Automatic reading versus the other kind (warning: long and boring!)

[wushijiao]

Anyone got any thoughts?

Yes. I once noticed that it was interesting that when I lived in Zhengzhou (郑州) in Henan (河南) as a beginner student, I'd notice the characters fro Zhengzhou and for Henan everywhere...maybe a few dozen times per day (on signs, papers, restaurants, beer bottles...etc). Thus, even though I was a very slow reader in general, I could read those particular characters in a block, and very fast.

And essentially, i think that's what Chinese reading is all about: learning to read in ever greater blocks. This can be done in place names, peoples names, but also in common collocations of words.

It takes practice, and an expansion of knowledge. But I also think that speed reading drills can be useful as well.

[roddy]

David Wong commented that "I don't let unknown words slow me down."

Sorry, but I don't think you get a choice in the matter. You still need to look at the thing and figure out if you know it or not, and I can't see how that can happen at the same speed as fluent reading. I'd love to hear how it can though . . .

The 'yes, we know' comments to the OP are missing the point, btw. This is about why, and like much language learning it's relevant to the process of changing monitored, conscious behaviour (teacher says 你好, you think back to last week's lesson and recall the correct response, put your tongue and mouthbits in the right place and say 你好 back) into automatic subconscious behaviour (teacher says 你好, and you respond barely noticing you're doing so, while tying your shoelace and listening to another conversation between your classmates).

And as I've said, it ties into why accuracy is important - if you aren't accurate your interlocutor gets pulled out of their automatic understanding of what you're saying and has to think. Get your tones and pronunciation off a bit, and nobody knows if you're looking to sell a watch or buy a ticket*. Sure, they can figure it out from context, and I'm the first to admit you should get out there and do the best you can at the time. But ultimately, you don't want people to be having to figure anything out - people don't have massive figuring-things-out resources, and using them up has consequences for the amount of effort they can put in to doing whatever you're trying to get them to do. A quick exchange on the street and you can get away with it. But sit down for prolonged negotiations with a business partner, or a relationship discussion with a life partner, and you don't want to be tiring them out any more than necessary. Read the example about the Israeli parole decisions. I wouldn't want to be presenting my case in mangled Hebrew, or even be after the guy with mangled Hebrew.

Those couples you see having stressful conversations in bad English and worse Chinese - would their conversations be stressful if they had a decent shared language?

I think pretty much everyone on here will agree that accuracy is important, so to an extent this is preaching to the converted. But I think it's interesting to have some more ideas about why it's important. It'd be fascinating to have some research on how the extra effort of understanding inaccurate speech depletes cognitive resources, but given the example of the book in which, eg, doing hard sums reduces the amount of willpower available for tasks later, I find it entirely plausible that there are measurable effects.

The book, Thinking Fast and Slow, is well worth a read. It's accessible, but by no means dumbed-down pop science. NYT review for anyone who wants to read a little more. It's remarkably widely applicable - I found myself reading it thinking things like 'Ah, so that's why I hate shopping.' See also this book on willpower, which ties in quite nicely.

Edit: Side-point: You need to be careful referring to your own experiences of what are actually subconscious processes. E.g. I've seen people say they don't sub-vocalize when reading their native language - missing the point that it's a subconscious process, and by definition you don't know if you're doing it or not.

*Come on, at least I didn't use horses and mothers.

[Silent]

David Wong commented that "I don't let unknown words slow me down."

Sorry, but I don't think you get a choice in the matter. You still need to look at the thing and figure out if you know it or not, and I can't see how that can happen at the same speed as fluent reading. I'd love to hear how it can though . .

I think it's really not that hard to keep up speed despite unknown words. Due to it being a language you're learning you won't read full speed, but it does not mean you can't keep up a pace despite unknown words. It's just a matter of alotting the same amount of time to each character. Look at it and move on. If you don't know the character you loose info. If you can't link it fast enough to the surrounding characters you loose info. It's finding a balance between effort invested and comprehension rate. Going through the text 'fluently' will make reading more enjoyable but this off course may be (over) compensated by the lower comprehension.

[Guest realmayo]

Well that just seems to miss the entire point! Of everything!

Okay: there will be words that you just know, directly and effortlessly. Same as you just know that 2+2=.... 4. But also known words that require effort for you to think about. Again, consider a difficult sum which you can work out if you spend time on it. That was all in the first post.

It's just a matter of alotting the same amount of time to each character

Can you really mean this? That you'd read super-simple words like hello and tomorrow at the same speed that you'd read ones which you know, but a lot less well? That is, words which you come across and have to pause for a second to work out if you know them or if you dont? I'm not saying that if you decide you don't know it there's anything wrong with just continuing. I'm just saying that most people's reading speed will be jerky if there is a mix of the super simple words that you just know, and more difficult ones, as well as unknown ones.

If you really mean what you just wrote, that you give the same time to each word on the page, then you will either be reading the simple words ridiculously slowly (you really need to spend that much time reading "hello"? if you see 2+2 then you will do the sum in your head just to make sure it really is 4?).

Or: by trying to read fast, so that you're only understanding the words you know really really well, you will be skimming over words which you would know if you read a bit them slower,but because you're reading very fast you don't understand them at all. Which means that you are not improving your grasp of words which you are slightly familiar with and which would become more firmly cemented into your vocabulary if you read (and noticed) them in a variety of different texts.

[Silent]

If you really mean what you just wrote, that you give the same time to each word on the page, then you will either be reading the simple words ridiculously slowly (you really need to spend that much time reading "hello"?

You're taking it very litteral here but yes I mean what I wrote. If you alot something it does not by definition mean you use up the alotment. You alot the same time to every character/letter/word whatever. If you can't figure it out in the alotted time you move on and maintain your speed. This does not mean that if you figure it out within the alotted time you have to wait till the alotted time is over. Actually it may very well be that you move on before time is up by just deciding you don't know it. This way you, as David stated, don't let unknown words slow you down.

Or: by trying to read fast, so that you're only understanding the words you know really really well, you will be skimming over words which you would know if you read a bit them slower,but because you're reading very fast you don't understand them at all. Which means that you are not improving your grasp of words which you are slightly familiar with

It's not this black and white. Oftentimes on first sight of the characters I know the approximate meaning or know that it means one of two or three things that I tend to mix up. With a little context added you get the correct meaning. You get this context by reading on, not by thinking and analyzing the character! IMHO in a language you're fluent in it works the same way. If you read a word with more then one meaning you're not going to think what might be meant here, you just continue reading and interprete in the process of reading on, sometimes even a few sentences on. Oftentimes you don't even read perfectly linear but take in a line, sentence or even alinea. Pick out the keywords and structure and go on from that, or when skimming, leave it at that. By 'stopping' at every character to think about it, to analyse it and try to 'translate' it you impair the natural reading process. Sure, by skipping it (too fast) you loose info. But then, by reading so slow and spending plenty of time on each character you don't know perfectly may loose info too as you will loose the main story lines

[roddy]

If you can't figure it out in the alotted time you move on and maintain your speed. This does not mean that if you figure it out within the alotted time you have to wait till the alotted time is over.

So unknown characters slow you down, and you don't have much choice about it. Which is what I said. Remind me again what your point is . . .

[Guest realmayo]

If you can't figure it out in the alotted time you move on and maintain your speed.

Can't you see that this is completely different to what I was saying about seeing something like "hello" or "2+2=?" and knowing what it means immediately & effortlessly? You're ignoring that entire concept. I'm not saying you have to agree with the theory of the guy who wrote the book which prompted me to make the initial post in this topic. But you could either acknowledge it as the context in which these discussions began, or say why you think it's wrong.

[jbradfor]

This does not mean that if you figure it out within the alotted time you have to wait till the alotted time is over.

Soooo..... if you read it automatically you read it faster, and then you don't wait the allotted time, so you spend less time on the words you know quickly. I don't mean to pile-on, but I'm also not sure of your point.

It sounds to me like you mean you have a maximum time you will spend on any one word. Good. But that seems different.

[David Wong]

Well, allow me clarify what I meant (and as I said that only half in jest, I feel that I need to defend the part of it that has some truth to it).

A better way of putting it would be that I don't always pause to look up characters that I'm not 100% sure about. Not being able to understand certain characters might place a significant drag on my regular pace if I were reading at a fairly advanced level, which I'm not. But as it is, my regular pace is pretty much just plodding along, and unknown characters don't take that much more time for me to process or acknowledge, before moving on and hoping that understanding them wasn't something critical to understanding the entire passage. Sometimes I don't even realize that certain characters are unknown, and assume that they're another character entirely. Those are the times I usually get in trouble.

Sometimes I luck out and could decipher from the components of the character, or recognize one of the two characters in a word enough to gauge if the word/character has a positive or a negative connotation, and often that's enough to allow me to proceed.

Allow me to illustrate with an example. Here's a passage I read not too long ago.

威利·卡瑞尔年轻的时候在纽约州的水牛钢铁公司做事,有一次需要到密苏里州的匹茨堡玻璃公司去安装一架瓦斯清洁机,以清除瓦斯里的杂质,使瓦斯燃烧时不致伤到引擎。

这种清洁瓦斯的方法是新的,过去也曾试验过。可是在密苏里州安装的时候,遇到了许多事先没有料到的困难。经过一番努力,机器勉强可以使用了,然而,远远没有达到他们保证的质量。

他对自己的失败感到十分懊恼,好像有人在他头上重重地打了一拳。他的胃和整个肚子都扭痛起来,烦恼得简直无法入睡。后来,他意识到烦恼不能解决问题,于是,想出了一个消除烦恼的方法,效果显著。

这个消除烦恼的方法非常简单,任何人都可以使用,可以分三个步骤。

第一步,不要惊慌失措,冷静地分析整个情况,找出可能发生的最坏情况—“没有人会把我关起来,或者把我枪毙,这一点我有把握。充其量不过丢掉差事。也可能老板会把整个机器拆掉,使投下的两万块钱泡汤。

When I read it for the first time, I either did not know or wasn't totally confident about the following words:

瓦斯 I had no clue but it seems technical and I guessed that it wasn't crucial to understanding the passage. Move on.

杂质 Not very sure, although I knew the component characters, or at least one of them. I recognize 杂 from 复杂 and I took it to be something negative.

引擎 I knew 引, but that's not helpful here. Only one occurrence of this word and it seemed technical, so I ignored it. If I had bothered to guess at the pronunciation of 擎 I could have guessed it.

懊恼 I knew 脑, and I might've even recognized 恼 if it had been in the context of 烦恼, so I took a guess here and lo, there's something I recognize a little further on which pretty much confirmed my guess. At the time, there's a good chance that I entirely mistook 恼 for 脑。

显著 Wasn't very sure, although I thought I recognized 显 from 明显, so I took it to be something positive.

步骤 I just knew 步 but then what follows, 第一步 confirmed my guess. This example illustrates how the penchant of Chinese for two character groupings help us when only one of the characters is crucial to the meaning of the word grouping.

惊慌失措 no clue but I knew 惊 and I guessed that it connotes something negative.

枪毙 I knew the first character and I think this was almost an automatic guess. I recognize the 死 primitive. The 比primitive gave me a clue to pronunciation. I knew the spoken word, I just don't recognize the written word.

充其量 had me completely stumped, but it seemed pretty important so I looked it up.

泡汤 I knew meant to make soup, so I guessed at the secondary meaning here.

I hope what I wrote makes sense.

[yialanliu]

I skimmed the posts and realized something as I was skimming.

Even with words I didn't know, I never slowed down and would think about it as I read (multitask).

So from that eprspective, I think it is quite possible to not slow down. However, I do beleive that there is a difference between reading and skimming but when skimming, the flying past at same speed is possible. Comprehension might be lower, but skimming is about saving time.

While I agree certain things become reflexive like the last few things the OP mentioned, I think a lot is actually trained. I believe language falls in the 2+2 category. Every language can be learned by an infant yet by the time we are older, it becomes much harder. This leads me to think a lot has to do with retention as has been proven by science that it's harder to retain things in your brain the older you are. Thus the adage of practicing is true. It just takes more practice time the older you are.

[anonymoose]

It seems that most people support not spending too much time dwelling on or looking up unknown characters and words whilst reading.

As a learner, I take the opposite view. I understand that spending more time flicking through a dictionary than actually reading the text can have a negative impact on one's enthusiasm and motivation. However, by skipping over the unfamiliar characters and words, then one is only reinforcing the parts that one already knows, and foregoing the opportunity to learn something new. In the long run, this only perpetuates the frustration associated with encountering so much unfamiliar vocabulary.

Personally, if my objective is to learn and improve my Chinese, then I look up every character or word I don't know. Admittedly it can be painfully slow at first, but as a result, I have reached a stage where I can now read practically anything in modern standard Chinese without too much difficulty. That doesn't mean that I know every single word, but after you reach 99% comprehension, the remaining 1% has much less chance of stumping you. Nevertheless, I will still look up the unfamiliar words, in order to take my Chinese that bit further.

[Silent]

Can't you see that this is completely different to what I was saying about seeing something like "hello" or "2+2=?" and knowing what it means immediately & effortlessly?

My statement has nothing to do with this. I just replied to Roddy who stated basicly that getting slowed down by unknown characters is inevitable. I just tell it's not inevitable, just alot a maximum time spend to each character/word. The price you pay is lower comprehension. This has nothing to do with automatic reading or not. It is a matter of balance between comprehension and reading speed. The same occurs in your native tongue. You can skim over a text or read it very intensively soaking up every detail and there's a whole array of 'reading modes' in between.

I think this has little to do with effortless or effortfull. Apart from that the knowledge level is far more diverse, you know them instantly/intuitively, you know them effortless but slow, you know them with a small conscious check, you know them after long intensive analysis and deep thinking, you know don't really know them but know what it relates too, you don't know them at all, etc..... reading is far more complex. Reading is also (re)interpreting what is written while you're reading. By looking up everything there's a fair chance you'll loose the main storylines in favor of (unimportant) details. You essentially impair the natural processing by reading too slow. Try reading extremely slow or listening too someone who speaks extremely slow in your native tongue you'll get lower comprehension too. This is because you forget the info provided before the context is taken in and interpretated.

[Guest realmayo]

Ok, you disagree with everything I wrote, as you're perfectly entitled to. Any futher chat with you on this topic is therefore pointless.

[roddy]

Well if you're going to mess around with more than one variable at once we're never going to get anywhere. Scientific method folks, scientific method.

[imron]

I agree with what anonymoose wrote above. It's important not to develop the habit of always skipping over the words you don't know otherwise your level will never really improve. That said though, I don't think it's necessary to look up every single word, otherwise it's very easy to overwhelm yourself. As I've mentioned in other posts, just set yourself a quota for the day and once you've looked up that many words, then just keep reading as you otherwise would. If you're doing enough reading, the words you missed today will show up again tomorrow, or the next day, or next week, or whenever, so you'll pick them up eventually anyway.

Getting back to the original question though, to get that automatic and instant recognition it really requires you to work at it, and I don't just mean reading more, but rather specifically working on getting instant recognition when reading. One thing I've found that tends to exacerbate 'slow thinking' in this regard is the use of SRSing. SRSing tends to encourage a 'set and forget' mentality and if it takes half a second to recognise a word many (most?) people would count that as a pass and move on, when in reality, if you want to be training automatic and instant recognition you either need to count it as a fail, or at the very least spend a bit more time drilling that word in your head practising to get that instant recognition before moving on to the next card. If I calculate the average time I spend per card when doing flashcards, it's consistently about 30 seconds per card - somes cards are instant, but for the ones that aren't I always make sure I've trained the word to the point of instant recognition (at that point in time) before moving on to the next one, and that's what sends the average up. This is not to say that I'll recognise the card instantly the next time I see it but at least now it's closer to that point and I'm confident that this is the skill I am training. For decks where you have large number of repetitions per day (more than say 50-60), it becomes harder to be so strict on yourself, but that strictness is necessary if you want to build that instant reflex for these words.

It's also important to do exercises where you practice improving your reading speed. Often your internal reading speed will settle on a nice, easy pace you picked up at the beginner/intermediate level, and once it's settled on that speed it's likely to stay there unless you make conscious effort to speed it up. With a bit of consistent practice, it's quite simple to double your reading speed (see my thoughts here).

[imron]

Also, just to add my thoughts on the "whether unknown characters slow you down or not" side of the discussion. I think what Silent is saying is that rather than "unknown characters slow me down" he's saying "known characters speed me up". Which might seem like six of one, half a dozen of the other, but a lot depends on your perspective and also your current language abilities.

If you are having instant recognition for a large number of words in what you are reading and you then hit an unknown word/character, it's going to slow you down. If on the other hand you are at the point in your reading where for most of the words you have to think for a fraction of second to get recognition, then it probably takes just as long to think "unknown word, skip" as it does to recall the meaning. In this situation words you can recognise instantly will speed you up slightly, but unknown words won't necessarily slow you down because you're already reading at a slower pace.

As mentioned above, this is probably the point where you settle internally on a reading speed that you'll stick quite close to even as your reading skills improve so it's important to always try to keep pushing that limit.

The faster your reading gets, the larger the relative impact/slowdown caused by unknown words will be.

[roddy]

True enough, I can see that being the case if you're still reading at a very slow speed. But I maintain that once you get any kind of flow going on unknown characters are effectively speedbumps. (hey, did you just ninja-edit something in, or did I just not read your post properly)

[imron]

Ninja edit. I agree that as your speed increases, unknown words will have more of an effect.

For arguments sake, let's say it takes you half a second to process an unknown character (i.e. to see the character, realise you don't know it, or only know part of it, and decide to either skip it or make an intelligent guess based on context).

If you're reading at less than 120 characters per minute (half a second per character), you'll find that unknown words cause little to no slowdown. As your reading speed increases however, the slowdown will feel more apparent.

At 200 cpm (0.3 seconds per character) an unknown character takes almost twice as long to process as a normal character. At 300 cpm (the bottom range of native reading speeds) you're reading at 0.2 seconds a character and it will take 2 and half as much time to process an unknown character. At 400 cpm, it's 0.15 seconds vs 0.5 seconds and so on and so on.

[Silent]

Ok, you disagree with everything I wrote, as you're perfectly entitled to. Any futher chat with you on this topic is therefore pointless.

Sorry, but I really don't know what you're referring to here and where your irritation comes from. The fact that one knows some things instantly and have to put in an effort to figure out others is pretty obvious. No disagreement there unless you insist there are exactly two systems as to me it seems there's an entire range from from effortless/instinctively to huge effort. From knowing exactly to having a rough idea to not knowing it at all. Effort needed depends on the degree of knowing.

Imho nothing new there, again we agree as that's how you finished your initial post.

If there's an disagreement over slowing down or not slowing down because of unknown characters I think it's more based on differences in definition and misconceptions of what I try to express. Apperently I'm not able to express myself clear enough.

[Guest realmayo]

Imron, you go some way to calculate the extra time involved with words you don't immediately know, and I'm not disputing that, but I'm not sure time itself is the obstacle: there are plenty of books (in English) which I wish lasted longer, because I enjoyed reading them. For me, one of the things from this book I mentioned was the real unpleasantness of having to switch gears and bring the slow, logical part of the brain into play one moment, then back to intuitive easy comprehension, then back to the slow part again -- the book proposes that we don't like doing this, it's an effort we seek to avoid, and that tallies with why it feels I have to force myself (against my will!) to read stuff in Chinese. As Roddy highlighted earlier, this isn't anything new about the "what happens", but just an interesting (for me) theory about the "why".

As for learning new words: I've long felt that the process of seeing a word that I know I should know, that I learned recently, and putting effort into recalling its meaning -- that this process is very helpful for cementing the word in my brain for future use/recall. Even though, according to the hypothesis behind this book, it's hard work. I'd worry that adopting your instant-recognition criteria for SRS recall would mean that I wouldn't allow a remembered-with-some-difficulty word to stretch out to a longer interval and so wouldn't require a second, later dose of brain-work to dredge up -- and it's this type of brain-work that feels, for me, where the real long-term learning is happening.

Silent: I'm not "insisting on exactly two systems", I'm paraphrasing what the guy who wrote the book is saying, that there are two systems. I guess it's a bit like being pregnant: either you are not, or you are -- and if you are, there are many different stages of pregnancy. In the context of the book: either it is easy/effortless/intuitive etc etc, or it is hard and requires the second system to click in, although the amount of effort the second system will have to do will vary.