Slow Learning

Dilatory Autodidact TY SwedishLast week I finished a protracted slog through the pages and exercises of Teach Yourself: Swedish.

“That took a while,” I thought, as I marked the textbook completed on Goodreads. (Yeah okay, logging textbooks on Goodreads, you got me. TBH if they were printed with an ISBN I’d probably have a Goodreads shelf of takeaway menus.)

In fact it took more than “a while”. The start date was March 2012. Back when we lived in Amsterdam. 3.5 apartment moves and five years ago.

FIVE FRIGGING YEARS. (Nothing against the book, it’s fine. Serviceable. Okay.)

I’ve always been a sucker for the stoic will-to-mastery of autodidacticism. Or rather for the *idea* of it.

Self-motivated, me-powered learning! It’s a daydream, one of those poisonous momentum-sapping ones where the fantasy gives a watered-down version of the satisfaction you’d get from the actual achievement, hobbling any real progress. Because that’s the only languge exercise book I think I’ve ever finished.

And language apps are handy, since they’re always pinging at you from your handset. Addictive when they get the gamification right. IMO yer Memrises and Duolingos are lacking, pedagogically speaking: TBH I’ve found them best for building vocabulary,  but the grammar doesn’t stick.

When it comes to educating humans, nothing beats another human.

Deploying that slowly-absorbed book-learning, that app-jacked rote learning, in conversations with Real People. Or signing up for lessons, group or one-on-one, where it’s harder to skim over exercises, or worse, let them slide. For, say, five years.

In truth, it’s a mix of inputs that is best. Practice in the wild is always critical, courses are great. TV, radio, films, books and podcasts too, though they’re best absorbed with some attention, even if intermittent.

The apps and lonely exercise books of the self-learner have their place too, whether it’s by providing new vocabulary and grammar to test out, or explaining the why behind language you might already be using.

But ivory tower learning in solitude just makes the whole process needlessly hard. It’s like training for MMA with shadow boxing only – how much training would you feel was enough before you ever stepped into a cage?

 

Advertisements

Google Translate and Plato

google-translate-ai-2016-11-24-01

So you’ve seen this news [New Scientist/Wired], right? That the Google Translate AI has supposedly invented a new internal language to help it translate language pairs it hasn’t learnt. Having been taught English <> Japanese and English <> Korean, it can then do the job for Japanese <> Korean.

The headlines position this as the AI inventing its own internal language, or interlingua, to handle those conversions. Every article notes the difficulty in anyone knowing exactly what is happening in there, the deep learning that’s going on inside the AI. (Engadget had the slightly more nuanced report on this, further from the “invented a language” headline.)

An invented language? That’s one interpretation. But if a language consists of signs, symbols that exist in the world, is that the best description for the process?

So Google Translate learns that English table means both Spanish mesa and Swedish bord. Does it then need to tell itself:

IF table = 1010101011101 = bord
AND table = 1010101011101 = mesa
THEN bord = 1010101011101 = mesa

?

That’s not how we meat-sacks use language. It skips over another interpretation, lacking from the reporting I’ve seen so far, which is either totally thrilling or utterly chilling, depending on whether or not you’re looking forward to the ascendancy of Skynet.

I (and I’m willing to assume you too) have an idea of a table, based on years of experience:

  • It’s a flat surface atop a number of legs (often 4);
  • It’s usually (not always) around thigh height;
  • Most are made of wood, or metal, or plastic;

All these things contribute to a mental representation of [table]: a confluence of images, physical experiences, language labels, and a heap of Venn diagrams of different properties that coalesce around the label of table. A lot of overlap with something like [chair]; less – but some – overlap with [dog]. A mess of connections and firings in the neural pathways, impossible to pin down, even while it’s possible to see where they cluster.

The Platonic form of a table, if you like. That is what’s triggered when I hear the English or Spanish or Whateverish word for [table]. English is my mother tongue, but if I were to translate a Stockholm restaurant reservation for a Spanish speaker, the mental process wouldn’t be “(Swedish) bord = (English) table = (Spanish) mesa”. It would be “bord <> [idea or Platonic form of the table] <> mesa”.

Look at that map up top again. That string of Japanese word/concepts on side, the English string on the other, the Korean somewhere in the middle but tending closer to the Japanese. The whole forming a neat oval, a cluster of meaning. OK, so the AI only has language input, there are no sights/feelings/memories of [stratosphere] associated, not yet.

But what if that oval, that pattern of pinpricks of understanding, represents the rough formation of a Platonic form; an AI idea?

But Books

We all remember our first mp3 player. The giddy joy of having days, months, years-worth of music at our fingertips. The delight at being able to ditch those overpriced CDs. The acceptable trade-off with sound quality.

As a music fan, that trade-off didn’t bother me much. The pros of convenience and accessibility far outweighed a dip in sound quality that was barely noticeable to anyone except audiophiles. Like I say: music fan, not stereo equipment fan. Yes, vinyl sounds warm. It’s lovely. It’s still not as good as live music. And it’s 100% non-portable.

Likewise with movies and TV. Seeing a film in the cinema is one thing. Watching a DVD is… well, who misses navigating through piracy warnings and over-elaborate DVD menus, anyways? Why not fire up iTunes or Netflix or Hulu and have a leisurely browse to see what takes your fancy at any particular fickle moment.

But books. Hmm. The original physical entertainment medium. The smell of the pages, the crack of the spine, the feel of the paper under your fingers. And perhaps more importantly than anything, the look of all those spines stuffed to bursting on the bookshelf.

It was obvious that this couldn’t last. That books would go the way of music and movies. It’s strange, in fact, that mp3s and DVDRips took off before e-books. Shouldn’t written text be the *easiest* thing to reproduce?

Though I’d had no problem with the death (ok, mortal wounding) of physical recorded music (what a weird idea that is anyways), I expected to have more of a problem switching from paper books to phantom illusory binary code in the ether e-books.

Or as I’ll refer to them from hereon, “books”. It’s not like hardback books and paperbacks had to be permanently welded to their individuating designations now, is it?

Because what isn’t to love about a book that you can read on a tablet, then pick up on your phone, when you’re stuck at the town hall? Isn’t it great that while reading Vlad Nabokov or dear Billy Self, you can highlight that wonderfully erudite and oh-so-fucking-abstruse word and look up what it means? That might be my favourite thing about (e)books – it’s so much less jarring that having to mark your page, pick up a dictionary, leaf through to the right word, get distracted by a rude word, put the dictionary down, pick up your book, drop your bookmark, swear, find the last page you definitely remember having read… you get the point.

Highlighting text, multiple bookmarks, searching for phrases or character names: all these things are now ridiculously easy and all-but-idiot-proof. Highlighting in particular – when I do that to physical books, I’m aware of how irreversible the process is, how I’m intruding on the reading experience for anyone else that picks it up later.

Yes, there are still questions to be ironed out. The expense of an e-reader in the first place. DRM, and the inability to lend of store-bought books. The business models of Amazon and Apple, and the ramifications for authors. But books are books, whether they’re burnt onto bark or broken down into bits and bytes.

Thanks for the Memrise

“Made it Ma! Top of the world!”

I’m always on the lookout for good language learning tools. And Memrise is my latest *essential* bit of free, online kit. Memrise provides a system for learning by spaced repetition. Essentially you’re asked to translate a term, either from or to your native tongue, and the system waits until it asks you the question again – it could be hours, days, or weeks, depending on how quickly and accurately you answered.

The technique of spaced repetition has been around since the 1930’s, but it seems to have benefited from the app boom of the last few years. There are scads of online and iOS/Android apps that make use of the concept: Anki being another personal favourite.

Memrise, like Anki, just provides the software. The courses themselves – Basic Swedish, Catalan Common Verbs, 15 Giants of Chinese History – are created by users. Anyone can create a list of foreign words, medical terms, or historical facts, that can be plugged in and learnt using spaced repetition.

Memrise uses the idea of greenhouses and gardens – for new shoots (short term memory) and flowering plants (long term) respectively. So far, so slightly twee. You plant new seeds, you water plants before they wither and die, you maintain your gardens, and all the while you clock up points. Utterly worthless, abstract points.

The smartest part of Memrise, though, is its Community. It actually took me a few weeks to stumble across this. This is where the gamification comes in, and it’s probably why Memrise ended up a winner at the 2010 Seedcamp startup event in London.

Via the Community tab you connect to Mempals, and give them High Fives or Thumbs Up. Most importantly, you can see how you’re doing on the leaderboards. This, for me, is where the app went from being fun to addictive. Now I wasn’t just learning, I was trying to rack up enough points to overtake my (new) arch-nemesis, noelmuller. Noel, of course, knows nothing about this. But for a few days I was desperately running through exercises in an attempt to draw ahead of him.

And I did! I made it, all the way to number 1 on the leaderboard! Well, not the *Everyone* leaderboard, just the Cohort version that only shows people who signed up at the same time. But still, I felt a smarmy sense of achievement. Come the next week, and noelmuller (*SHAKES FIST*) had overtaken me again, and now he’s about a hundred thousand points off in the distance, but hey, y’know, I had my moment.

The site has a few small drawbacks. Since the courses are created by users there are often a few errors, but those are generally corrected, sooner or later, and you can always leave a note that there’s an error on a specific phrase. There are user-submitted mnemonics, which are meant to help lodge words in the brain, but I don’t find those incredibly helpful. Perhaps it’s better to come up with your own mnemonics, since the associations need to be personally obvious to be helpful. I also prefer associating physical gestures to help me fix a word in my memory. Yes, it looks dumb, sweeping your arm around the room as you enunciate rummet, but it helps.

Morning Pages and Grand Theft Auto

Grand Theft Auto III for iPad

Grand Theft Auto is a stoopidly entertaining game. You know the drill, progress from smalltime lowlife to city-straddling crime kingpin via a series of improbable car chases. The port of GTA3 to iPad is a real achievement, bringing a “full console” experience to a portable format. In a session I’ll typically complete one or two small missions, crash a bunch of cars, and spend half the time running cluelessly around the city looking for health and armour restores. At the end, I’ll quit the game, check the time, and think “f*ck me, was that really two hours?”

Fun as it is, GTA is a massive time sink. This, I had to remind myself, is why I never bought an X-Box or PS3 – because I would lose whole weekends to the damn thing.

My big problem – and I know I’m not alone here – is productivity. Actually getting the things done that I say I’m going to. The blogpost, that short story rewrite, the legendary screenplay (“think Goodfellas meets The Princess Bride with shades of Hable Con Ella”).

It’s clear I’m not writing enough. How to fix that? The simple answer would be stay up late/get up early. I’m working on those two. Honest.

I love tools that make me feel more productive. Initially I thought 750 Words might be another productivity-boosting tool.

The idea – supposedly inspired by something in “The Artist’s Way”, though it’s a common enough concept – is to produce a burst of untrammeled automatic writing in the morning, before doing anything else. The 750 Words site gives you a place to do that, and software to log those words. You can also track variables like hours slept, mood, whether you got laid the night before… whatever you feel the need to track.

The point of this exercise is that it’s a massive clear out, a total brain dump that clears the mind of extraneous worries so that it can focus on the creative task at hand.

Now this is fine if you’re then going to spend another 8 hours writing. I was doing something similar, on a smaller handwritten scale. But what if you’re struggling to carve out enough time as it is?

750 Words tracks you day by day. It knows when you’ve skipped a turn. Then you feel guilty, because not only are you not keeping up with your writing, you’re also not keeping up with your pre-writing exercises.

There are sites that use gamification well, such as Memrise. There I’m keen to collect points and creep up a leader board – there, it’s all in service of learning, i.e. the work at hand.

With 750 Words I felt like it was just more time I should have spent doing something more constructive. Maybe I’ll go back when I’m struggling with my second novel. But for now, no.

So I deleted my 750 Words account. I uninstalled Grand Theft Auto. Instead, I rebooted my blog. Writing with a sense of purpose. FAR more productive.