Hacker News Reader

Comment by wallflower | Hacker News Reader

When you are learning a new language and in the beginner to intermediate chasm, reading in your native language is like speed reading compared to reading in your new language. It depends on the context of what you are reading, of course but it can be 10x - 100x slower.

However, many polyglots write in most articles about reading in a new language about one of two approaches:

1) Reading and understanding everything (to the point of dictionary lookups ad nauseam) 2) Skimming the text as fast as possible to build your pattern /phrase recognition abilities.

I wonder what applying speed reading techniques to reading a new language would do. Would it help you "skim" faster? Skimming the text of the new language as fast as possible is still pretty slow. Does anyone have experience with reading quickly/reading a lot to build their new language ability?

View on HN

Based on my experience and on my bias for what language acquisition researchers I favour ( :-) ), I don't think #2 will work as you described.

The key to language acquisition (acquisition giving you the ability to use the language without undue thought or planning) is repeated exposure in a comprehensible way. So you would think that skimming would be ideal because you can be exposed to the patterns more quickly.

The problem is that research has shown that you require 95% comprehension before you can infer the meaning of unknown language based on context. So unless you are already very familiar with the vast majority constructs and vocabulary, you will not be able to infer anything new and you will not learn anything.

My opinion is that native readers can skim because they already are very familiar with the structures and can guess what's going to be said. This allows them to skip over most of the sentence and still pick out the meaning. The less familiar you are with the structures, the less success you will have (and research shows that the drop off is very dramatic).

On the plus side, the average novel has thousands of unique vocabulary and virtually all commonly used grammar. If you work through a single book, you will cover a huge chunk of the language. And since authors often stick to their favoured vocabulary and sentence structures, once you get half way through, you can usually read without a dictionary. Certainly if you read a second or third book in the series, it can go very quickly.

As an interesting(?) side note, I met my wife when both her English and my Japanese were quite poor. This meant that we had to listen very intently to each other. When the other person got it wrong, we knew it was a mistake and not a matter of not listening. I find that most people (me too) do not listen well in their native language and "skim" conversations -- listening to a small fraction and making the rest up. This is just not possible for beginner-intermediate language learners and ironically it formed the basis for excellent communication between my wife and I. Now that our other language skills have improved, we do not communicate nearly as well!

>repeated exposure in a comprehensible way

like a toddler?

One of the main contributions that Chomsky had on language acquisition was to provide a very solid argument that language acquisition in adults is not different than language acquisition in children. So, yes :-). This is the generally accepted point of view at the moment. But this area is very, very difficult to test so there are still some people who disagree.

Which is of course false. When you're learning a language as an adult you can read and memorize the other language's grammar from a book. This significantly speeds up your learning.

Sorry, I was speaking sloppily on a technical topic. Language acquisition is the mechanism by which language that you can remember springs unbidden to your mind when you need it. Chomsky argued that the mechanism for language acquisition is the same for adults and children. It doesn't necessarily follow that the activities will be the same.

The person that posted was correct (as far as I understand it). Repeated exposure to comprehensible language (input) leads to acquisition (this is what Stephen Krashen argues, and which is the currently most popular theory as far as I know). The mechanism is the same for adults as well as toddlers (this is what Noam Chomsky argued, and which is generally accepted).

Memorising grammar is not necessary for language acquisition in children, nor in adults. The question is whether it can speed things up. Even if you agree with Krashen, it probably can because it provides a framework for reasoning about the meaning of something. It becomes an extra tool for creating comprehensible input rather than relying on context alone.

Many people believe that using grammar rules to generate output speeds the acquisition process. My understanding of the literature (now about 5 years out of date, unfortunately) indicates that this is not at all popular. In fact, it is said to be counter productive. My experience teaching English as a foreign language also supports the notion that constructing output from grammar rules should be avoided.

How is toddler learning by 'repeated exposure in a comprehensible way' when it lacks a vocabulary/understanding? My previous post was meant to point out fault in your reasoning.

Toddlers learn vocabulary and grammar from context. Here's one (somewhat limited) technique for teaching adult learners in a similar way: https://www.youtube.com/watch?v=j6De52Pzr8c It's called TPR (Total Physical Response) and relies a lot on tying physical actions to each concept. You can see that with the physical actions it is quite easy to create a context which is understandable for things like commands.

You can imagine that TPR can not teach all kinds of grammar or concepts (for example, how would you explain government). There are other techniques for creating a context that allows comprehension of these kinds of things. One of the keys is to teach yes/no questions early so that you can probe your student's understanding without them having to speak. It takes time for students to acquire the language well enough to produce.

This is quite an involved topic and many of the teaching techniques are quite new. I have used TPRS (Teaching Proficiency through Reading and Storytelling -- not connected with TPR) in the classroom quite effectively.

If you are interested in modern theories of language acquisition, I recommend reading "Principles and Practice in Second Language Acquisition" which is available for free download from Krashen's web site: http://www.sdkrashen.com/

Although it may have escaped your notice, I've tried to point out that this is not the only competing theory in language acquisition at the moment, but it is the one which has the most acceptance now (or at least 5 years ago when I was actively reading up on the subject). Learning language primarily from grammar rules has been discredited as an effective technique for at least 50 years now. Unfortunately educators have not kept up with current research. Having said that, my experience is that some knowledge of grammar can be useful as a supplement.

I am just coming off of 2 years intensive Norwegian classes, and passed the required tests at an intermediate level, reading and listening were at a high intermediate level.

Both of the methods have their strengths. 1. Reading and understanding with dictionary: This takes a really, really long time just to read children's books and news articles. You wind up doing a lot of writing because of memory constraints, and a lot of rereading paragraphs to get the context. However, this is great for increasing vocabulary. Until you are to a certain point, however, it is probably better to learn main vocabulary words first, and then words that you will be using (interests, work, etc), and then work on the nuances.

2. Skimming: I'd not advise doing it as fast as possible, but this is good for figuring out key words you might not know and gives a clue on context. The second read, without a dictionary, you can pay more attention to details and figure some things out as you go along. I do this often and it helps me have a better grasp on the context - which is really more important.

> This takes a really, really long time just to read children's books and news articles.

Movies with subtitles are much better for this. An hour or two per day for a couple of months and you get like 10x further compared to reading.

http://journals.plos.org/plosone/article?id=10.1371/journal.... but make sure sublang=audiolang

If you can find them, but not everyone can.

They don't usually dub shows here - so many adult shows are in English. I think a good part of the movies produced are in English, and they stick Norwegian subtitles on the screen, which most folks ignore. I've found that they aren't exactly word for word - instead they are adjusted for style and culture at times. I'd have had some thing completely wrong on my own.

Children's shows on the state TV channels (think BBC for Norway) were the best find. Clear language without heavy dialects (some norwegians cannot understand each other's dialects well) plus subtitles if I wanted, transcripts on the website, and simple language. And always dubbed - kids don't start learning English until around 6 years old. Video games are usually in English. I spent many hours watching Spiderman cartoons.

However, I've met plenty of people that learned English solely through television shows and movies.

Yes, only children programs are dubbed here, although I disagree that people ignore them. We even subtitle programs movies in languages that more than 95% of the people don't speak, and assume that people can read the subtitles.

You don't see it as much today as previously, but sometimes you could hear on cinemas that people was laughing in response to the subtitles rather than the spoken words, even for English language movies. It can still happen with movies with an older audience, though today most younger people will have watched enough movies and series downloaded from the net (without subtitles obviously) that they are more used listen to the words, and in general more exposed to English.

One thing about those subtitles on movies and shows are that they are often done using the audio alone. I have seen some hilarious examples where the actor uses a generic term, while the subtitle use a specific term that do not match the context on screen.

> Reading and understanding with dictionary: This takes a really, really long time just to read children's books and news articles.

Not really a problem anymore with e-readers though: just click the word and there's the definition or the translation.

Not for all (7000?) languages though :-(

That wasn't nearly as possible for me - I gave away electronics back in the states and got new :D

It does work much better online, however, with the translate extension on chrome - though sometimes this stuff is hit and miss, at least with Norwegian - especially with phrases where a preposition changes the meaning.

I'm in the first camp of looking up every word you don't know. In my opinion speed reading doesn't do much for you if you don't know enough words. I saw an article that explained how fast our comprehension drops when there are words we don't understand in a sentence. This was done by using a slider that would hide a percentage of words in an English paragraph. Basically if you know about 90%+ of the words you can understand everything, but things drop off very quickly from there. Once you have no chance at understanding the thing you're reading I think the benefits of having read it are marginal compared to just looking up the words you don't understand.

Using a Kindle/iPad/laptop really helps with ad nauseam dictionary lookups, since you can just tap the word and get the definition right there. The Kindle even saves looked-up words to a vocabulary list to help you remember them.

By default, this only works for advanced learners. The tap-to-define action shows a same-language definition, and not a translation into your native language.

If your comprehension of the text is 50℅, you probably won't get enough from the same-language dictionary definition.

IIRC there are ways to hack around this, and install a bilingual dictionary to replace one of the Kindle's regular dictionaries.

Speed reading cannot apply to a new language. It's essentially a way to look at words, but see a dream out of them, given that your brain stores visuals for them. And you won't see anything, if your brain doesn't associate new words with anything yet.

Reply on news.ycombinator.com