Languages in Europe by country, age group and mother tongue vs. learned. Based on 2012 Eurobarometer data
What is languageknowledge.eu ?
Languageknowledge.eu is a visualization of language knowledge in Europe by Jonathan Van Parys, launched to coincide with the 2012 European Day of Languages.
People like Martin have done such cool things with demographic and migration data. So when the European Commission published a new Eurobarometer survey on languages this summer, I thought it would be fun to do something interactive with language data in order to visualize language dynamics in Europe: which languages are growing in popularity, which languages are most spoken in each country, etc.
For ideas, questions or corrections feel free to email me at jvanparys@gmail.com
Thanks to @pvpbrussels, @odemarne and @madewulf for their feedback as I was working on the site.
I’d also like to thank the European Commission for making the raw survey data accessible and TNS for answering questions on the survey’s methodology.
Why not visualize it on a map?
You might be wondering why I didn’t just choose to visualize this data on a map, given that it has an obvious geographical dimension to it. My take is that maps aren’t so good at comparing countries because area size and position in space have strong meanings when you think about data, so unless geographic location or size explain something about the data, it just makes it harder to make sense of it and potentially sends confusing signals about the ranking of countries.
Plus lists and horizontal bar charts are particularly useful for direct comparison, especially when people can play with dimensions, as you can here with the age-groups and types of knowledge. Resizing and reordering bars just makes rankings and changes in position more obvious.
Where does the data come from and how are the statistics computed?
Language knowledge statistics are computed using the Eurobarometer survey published in July 2012. By the way, there are many other Eurobarometers out there: it’s well worth taking a look.
In early 2012, a little more than 27 thousand people all over Europe above the age of 15 were polled for the Eurobarometer; grossly-speaking, about one thousand per country. I grouped people between ages 15 and 34 (Younger), 35 and 54 (Middle) and 55+ (Older) to highlight how language knowledge evolves between generations.
In the survey, every person is asked to state their mother-tongue (Native) and which other languages they speak well enough to have a conversation (Learned), and how well they speak those learned languages (basic, good or very good). In line with other studies, I consider a person knows a learned language when they state their knowledge of it is good or very good. I cross-checked the statistics on languageknowledge.eu with those that appear in the official Eurobarometer report and they are fully consistent.
You may be surprised to see that certain languages you expected to see aren’t present. The reason is that the set of options offered to respondents in the survey only included the 30 or so languages included on this website.
Data on sub-titling was retrieved from Wikipedia’s article on Dubbing. The reason I included it is that, as you will see, the use of sub-titling rather than dubbing or voice over in the media is highly (and positively) correlated with the knowledge of foreign languages. Sub-titling on television may be the cheapest language school there is – that some countries still allow dubbing is just non-sensical. The other determining factor in explaining foreign language knowledge appears to be the size of a country’s native languages: the more widespread they are (as a native or learned language), the less a country is likely to be multi-lingual.
How reliable is the data?
This is obviously survey data based on self-assessment, so what it says is only as good as the representativeness of the sample and the ability of people to self-assess their knowledge of a language. I’ve noticed that some small languages may be over-represented in some countries because of sample bias. This being said however, I’ve found that results from the Eurobarometer language survey match more in-depth local surveys, are consistent with other European language learning surveys, including the first European survey on language competences, and tend to be correlated with compulsory school languages in countries.
