Saturday, August 29, 2009

inference is better than instruction

A Science Daily article from March 2007, Kids Learn Words Best By Working Out Meaning, states that the kids observed in the word-learning experiment could acquire new words better by watching, observing, and guessing on their own than by being told directly the names of new objects.

I wonder who will be the first commenter to tell me, "adults can't learn like children." The subjects of the research were toddlers, 36 to 42 months old. If adults can't do what children can, then it would be a pretty sad world. You'd have a hard time convincing me that adults cannot use the same innate strategies.

Take for instance, the situation where you meet a new person. You meet them, they tell you their name, and you can't recall their name less than a minute later. Why? Because there was no connection made. Even when we try to put that name into memory it often is impossible to recall. And when that person comes from another country that uses a language completely different from our own, it is even harder to remember their name.

If the situation were different and you were not told a certain person's name, but those around you kept talking about that person, your curiosity would grow. First you might not even know that the word is a person's name. Once you realize that it is, you start to wonder who it could be. Every time you hear the name, you try to connect it to what you heard before. You make some guesses. You know the name before you even know who the person is! When you finally figure out who it is, the connections are so strong and so familiar.

Now what adult couldn't do that? Hopefully you can see how the situation creates a powerful connection that wouldn't be easy to disconnect. Even though after awhile, I can't recall the names of people that I used to know, I still know exactly who the person is after someone mentions that person's name. I think the same goes for vocabulary acquisition in foreign languages. Of course, if the connections to the words are weak to begin with, then it's possible to not remember the meaning of words that you haven't heard in a long time. Even though you thought you knew those words, you didn't.

If you study a word beforehand, or look it up in the dictionary right away, what you get is a weak connection that requires you to go through your native language which makes the connection even weaker. Since you've got that weak connection already, there's no wondering, no curiosity, and no need to grow. So don't expect that you can make up for taking shortcuts. In fact, a dictionary shortcut is the long way around.

Say for instance, you are doing some reading in your target language and you come across a word you don't know and you can't figure it out from the limited context you are given. If you look the word up, what you are doing is going through a process. Every time you do that process, you build a habit. Even if you do it in your mind by recalling what you found in the dictionary when you looked the word up, you're repeating the process. Perhaps you will get good at that process and it will become automatic, just a split second for you to recall the meaning of the word. Even so, that is still far slower than native speed. That is the reason why most language students will never reach the ability of natives.

Most language learners will find this limitation to be acceptable. As long as they are speaking the language fluently, they will be happy. That is why they will not aim for higher standards. Others believe it is impossible or that trying to attain native fluency is unrealistic or would take too long. I find that these people have pretty strong opinions or a strong reluctance to take an honest look at what I write about. If they want to achieve lower standards or whatever, it is ok with me. I'm not writing to try to convince everyone to change their ways.

This blog is more of a documentation process. Some day, when I've finished learning Chinese, somebody will ask me how I did it. I am keeping a detailed record here for everyone to see. Then when I go on to learn another language, I will use only what I think works best and then I will hopefully be able to have a language learning experience that is not filled with trial and error.


  1. Strangely I have always thought this in a very non explicit way, but having it stated like this and thinking about it, this is brilliant (though the example using the name of somebody is rather weak).

    In the end I am starting to think that looking up words that you don't know immediately in the dictionary is actually bad practice. Having the words time to (let's say) fester in your mind WILL make the words stick better and the connections between word and meaning stronger

  2. I like your writing, Keith. You make good points. I just wanted to ask if what you write is backed up by your own experience or if it is just theoretical stuff you subscribe to. By the way, I agree with much of your reasonning in general.

    More on the topic at hand. Keith, you say one shouldn't look up the word whose meaning one can't guess from the limited context, but you don't say what the learner *should* do in this situation.

  3. I've long held a policy of no dictionary usage, mostly because using a dictionary meant that I'd be looking up a word in my native language. However, I'm now proficient enough in my second language that I can easily use a L2->L2 dictionary. Rececntly I've been using it to look up words that are very difficult to define, even by rich context, and it has been a rather large help. I've noticed that unlike using a L2->L1 dictionary, that the L2->L2 dictionary has no lingering translation; the meaning sticks rather than a word. I'm interested in what everyone elses L2->L2 dictionary experiences are.

    I've also noticed that as a word gets better defined in my mind through experiences, the more likely it is to find itself a mate in my native language. If someone were to ask me what a word meant in my second language I could easily give them my native languages equivalent even though I had never actually learned the equivalency through a dictionary or instruction.

    This has got me thinking that if one were to learn a word via translation that no actual harm would be done, so long as they had acquired enough experiences to firmly place the word in the semantic web. But that begs the question, when exactly has a word gained enough experiences to be safe to use? And I think that is the exact reason why using translations is seen as harmful, because you can't know exactly when a word is ready (this is ignoring the fact that translations don't often accurately respresent a words true meaning, which is harmful in its own right). On the flip side, by only learning through context the word won't be available until it has had enough experiences.

  4. I only have positive experiences with monolingual dictionaries. I can't recommend using them highly enough for intermediate to advanced learners.

    You have very valid observations, Spirit.

  5. This is why someone can ask me how to say something in Portuguese five times, and I tell them how, adding "I DID tell you this yesterday." And then they see some other word a few times in a movie and somehow remember it. I believe that by hearing the word enough without knowing the meaning(in meaningful context), you will burn the word into your memory by the time you eventually learn the meaning.

    That said, I would say that bilingual dictionaries mainly hurt learners because they are unable to make strong enough connections between an L2 word and its meaning. Instead, they have a shaky connection (as you said) will a definition in L1.

    This is why I would remember some flashcards over others back when I was learning Spanish. I seemed to know two words equally well when I went through the cards, but the next day I was unable to remember one. It seemed like they both stuck equally well the previous day, but one was deeper set in memory than the other because of some prior memory or a mneumonic(give that crap up). That just shows that even if you think you remember a word by translating it, it can never beat a more meaningful memory.

    Now, as all of the initiated know, FC's are not ideal for learning, but I'm sure you get my meaning. :P

  6. @Alexei: While watching Chinese TV dramas, I have enjoyed learning new words solely from the content I watch and I have found those words to feel real to me in comparison to words that I have studied which never feel real and always require an internal checking process.

    What you should do when you come across an unknown word depends upon your situation. In an ideal situation, you can just ignore it and it will keep returning and then one day the meaning becomes clear and you will recall some of those other situations and you will wait for some more situations and you will know without a doubt what it means. This is my recommended strategy for the first 80% of vocabulary. Once you can understand 80% of words like a native, then (I think) you will be able to ask natives what a word means and understand the response all in the target language just like a native. One of the reasons I say this is because the least frequent words are not frequent enough to learn from context unless you are lucky enough to come across the right context.

  7. @Keith: You are absolutely right about the least frequent words. The less frequent a word is the less likely you are to come across it, the more difficult it is to deduce its meaning, so you probably will have to use an aid - look up its definition in a monolingual dictionary or ask a native about its meaning.

  8. Keith, While I can see your point (and I now have a Thai TV habit due to reading your site), I'm sort of addicted to the little details of learning languages. That, and all of the resources available: Dictionaries, grammar books, course books, audio, web sites, etc.

    Some people collect antiques (and I do a bit of that too). I collect scads of Thai learning materials. So there is no two ways about it, I like stuff.

    My teacher is quite polite, or I'm sure she'd say that I have a fixation when it comes to the little details of learning Thai.

    On the subject of my Thai teacher... today in class we started talking about her years of experience teaching at AUA.

    I did not realise that her direct working relationship with Dr Brown spanned ten years, or that she actually taught the new AUA way (one or two teachers in front of a silent class).

    I knew she worked there full-time for twelve years before heading out on her own, but I also didn't put her and Dr Brown's theories in close proximity. Until today.

    I asked if she'd be willing to answer questions about AUA as well as about Dr Brown, and she agreed.

    So if you wanted to send a range over, I'll type out her answers and send them back (she keeps far away from computers).

    If she can't answer some of the questions, she said she knows others who worked with Dr Brown before he died and she could try them too.

    Note: I don't know how detailed her answers will be... but I can always ask David Long too if you like.

  9. omg! tried this method out, i know a bit of japanese, but i thought why not try it for a bit, it was soo fun!

    i "learnt" 2 new words, well that i could make out i know it's not alot and i don't even know if there right or not, anyway i usually listen to stuff in the background but this makes SO much difference

    i THINK i learnt

    salt and pepper (shiokoshou)
    i've also heard "this programme is sponsored by... so many times i know what it means

    i NOT sure if my inferences are correct because to be honest i've been to scared to look them up because they make sense to me.

    your blogs make so much sense also.
    your a genius!



No profanity. Please be considerate of others. Thank you.