Language Learning and Programmed Instruction

December 2006

A while ago, one of my friends asked me what I thought of Michel Thomas's technique for teaching languages. His tapes essentially consist of him teaching a couple of people a language. You listen to them, pausing as necessary. He essentially gives you a series of sentences in English to translate getting progressively more difficult with the odd point of grammar thrown in, but without much direct reference reference to grammar - no talk of conjugating verbs and so on.

My instant reaction was that what he does is basically programmed instruction. And then I looked on the web and was disappointed at how little I could find about programmed instruction and what research has been done on it. I couldn't even figure out if there was a generally agreed definition. Anyway, I think of programmed instruction as mastery testing with sequences of questions for which there is automated feedback.

Programmed instruction gets a surprisingly bad rep, particularly when you consider it's what is behind lots of video games. I think there are a few reasons for this. It's not easy to do it well so lots of examples of it aren't very good - Michel Thomas does a good job however. It's only really suitable for certain types of learning - learning grammar is something that it's fine for. It also doesn't work as soon as you get 'stuck' or want to ask about something. The main reason though is that by itself it's boring. That's why video games need cool graphics and reward mechanisms that motivate people as well. I think the problem of dullness ends up being the weakness of Michel Thomas's technique - it does get quite samey after a while which makes it easy to give up. I don't know anything about language teaching, but have been on the receiving end of a fair amount and I do think however it's an improvement on how grammar is usually taught. You just need a change of pace occasionally and the odd thing to keep you motivated.

I have a good memory so I'm good at learning languages and picking up grammar, but I've seen other people struggle a lot. They often end up almost learning a phrasebook off by heart rather than actually figuring out the rules that will let them construct new sentences for themselves. They often get so good at learning phrases off by heart that people think they've mastered bits of the grammar when they haven't.

So for fun, here are some of my pet gripes about the way that languages are often taught.

Teaching grammar by making you use exactly the same construct again and again in quick succession. I'm thinking of the sort of exercise where you have to say 'I must go out' then 'I must wash' then 'I must get up'. The problem with these is that once you've done the first you just copy it in all the others and you don't actually remember anything. Exercises about grammar in fact become about vocabulary.

Being authentic in a boring way. It's quite fun to get the odd newspaper article read, but really I want to spend my brain cells trying to learn the language rather than trying to add up menu prices. Don't distract me when I'm still getting to grips with the basics. Also, remember, it's far more important that I learn to say things that I find useful than that everything has to come from a genuine conversation among natives.

Overloading me with vocabulary and phrases I don't need yet. I possess a dictionary. I really don't need or want to learn the words for every single country in the world in my first week learning the language. Give me some grammar in the first lesson so I'm starting to feel empowered rather than that I'm learning stuff by rote. If you want to throw in lots of obscure idioms to make things authentic, at least be explicit about what I do need to remember and what you're just throwing in for exposure. I understand that there's a skill in figuring out meaning when you don't understand all the words, but at the same time I need to learn not to rely on that too much.

Shying away from grammar too much. If I've reached the stage of trying to reverse-engineer the grammar out of frustration, something is wrong. And point out irregularities please too if they crop up, otherwise my brain will fry as I try and reconcile everything.

Language Learning on the iPhone

I have discovered that two of the most important features of the iPhone are:

i) that it is comfortable to sit and read from your iPhone when sitting on a sofa

ii) that the speed of web browsing is sufficiently painful, at least for those of us still on a 3G, to make using it remarkably distraction-free.

This combination makes the affordances of the iPhone interestingly different from a laptop or desktop. One particular effect is that apps designed to help you learn things of a slightly repetitive nature suddenly become more compelling. My first experience of this was with an app designed to teach different bird songs, with the result that you may now well spot me suddenly stop as I walk across campus and listen attentively.

Apps for helping with learning foreign languages are another thing that, in theory at least, should work so much better on the iPhone than on a more conventional computer. Unfortunately, the choice isn't as good as it looks at first glance when you see the number of apps available - none of them seem to be quite perfect.

The languages I was interested in were Chinese and French. Chinese vocabulary is horribly difficult to learn and mine is embarrassingly bad considering how many hours I spend learning the language. On the otherhand my French vocabulary is considerably better, so they made a good comparison. Chinese is also interesting because you can represent words as either pinyin or as characters.

So should you be thinking of designing an iPhone app for learning vocabulary, here are some things I found that you may wish to consider:

1) First of all you have a choice of multiple-choice, flashcards (where you tell the app if you got it right or wrong) or free text-entry. These all have their pros and cons and I'm not sure one is strictly better or worse than any of the others, but they are very different to use. If you do use flashcards, make sure it's clear how you turn them over. I had to spend a surprisingly amount of time tapping, double-tapping and swiping the screen for various flash card apps until I hit the magic combination needed. 

2) Don't try and use photos or pictures instead of English words. It's a nice idea trying to associate the foreign words with the concepts directly, but in practice it just becomes a really frustrating guessing game trying to figure out what the pictures represent.

3) Let the user pick the direction to go in e.g. from audio to english, english to pinyin etc. Additionally, make sure that if the user doesn't want to see one version, they don't have to. For example, if you are trying to learn Chinese characters, you don't want the pinyin displayed as well. Likewise, if you want to recognise the pronunciation of a word, you don't want to see it written down too.

4) If you are going to use leaderboards, don't make the user type in their name every single time they get a new score on it, and also, show where they have come on the leaderboard if they get onto it rather than just telling them that they are on the leaderboard.

5)  The vocabulary lists are important. I think ChinesePod/FrenchPod have the nicest solution to this, where you add vocabulary from lessons on their website which then syncs with your iPhone. For me, this was much more of an issue with French than Chinese, because my French vocabulary is so much better, and I wanted vocabulary that I didn't already know. 

6) There needs to be a way to remove a vocabulary item once you are happy that you know it well enough, and a way for me to gradually add vocabulary as I feel like I want to add in a few more words as I feel ready to cope with slightly more.

7) If you are doing a demo version, please don't use the months or days of the week, pretty please? I know you are just showing how the app works, but it's incredibly uninspiring and less likely to get me hooked as I won't have felt like I have learned anything. This is particulary bad for Chinese where the days of the week are numbered from one to seven and the months of the year from one to twelve.

8) Don't assume that because I have got a word right once that I never need to be tested on it again. It is also more fun if words are chosen randomly from a bank of words than if the words are just a shuffled version of the list. The latter also has the problem that you find yourself getting the right answer by a process of elimination rather than by actually knowing the meaning of the word.

Overall my favourite apps were the ChinesePod and FrenchPod ones from Praxis Language, although they could still be improved. However, they do require a premium subscription to their site to use it. If I were more serious about learning either language I'd jump at like a shot, but it is probably a bit too expensive to justify for my more casual learning attempts.


January 2010

Following on from my previous blog post about language learning, I want to mention lingt. This is a site for learning vocabulary, specifically at the moment Chinese vocabulary although I believeI remember reading that they are planning to expand to other languages.

I'm very much liking the site at the moment. I know it's not very sexy to talk about things like rote learning but for me earning vocabulary was the part I found difficult when I was studying Chinese, and I really wish this site had existed then. I'm doing a five minute stint on the site over breakfast in the mornings and seem to be keeping it up. It doesn't tell me how many days I have logged on in total, but I managed 13 consecutive days in one stretch at one point. I am also level 8 whatever that means. It's interesting because it very much addresses the long-term nature of learning rather than just the short-term, figuring out when you are most likely to have forgotten a word in order to test you on it again. I have never properly looked at the Supermemo software but I'd be curious how much they have in common.

I am of course a games player and have to admit that the aspects of the site that attempt to be game-like don't really work for. It's certainly really interesting to think about how one might improve that part of the site. If you were going to make a site like this into a game, how would you do it?

Nonetheless the actual testing has been executed very well avoiding the classic pitfalls I talked about in my previous blog post, and I think it's one of the most interesting learning-related sites that I have seen in maybe the last year.