Well now I have to talk to everyone I know who natively speaks a non-English language and be all "do this do this".
As a teacher and tutor, I'd imagine that writing the book could be as much a learning experience as reading the book, and I'm half-tempted to try this.
AP Latin classes do something similar by rote-translating the poems of Ovid, Catullus, Virgil, Cicero, Horace. The metered style simplifies the identification of matched phrases in samples, and would be a useful property if one were to similarly craft a book merging a native work and a translation.
Edit: Graphic novel would also be a good medium, as in Scott's example, since in addition to separating phrases, there is also visual context given for each phrase. Actually, I had a few friends in high school who learned Japanese largely for—and possibly by—translating imported graphic novels.
Edited at 2012-11-14 08:47 am (UTC)
There was an old, old school Korean textbook that did this, kind of. There would be the Korean phrase, the "proper" English translation, and then a literal translation of the Korean. An example (please excuse my awful Korean):
어닐애 밥을 먹었어요.
Today I ate rice.
Today rice ate.
In addition to lessons about things like particles (애,을) or past tense verb conjugation and how to use them/make them.
My friend who speaks pretty good Korean was a big fan of this particular textbook. As well as rote memorization: people piss all over it but sometimes it's inevitable, as well as being the most efficient method.
Yeah, I'm a big fan of rote memorization as effective, I just wish there was some other effective way.
I shared this on Facebook. The nice lady I tagged is a senior at UCSCU studying computational linguistics. You should go read her comment there
It worked for watership down in a small way. I'm also trying to imagine if there's any automated way of doing this; say having a bilingual news site, and reading the news in Japanese, but whenever you hover your mouse you get the English translation, so you can always just read in English, but it goes faster the more common Japanese words you absorb?
I was thinking about Watership Down, too.
I think it could work to introduce words in the new language fairly early.
I actually like your idea of the footnote. It's reminiscent of the way Magic: The Gathering teaches players the new keywords: in the first set where a keyword is used, there is "reminder text" on all the common cards containing that keyword, in parens and italics. Then, in the next set, the reminder text is gone, and you just see the keyword on its own.
A simple way to do this that wouldn't be obtrusive to the reader who -doesn't- want the flow interrupted by reminder text would be to have the novel be in hypertext format, using the anchor tag
's "title" property.
I had an idea for a game that would teach you Japanese. It was basically a noncombat Metroidvania, and at the beginning, your only ability was "Sore wa nan desu ka?" "What is this?" So you'd only be able to ask people about things that were close to your character.
With that ability, you could find out words in Japanese, and use them to get past various people. Later you'd get "Kore wa nan desu ka?" "What is that?" so you'd be able to ask things that were close to other people. The game would go on like that. You'd learn blue, by getting the power iro and asking (I think, I've forgotten a lot of Japanese) "Sora wa nan iro desu ka?" "What color is the sky?" and an NPC'd say "Sora wa aoi desu." "The sky is blue."
Eventually, you'd get quest to do things in Japanese. Buy things at the store. Read a kid's book or manga. Go on a date. There could be progress bars that kept track of how many words you knew. The game would detect when you forgot words, too, when you'd have find ways to get them, and it would make sure to bring up words you had forgotten in quests. Every so many words, you'd gain a level, the levels wouldn't be numbered, but named. (I can't be the only one who gets progress in those types of RPGs just to find the names of things.)
The end goal would ideally be something legitimately fun and game like as an end goal. Become boss of the Yakuza, or an astronaut in the Japanese Space Program, work for a game or animation company, score with a Japanese girlfriend. I'm not picky.
The point should never be to take the memorization away. That's impossible. Memorization isn't fun in and of itself. But it's not actually boring either. We memorize things every day. The way home from work. A new friends name and interests. Faces. A conversation with your girlfriend. Every. Single. Day.
It just needs to be gamified. It needs to be treated as a worthy challenge with progress, and each word is a secret power that can be combined with other words to make a magic spell. Giving people powerful ways to combine words will reinforce memories, and make each word that much more memorable. And the spell book is already there, but instead of being called a GRIMOIRE it's spelled DICTIONARY.
Edited at 2012-11-14 09:21 am (UTC)
It didn't work for me. I'm not sure how much it was that I needed to make the test larger and how much it was that gender essentialism makes me crazy.
Have you read Watership Down? it famously does almost exactly this, with an entirely made up language (only rather slower, and culminating in only a single sentence not a whole chapter).
I think it could work quite well, if the book was sufficiently interesting. And I'd like to see it done. The thing I don't think you could get from it is the correct way of pronouncing things (languages which share the Latin alphabet do not all share the same orthography).
The only way I've ever got beyond "hello, my name is naath; do you know where the train station is" sort of conversations is by spending time in a country where the language is spoken and *avoiding English*.
Personally I find that speaking two natural languages gives a lot less insight and expressive power than knowing a programming language does. I've had many enlightening and fruitful conversations in Finnish with Finnish-speaking people, but it's hard to imagine I would have missed a lot of the same insights only speaking English with English-speaking people. But if I hadn't learned Haskell, I *would* have missed many conversations with GHCi. I have had my assumptions challenged, I have untangled my understanding of concepts like identity and change, I've explored properties of different constructs and algorithms, all by asking GHCi questions.
There's no illusion of transparency when your counterpart is a compiler. You either specify what you mean precisely, or the program doesn't work. Well, okay, there's a kind of an illusion - that being that even simple-looking programs are incredibly likely to be wrong (aka buggy), but in ways that are invisible until you find the particular test that exposes them. You need precision and humility. You learn to make deep models and then poke holes in them. Now I think you're already pretty damn strong at those things... but I think that learning a programming language would do more to increase your power than a natural language would.
To save you a little possible future effort if you do make the strategic decision to put effort into learning programming, I'm going to throw in a few recommendations. Learn Python the Hard Way
might be the best starting point (operative word being "might" because I don't think there's been research on this). It's really
big on the precision and humility. As in, you're going to hate it because it takes an approach that's even more boring than rote memorization and grammar rules, but, well, that's what it takes when you have to get every character right. Oh, and the nice thing about programming languages is, once you know two (from different paradigms), you'll be conversant in all of them. So in the even more distant future when you've gone through the really painful stuff, you could try Real World Haskell
. It's more conventional and more oriented toward people who want to actually be programmers rather than just people who know how to program, but it's one of the least stupid Haskell tutorials around, and then I'll finally be able to explain what that signature about naming things that I once used means.
Second that, Ilya Frank's method seems to be fairly close. It would also allow you to mix the two languages in whatever proportion you're comfortable with at the moment, without "breaking" the grammar and syntax of the language you already know (which is what squid314's method would arguably do).
However, I've tried to read some French this way, and found that it's kind of hard (for me at least) to read words you have no I idea how to pronounce. So some kind of interactive Frank-based book with IPA transcription and possibly recorded or TTS pronunciation would probably be the best.
Can't wait for speech recognition systems to become sufficiently advanced to be able to recognize pronunciation errors and correct them.
Edited at 2012-11-14 09:12 pm (UTC)
I'm now learning German. And doing a reasonable job despite putting no effort into the course. I think it's for an entirely different reason.
Namely, I know lots of German *songs*, thanks to classical music. Yesterday I learned the word "Huter" (imagine an umlaut there), which means protector, and I thought "Oh, right! Like "Erkenne mich, mein Huter" from the Passion of St. Matthew!" And that way I'll never forget what it means and I have the gender for free, and in general I have kind of the rough sense of what sounds right and what sounds wrong grammatically because I've heard it in a million songs. I wouldn't have to do your weird syntax game, because German syntax already sounds normal because that's the way all the songs go.
Sure, I had to do a *little* grammar memorization, but the haze of familiar references fills in the gaps, like a friendly musical exocortex. I'd probably be terrible at Japanese because I have no cultural references in Japanese -- I'd probably have to spend years watching subtitled anime without really trying to learn, and then I'd take a class and be all, "Oh, that verb was in that cute scene in Kenshin!"
I wrote a few sentences with English words and German syntax and my brain doesn't recognize them as that, but as English sentences with garbled syntax. In the reverse it's the same. So it seems that my knowledge of the syntax and words of a language are not independent. So i wonder if it makes sense to learn them independently.
So i do suspect that courses that teach grammar explicitly and words out of context are doing it wrong.
Still i think there is no way to make learning a language not hard work. But falling in love with someone who only speaks your target language, or falling in love with a subject which gets written about only in your target language, can make it feel like it isn't hard work. The latter is how i learned English without even seriously trying.
Shaul Mofaz (Antichrist) will be Israeli Prime Minister( and
After from earthquake in northern Israel and Syria (december 2012)- and become dry lake Kinneret/ govern -
period mofaz 427 day and will be killed In Jerusalem:march 2014
Saudi Arabia king will be dead and after
start War for get reign :november 2012
unprecedented Floods In many regions of the world in: september-.. 2012)-
will be eruption Yellowstone?
and will be send out ashes During the eight day
earthquake in East Mediterranean Sea - with hundreds thousands killed on :december 2012
will be occupy syria by Jordan army on
december 2012-February 2013
and will be war Between Turkey and Jordan in the north-
eastern syria april 2013
and after to leave american army in iraq Jordan army will
be occupy iraq on may-june
, some of Sufyani armys (Jordan) will be sink
into the ground./ This place will be known as Baidah and
will be located either between Makkah:
(and Madina-volcano (alaise
Will be killed pious man in Mecca on
2014- return Jesus perhaps on...2013.
will be coming to earth from the sky for to help Imam Mahdi propagation
freedom and justice and beneficence in the worldhttp://emam-mahdi-1431.blogfa.com
http://en.wikipedia.org/wiki/Interlinear_gloss seems relevant.
I vaguely remember reading about a language learning technique that's actually been used in the past that's somewhat similar. Rather than gradually transitioning the text from the learner's native lanugage to the language being taught, a novel (Felenon's Les aventures de Télémaque, I think) is presented side-by-side in its original language on one page and translated into English on the facing page. The idea being that an intermediate student who already knew the basic grammar constructs and a few dozen of the most common words, plus could make a guess at other words via cognates and loan words, could try to stumble through the French text, referring to the English text as needed.
The utility of this version seems like it'd be limited to improving an intermediate student's skill, and would be more successful at teaching a language that's closely related to one the student already knows. From personal experience, I've found it very effective at getting the knack of reading Middle English, but short of Scots, Frisian, or maybe Africaans, that's one of the easiest language-acquiring tasks for a native Modern English speaker to accomplish.
Your proposal, by improving the bootstrapping process by gradually introducing intermediate versions of the language, seems like it could be a considerable improvement.