Friday, January 02, 2015

Catastrophic language learning in the UK – is online part of the answer?

Learning a second language is one of the toughest learning tasks you’ll ever tackle. Let’s be frank here, language learning in schools is nothing short of catastrophic. The whole population has to endure years of classroom instruction and the vast, vast majority come away with next to nothing. It’s lemming learning with young people falling off the cliff in their millions.
Learning a language outside of school in the UK is just as bad. I worked with the CEO of a major, global language learning company, who cheerfully told me that his market was not language learners at all but false starters, people who wanted to learn a language but would almost certainly fail. Some of the product, especially the BBC product, he claimed, were so bad that it was actually impossible to learn a language from the usual book and CDs.
In 2016, the Department for Education (DfE) data showed that the attempt at a forced rise in entries for GCSE languages through the EBacc staggered to a halt. Since 2002, entries for A level French have declined by about one third, and those for German by nearly half. Although more pupils are taking A levels in Spanish and other languages, these increases have not involved enough pupils to make up for the shortfalls in French and German. Language learning in the UK is a disaster. A British Council report also showed that the way languages are taught, in particular, the examination system, is preventing rather than aiding language learning.
Compare this with those young people in other countries learning English. I’d argue that the majority of those who learn English as a second language do not learn it at school but through immersion, and informal online learning. This is drive by the status of English as a global lingua franca and therefore usefulness. Yet we continue to harbour the belief that schools drive language learning. They do not, in fact there is evidence to suppose that the inhibit language learning with their rigid pedagogy and examination system.

I live in Brighton and for the last 3O years the town has been full of young people ‘coming here’ to learn English. They come here not just for the instruction but for the immersion, context and culture. That’s how you learn a language. However, given the resources available to a young person online, millions are learning English through music videos, YouTube, Netflix and hundreds of other online channels and resources.
But  that's not to say that we should give up.Indeeed, what's needed is a more proactive look at online learning to provide what schools don't provide - practice, immersion and context.
Duolingo
We may be seeing a breakthrough here, similar to that of the Khan Academy in maths. I’ve been using Duolingo for some time and recommended it to my son, who has a German girlfriend, and wants to pick up some basic German. People who have used it give it nothing but praise. The mix of exercises is good and, most of all, the commitment to practice, personalised practice, forcing you to refresh and strengthen your skills through spaced practice, is its real advantage. Apple named it ‘App of the Year’ in 2013.
1. Interface is exemplary
It’s simple, uses colour judiciously, with and the course, along with progress, is clear. I also like the way they make improvements to the interface. It’s getting better and better.
2. Personalised (Time)
You are not stuck in a class timeline but can go at your own pace, doing it whenever you have the spare time or are in the mood. This, in my view, is essential in language learning.
3. Personalised (Goals)
You can set your own goals; basic, casual, serious, regular, insane (aiming for 1-50 sessions per day). You are then tracked against these goals.
4. Personalised (Practice)
You progress according to effort and competence. In language learning this matters, as it is easy to get left behind. More specifically, it knows when you get a something wrong and brings it up again and again, until you get it mastered.
5. Chunking
The ’18 item’ chunks are topic based, and are tough enough to push you but not overlong so that you get demotivated. Within these test items you go back one step if you don’t get it right.
6. Practice
The focus on practice (not exposition & reading) is exemplary. It doesn’t let you away with short-term memory and performance. You’re forced to practice and refresh your skills, based on actual competence.
7. Gamification
There’s lots of hype about this but Duolingo really does put this literally into ‘practice’ with reward and punishment designs that are finely tuned.
8. Online
Language learning screams out to be online.  The advantages of online are many – the reminder emails & links to practice is a good push service as is the fact that it gives you good feedback. The personal tracking, available online is also a boon.
9. Mobile
The mobile app is good and also has a pronunciation checker. This is essential for a language programme, where you want to take advantage of those spare, dead-zones to brush up your skills. It also allows you to use it in situ, say on holiday.
10. Evidence
The ‘Duolingo Effectiveness Study’ by Vesselinov & Grego,
Looks at the effectiveness of Duolingo was measured as language improvement per one hour of study. It is a promising and convincing start on evidence. They also claim that it takes 34 hours on Duolingo compared to 55 on Rossetta Stone to do a semester of Spanish. I believe them.
11. FREE
It has no Ads and costs nothing, zero, zilch, zip, nil, nada, nought…… and I don’t mean Freemium, where they lure you into paying for other services – it really is free. Compare this to the cost of current education in school or language learning courses from CD or online.
This comes in late January 2014, with dashboards for classes and students along with some management tools. This is exactly the sort of tool teachers and language learners should be using. I think there is a strong argument for insisting that ALL language learning in schools should be using online courses such as these. My guess is that Duolingo is getting close to delivering GCSE standard language skills on its own and that many students could simply be set that goal without the need for much teacher intervention.
Social good business model
Ever had to type in those annoying letters on a web form? That’s ‘Captcha’. Luis von Ahn, who invented this security technique, took his Captcha idea and used it to harness all of this global brain power to decipher failed OCR words, especially older faded text, with ‘Recaptcha’. That’s why you often get two words – the system knows one, not the other. Over 10% of the word’s population has helped digitise books. It’s a crowdsourced social good.
Duolingo was started by that same Luis von Ahn when he saw that similar techniques could be used for translation. People learn a new language for free but translate at the same time. Duolingo works as you ‘learn by doing’. You move on to learning with real content by translating real content. Duolingo combines the translation attempts instantaneously. This is being done for the likes of Wikipedia. So learners create value while learning. This is a great business model for education. One social good pays for the other.
Or does it? Look at the URL. Duolingo is a for-profit organisation. Luis sold Captcha for $25 million to Google. I’m with Luis on this one. It’s important to get real revenue and that’s coming from real customers like CNN and Buzzfeed. Keeping it as a not-for-profit would have clipped its commercial wings and made it rely on grant income to progress and survive.
Conclusion
With the increase in quality of these tools, voice and pronunciation recognition and adaptive learning, it makes no sense to keep language learning trapped in the one-size-fits-all classroom, where generations of language learners have wasted their time. We should be moving quickly towards an evaluation and large scale trials using these tools.

2 comments:

Seb Schmoller said...

It's worth watching "Duolingo: the next chapter in human computation" a fascinating and funny 17 minute 2011 talk by Luis van Ahn, the inventor of Duolingo: https://www.youtube.com/watch?v=cQl6jUjFjp4.

What's very striking about Duolingo - apart from the way it enables free language learning - is the way that the work the learners do in Duolingo is contributing to the translation of texts, just as the work we do deciphering reCaptcha images is playing a part in the digitization of text.

Will said...

I agree. Schools have no idea how to learn languages. Duolingo is fantastic. I have finished the Italian section with great results. However, to increase progress, I think a mix of other tools help immensely. A good old fashioned grammar book is essential as a reference tool as new topics are introduced: verb tenses, etc. and I found audio courses such as Michel Thomas or Pimsleur surprisingly extremely effective. ( but expensive, unlike Duolingo).

My wife had no interest to learn a language, but the fun, well thought structure of Duolingo has hooked her too.