Perhaps you've heard of learning Norwegian in 3 months?

Or maybe, learning Norwegian in 10 minutes a day?

Well, how about learning Norwegian in 150 seconds?

First up, I should explain that I don't actually speak Norwegian yet. I'm learning, but I'm a total beginner. I wrote this program to help me learn, by thinking about what helped me learn other skills.

I started trying to learn Norwegian with a program called Anki. It's a good program, used by many people around the world. But it has 5 'problems', in my view.

  • You are distracted from learning by continual questions about 'how well you're learning'. You have to press different keys to indicate how easily you learned a word, so that it can show you the words you had most trouble with more often. But this can be an annoying distraction - and what if those words weren't very useful words anyway?

  • The interface is distracting - there are lots of icons and menus, and words appearing that aren't in the target language.

  • It tells you when you've done 'enough' and makes it a little inconvenient to continue. Why discourage people? I think it's better to 'addict' them to learning.

  • Some of the people who designed the flashcard sets for Norwegian Bokmål may not have had a coherent plan for their choice of target vocabulary.

  • Sometimes you will get a card showing 1 English word mapping to 3 possible Norwegian words. e.g. conjugations of a verb. This is hard to remember. I prefer the idea of 'flattening' a language as much as possible - i.e. you pretend that all the conjugations are different unrelated words. After all, who teaches 3 year olds how to conjugate verbs? Once you know enough words, your brain will figure out the rules for verb conjugation anyway.

This program provides an alternative approach to learning the language. The underlying principles are:

  • Focus on the words that matter most, to allow you to start learning from natural texts early. One of my friends teaches in a Finnish high school. Apparently, they found that foreign students are able to integrate well in normal classes once they know about 300 words of Finnish - that's all! According to this paper, 2000 'useful' high-frequency words will get you around 90-95% coverage of an English book's vocabulary. Considering that English has over 250,000 words, that's quite interesting. So this program makes you practice a small set of extremely useful Norwegian words repeatedly until you know them without thinking. However, learning the same thing many times is boring, so we need...

  • Gamification. Flashcards are work. Games are fun. Here, the goal is to get every word right, every time. React immediately, over and over, without time to think or take a break. If you make one mistake, back to the very beginning you go, as 'punishment'. I find this type of challenge works very effectively to teach motor coordination skills, and I think it can work for language skills too. For example, watch this: BIT.TRIP Runner or try playing this: Marvin Spectrum. You can feel yourself getting better as you play.

  • No distractions while learning. Simple interface. Few options. No menus. Once you start playing the game, there should be literally nothing on the screen except the game elements.

  • Use other senses, besides vision. Here, sounds are used. The sounds should help your brain 'play the game', by giving timing clues, and inducing 'trance-like' learning, e.g. flow.

  • No excuses for not trying. It must be easy to install or run. Just go to the page. No choice of flashcard sets, no downloads, no updates, no settings or preferences. Just make the most obvious options the default settings. This removes immediate barriers to learning.

  • Reading requires mapping words to concepts very very quickly. Therefore, practice mapping words to concepts very very quickly.

  • Measuring progress should be 'obvious', by which I mean 'monotonic'. A score of 19 should always be better than a score of 18. This implies presenting the words in a fixed order (here, the order is 'how frequently the word appears in Norwegian newspapers'). This also means that the learner gains the most value in terms of % language coverage at an early stage.

  • No clues about pronunciation. The focus is on reading/writing, not speaking/listening.

  • With written language, knowing the grammar or some idioms won't save you if you don't know the vocabulary (this is different to speaking, which can be very idiomatic). But knowing the vocabulary might let you guess the meaning of a sentence, from which you can deduce the grammar and idioms. So: let's focus on vocabulary.

  • To prevent people learning 'the sequence' (en, to, tre) rather than 'the mapping' (one->en, two->to, three->tre), there is an option to skip words occasionally at random.

  • Sliders and checkboxes in UIs guide users and prevent them from accidentally configuring invalid settings.

But does it work? Write to me (website@thisdomainname) and tell me what you think.

Update (Aug 2012): After playing it myself for a while, my feelings are that it is somewhat useful but not exactly super-fun. The big problem is that the user needs to do the work of recognising when they made a mistake. That's not fun. I may create a second version which prompts a user to pick an appropriate response from a selection. Also it perhaps needs some zombies or ninjas - something like the games 'Word Zombies' or 'Zombies vs Literacy'. Time to learn how to draw zombies. :-)