To recap briefly (you ARE reading these in the correct order, aren’t you?) – I’d figured out that speaking was vital, I was trying to think of ways to inject alcohol into mp3 files, I was convinced that interval learning was central, and I was certain that grammar itself isn’t difficult – it’s our explanations of it that turn it into a nightmare.
You can probably hear a lot of certainty in that paragraph, can’t you?
Oh, I felt a lot of certainty, believe you me.
And I was so frustrated about everything that had niggled me in the Michel Thomas Spanish course, including the infuriatingly hesitant learners I was meant to be picking up the language from, it seemed that the obvious thing to do would be to build a much better Spanish course. In fact, it seemed so obvious that I got in touch with a friend of mine who was a very successful entrepreneur, explained what I thought I could do, explained that an mp3 course offered the possibility of massively undermining the absurdly high costs of traditional language courses, and generally enthused at him.
Pleasingly, he really liked the idea. He liked it enough to put his reputation on the line and talk to some venture capitalist friends of his, to ask them to fund a prototype. I can remember thinking that this business lark was an absolute piece of cake, and I couldn’t for the life of me work out why everyone else wasn’t starting their own business too.
And then he got back to me, and said ‘You’ve got to pivot.’ ‘Pivot?’ I said sternly. ‘I’m not trying to build a bloody gyroscope. What are you talking about?’
It turns out that pivoting is what start-ups do when they’ve got it completely wrong. It’s start-up language for ‘having a better idea’. His venture capitalist friends had said they’d rather invest in selling ice-creams to penguins, because the online language teaching space was already ridiculously competitive, and I didn’t stand a chance. They’d give better odds on the life expectancy of a snowflake settling down optimistically on Satan’s nose.
After sulking furiously for a while, I thought (and I apologise for the language) ‘Sod this for a game of soldiers. Pivot my arse. Call that advice? To hell with you, I’m bloody well going to build a language course anyway, and I’m going to build it for Welsh, and I don’t care if it doesn’t work, because I love Welsh enough to put the time in anyway, so up yours, you faceless venture capitalist idiots.’
Yes, yes, I know, you don’t have to point it out to me. I pivoted.
And then I told Catrin. Catrin was getting a little tired of my ‘this would be better than a job’ ideas, not least because we were living in a one room refurbished pig-sty on her sister’s small-holding and had to cross the yard to go to the bathroom, and unbeknownst to us the beautiful little stream that we rather loved was developing a cunning and heartless plan to flood us out.
Catrin shocked me. She didn’t roll her eyes. She didn’t sigh. She said ‘That is actually an extraordinarily good idea.’ I almost fainted. ‘Just one thing,’ she pointed out. ‘You’re going to need someone to do a southern version too. Someone as obsessive as you, who’ll give up ridiculous amounts of time just because they think Welsh is important.’
‘No problem,’ I said. ‘I’ll give Iestyn a call. He’s much more obsessive than I am. He drove from Maesycwmer to Penrhyndeudraeth on his birthday just to see if Cymuned was full of racist sociopaths or the kind of amiable lunatics he could get on with. He’ll do it.’
I was right, of course. After asking Catrin to marry me (SEVEN times, in a row, before she stopped laughing, put her drink down and asked if I was being serious), and after deciding to learn Welsh, asking Iestyn to set up a thoroughly quixotic company with me to teach Welsh was definitely the third best decision of my life. He’s one of the most impressive people in Wales, but even more important than that, he can actually deal with tax returns.
As you know by now, I love rambling on. But it’s getting late, I’m on my second pint of rum and coke, and I’ve got a dreadful case of the hiccups – so let’s pan the camera back a long way and look at the rest of the story as though it was a mountain range (yes, with hobbits and orcs running all over the place).
I wrote the first 15 lessons of SaySomethinginWelsh based on the ideas we’ve already talked about. Along the way, it struck me that you could build up to some of the more complex sentences in stages: ‘I wanted to say.’ ‘What I wanted to say.’ ‘I can’t remember.’ ‘I can’t remember what I wanted to say.’ That one turned out to be a keeper. Iestyn adapted the lessons to a southern dialect (in fact, I think he decided to try and revive Gwenhwyseg, an admirable aim), and we stuck them up on the internet. January, 2009, four days before my daughter was born – ideal timing, I’m sure you’ll agree.
Then I set up a forum, and told 20 friends on Facebook about it.
Then Iestyn went on holiday to Europe. FOR EIGHT MONTHS.
We kept producing more lessons, and Iestyn had to record them in a variety of implausible locations, and upload them principally via free Wi-Fi at a range of McDonalds. I felt little pity for him.
Oh, and the forum exploded. Figuratively at first, and then literally, and we had to change providers, which we did with the help of a programming expert from Birmingham who had a holiday home in Borth-y-Gest – not the kind of person you’d usually expect to be contributing his leisure time to the future of the Welsh language. Once he’d got us on a better host, he then wrote a piece of custom software to make it possible for us to create practice sessions automatically – which was probably the single most important part of turning the lessons into something people would pay a small monthly fee for. SaySomethinginWelsh would not have succeeded as it has without the love and support of an Englishman – and Spenny was just the beginning.
As the forum expanded, so did the screams of pain. People formed support groups to deal with the trauma of Lesson 6. I looked at it, realised how badly it failed to keep to the guidelines I had in mind, and rebuilt it into 6.1 and 6.2. People formed support groups to deal with the double trauma of 6.1 and 6.2. I looked at it again, and finally worked out what the real problem was – and it was a problem that would have required rewriting the entire course, so I decided to put it off for a while.
We finished Course 1. People pointed out that they could more or less speak Welsh, apart from the minor drawback of not really knowing any actual words. I wrote 10 vocabulary lessons, so that people could survive Bootcamp and say what day they wanted to do something. We finished, a year or two later, Course 2 – and then did 10 vocabulary lessons for Course 2, because I still felt guilty about people I’d seen struggling on Bootcamp.
Bootcamps, by the way, were weeks in Tresaith where English was banned. They had a life-changing impact on a great many of our learners, but even more importantly for the course, they showed me in glorious technicolour exactly what I’d forgotten or taught badly.
Work on Course 3 started, and we had a request from Maga, a Cornish version of something like the Welsh Language Board, to provide a similar course for Cornish – by this stage, we’d had a fair bit of press coverage, some snippets on the telly, and learners were saying all sorts of lovely things about us on the interwebs.
Maga even offered us money! Ho ho, we thought, money for old rope, we’ll get them to translate it and then they can record it and we can take our kids to the Bahamas for a year or two.
And so began what I think might just turn out to be my life’s work – because we were utterly wrong. We couldn’t just let them translate it. It was all over the shop – not only had three years of practical research shown me all sorts of things I’d done wrong to begin with, but Cornish just didn’t map to Welsh.
Puzzled, I put together a very basic spreadsheet-style algorithm to build a better course. I populated it. I wrote the course – 25 half-hour lessons, so several months of work. It was absolute, unmitigated rubbish. It wouldn’t have turned people into Welsh speakers, it would probably have made them want to join the much-lamented Welsh Mirror’s campaign of hatred against the National Eisteddfod.
With moronic determination, I turned to word frequency lists to build a better course. I populated them. I wrote the course – 25 half-hour lessons, so several months of work. Is this sounding familiar at all? Yes – it was absolute, unmitigated rubbish. It would have made people join UKIP.
Chastened, older, exhausted, and hoping to sleep again when the children left for university, I realised that there was no mathematical short-cut.
I couldn’t use an algorithm, I couldn’t use a spreadsheet, I couldn’t use frequency lists.
I needed to map out real conversations that learners would actually want to have – conversations that they would have when they were getting to know people, not ‘Can I have a loaf of bread and isn’t the weather lovely?’ conversations. Conversations that it would take a truly horrible amount of work to map out in the kind of detail that would be needed.
And that’s what I’ve been doing pretty much ever since.
We did some testing with initial lessons in Italian and Spanish and Cornish and something else I can’t remember, just to check some other ideas I’d had on the way – fascinatingly, people believe that sentences with ‘and’ and ‘but’ and ‘or’ in them are longer and more difficult to remember than sentences with multiple dependent clauses in them (like ‘I met someone in the pub last night who said that he wanted to talk to your sister in work next week’) – even when the multiple clause sentences are significantly longer than the sentences with and, but, or.
As I mapped out the conversations, and tested the extra ideas, I started to build them into a new framework – designed from the very beginning to be adaptable as easily as possible into multiple languages. It’s challenging – you have to be permanently aware that what can be inferred in one language might have to be introduced as an entirely new item in a different language, so you need huge amounts of redundancy to create space for adaptations – but based on the work we’ve done so far in adapting the new framework to Welsh, Cornish, Dutch and Spanish, it’s excitingly robust.
So, mistake after mistake after mistake, coupled with feedback from thousands of passionate learners, coupled with a willingness to write completely awful courses, throw them out, and start again – eventually gets you to the point where it no longer seems insane to believe that people can use the materials we’re developing to acquire genuinely impressive confidence in a new language in a matter of hours.
The new framework was the inspiration for the initial tests we’ve been doing in accelerated learning, of course.
And in the last few days, the SaySomethinginWelsh community has responded to my cries of self-pity about having a bad back (shameless of me while Iestyn’s wife Cat lies in bed unable to sit up because of her Cauda Equina Syndrome) by forming a working group of about a dozen people in a matter of hours to take over the tedious work of importing sound files into the software Spenny wrote for us.
Which means that we will have enough Welsh lessons based on the new framework for us to test what is possible in a two day intensive course in the very near future, and enough Spanish for the same test very shortly after that.
Yes, the new framework was the inspiration for the testing we’ve been doing on accelerated learning.
Are you working on something? Are you trying to build something? Make mistake after mistake after mistake, as many mistakes as humanly possible, and as fast as you can – and then get honest people to tell you what they hate about your work. Then keep on making more mistakes, different and better and bigger mistakes, until one day some of the honest people you’re torturing tell you that they don’t actually hate it all that much any more. At that point, you will have developed a very real level of expertise in what you’re doing, and you might just have a chance to change at least a little part of the world.
No, stuff that, I take that back. Go through that process, get back up every time you’re knocked down, and you’ll have a genuine puncher’s chance of changing the world.
Go on, then. What are you waiting for?
[Sorry, no pictures this time – that function doesn’t work on the tablet. Did you like the pictures as much as I did, or were they really just a waste of time?]