they are eclipsed by the former colonial languages.


This second table shows the results of the same calculations using the IMF's estimates for population and GDP in 2020. The Russian future population is probably inflated since it relies on increases in the Ukraine, Kazakhstan, and Uzbekistan, all areas where the Russian language is less welcome now than in the past. Several Asian languages have gained as economic growth affects more people and increases living standards. However, four of those "rising stars" are dominated by Islamic cultures which have little to offer me. Even though Malay will surge into the top ten, its two leading states have cultural issues. Mandarin and Hindi combine high growth forecasts with strong media markets and long, literate histories. They are the most interesting choices outside of Europe.

In spite of this Asian growth surge, the colonial Romance languages will continue to be the best choices in the short term. French, Spanish, and Portuguese are widely distributed across multiple governments with western ideas. Each of them is relatively easy and they have a network effect: learn one and the next one is even easier. You could become fluent in all three (and perhaps Italian) before you could become fluent in Mandarin Chinese.