If you visit Japan and try to talk to thirty year olds in English, the first thing that may surprise you is that most of them understand very little and can speak even less. This is in spite of the fact that English was taught to them as a second language for many years in school.
Sometimes you’ll meet people who can get by in English. However, for the most part, even those who have spent some time abroad during their adult life generally have a very strong accent. They simply never learnt to pronounce English correctly.
Occasionally you’ll come across someone who speaks flawless English. There are two things you can be fairly sure about: (1) that person lived abroad as a child and (2) that person primarily did not learn English from a teacher, but rather from other children.
For many children in Japan it really is impractical for them to move abroad for a few years to learn English. So how about arranging play-dates with native English speaking children? Well, there really aren’t many native English speaking children in Japan, so we need to think of a new solution…

We now have skype and indeed other internet based video conferencing technologies. One of the great things about skype is that once you have your computer with mic and web-cam along with a decent broadband connection there are no more incremental costs.
Suppose we were to carry out the following experiment involving 4 year olds in England and France. At the start the English and French kids don’t share a common language. We have two English children sitting at a booth with a web connection and also two children in France. The children can see each other using skype and we’ve set-up an interactive fun game that they can play. If we let them play like this for one hour a day for a year, we’ll come back and look at their language ability.
Our hypothesis is that with this approach we can achieve a remarkable growth in second language ability for a very modest cost with only a small amount of staff training required. It is not necessary for the teacher based in France to speak English nor for the teacher in England to speak French. The children will be able to teach each other their respective languages though play and interaction over skype.
Our suggestion is that after a successful trial run, this method of teaching should be rolled out in all state-run schools for kids starting at the age of 3.
The skype based interaction between the kids needs to be more fun than forced.
If they get along well, then the same kids can meet up each day (via skype).
Small groups, such as 2 +2 may work better than one on one.
If we observe one session, with 2 English and 2 French kids, we will probably find that one language will become dominant. There may be one French child who dominates the talking and keeps proceedings in French. We will need to think about how to ensure that the French kids learn English (and vice versa. )
So if there is one 2+2 group that (almost) always speaks French, that’s fine, that group can be maintained, but half the time we may need to send those kids into different groups to ensure that the French kids are exposed to English.
To encourage one language over another we can try the following:
(i) The audio-visual interactive game that the kids play will be in one language to encourage use of that language.
(ii) Each child will be a member of two groups, one for their native language and one for their second language. So they will know that consistently when they see A and B ( via skype ) they’ll speak say French, but when they see E and F, they’ll speak English.
(iii) The booths can be colour coded, to build up an association between a colour and a language. That way they’ll know what language they’re expected to speak.
(iv) Suppose we want the kids to speak English, we may find that if we put the two English kids together at one computer and the 2 French kids at two separate computers that may help to encourage the use of English.
After initial trials we may be in a better position to understand how to influence the language that becomes dominant in a session. This could be useful if for example we find that there is a greater demand for kids in Italy to learn English rather than for English kids to learn Italian. We suspect that with some clever tricks one English kid may be able to teach 3 Italians to speak English. So a 3 to 1 ratio would be fine. However, iIf ( as we suspect ) the ratio is even more extreme we may need to get Italian kids with a few months experience of speaking English to teach those who are just starting out.
The initial trials should be carried out in under-privileged. It can provide a more useful proof of concept than if the schools chosen were elitist private schools.
If you're interested in the ideas of children learning from other children, then have a look at
this video of a TED talk given by Sugata Mitra: The child-driven education.