Extensive reading and reading speed

[realmayo]

Wow novels four times - I was just assuming read once, then re-read once! Still, seems like it worked. Did you enjoy the books, even fourth time around?

[Moshen]

Quote

My arbitrary goal is to reach an average of one unknown word every 3 pages (i.e., the majority of pages have no unknown words). 

 

I'm a native English speaker with a much larger than average vocabulary, and I've read numerous books where I might have an average of one unknown word every page or two!  Most of the time this does not send me to the dictionary, because I'm able to guess from the context what the words mean.

 

So I'm not sure your standard is a good one.

[Woodford]

10 hours ago, realmayo said:

Did you enjoy the books, even fourth time around?

 

I think that because the first three times were all digital and my fourth time used the physical copy of the book without the help of Pleco, the fourth repetition was always an interesting experience. However, it was all such a chore that by the end of the process, I probably felt a bit too numb to enjoy anything to the fullest!

[tsitsi]

@pinion @Dr Mack Rettosy Thaks for your inspiring stats, how do you track your reading speed? I'm just starting on my reading journey And where did you find your reading material to get you going?

[大块头]

17 hours ago, tsitsi said:

how do you track your reading speed?

 

I do all my reading on my phone, and Android's "Digital Wellbeing" function tracks how long you use your apps each day. Using that to time my reading is easier than futzing with a stopwatch, and it also allows me to separate out the time I spent reading entries in Pleco.

[tsitsi]

True, I guess I could track some reading exercises via time spent on app. but I'm hoping there is a web app or extension I am yet to find for desktop that will auto-magically do it for me ??

[Dr Mack Rettosy]

Quote

Thaks for your inspiring stats, how do you track your reading speed? I'm just starting on my reading journey And where did you find your reading material to get you going?

 

I do all my reading on an Apple device and use screen time to track. I use the Pleco app's e-reader, and most of the content is purchased within Pleco, like the graded readers they offer. I highly recommend Mandarin Companion for starting out. The Rainbow Bridge series is also excellent, but a little more challenging for beginners. I found having started with RB, my reading speed had a massive jump switching to MC (see my updated progress below)

 

 

That said, I'm looking forward to the day where I can read native materials on paper, but that's a long time from now.

 

@pinion Thanks for your reply to my questions. I haven't been around much on the forums, but saw your reply some time ago and really valued your feedback.

 

 

[黄有光]

On 7/29/2021 at 10:02 PM, Dr Mack Rettosy said:

That said, I'm looking forward to the day where I can read native materials on paper, but that's a long time from now.

Same!  In fact that's what I'm working towards now. Looks like I'm on track to meet that goal by 2025, but we'll see.

[phills]

Thanks @pinion for the chart.  It's really inspiring!

 

I tried to do the same (build up my reading ability) during this COVID period, although I only gotten through about 2 million chars so far.  I posted about my experience in another thread. 

 

https://www.chinese-forums.com/forums/topic/61248-reading-material-chasm/?do=findComment&comment=481601

 

I haven't focused as much on speed as on parsing sentences & stamina.  I'm plateauing right now at around 100-150 chars per min, but it's nice to see I'm on track per your chart, and my speed will eventually get there. 

 

--------

One question I have for the group is how they've approached building up reading skills when dealing with both Simplified and Traditional character texts at the same time.  I only recently added traditional chars to my repertoire, after I became "mostly" satisfied with my ability to read simplified chars.

 

Learning traditional chars wasn't bad.  I just drilled on a 500 char deck that I got from this site on another thread.  However, one side-effect I noticed is that it seems to have stopped/slowed-down my improvement in my reading speed when reading simplified chars. 

 

I haven't forgotten how to read simplified chars, but I'm not improving as much, and occasionally get confused like between 曾 and 会(會), simple characters that I've long mastered previously. 

 

Now I "know" how to read traditional chars (for basic purposes of street signs, labels and short passages of text), I'm wondering if I should shift back to focusing on mastering simplified first, rather than diversifying. 

I'm curious how others have approached this.  Continue deep-diving in one form first, or keep flipping between the 2?  Also, for people who've reached native reading ability levels in both, do you really read both at the same speed or are you faster at one version or the other?

[Lu]

On 9/3/2021 at 8:33 PM, phills said:

I'm curious how others have approached this.  Continue deep-diving in one form first, or keep flipping between the 2?  Also, for people who've reached native reading ability levels in both, do you really read both at the same speed or are you faster at one version or the other?

I don't claim anything near native reading speed, but I read Chinese no problem, both simplified and traditional, at about the same speed, with a slight preference for simplified. I guess what I did was alternate them, but for long periods with each type. I first learned traditional, then spent a year in Beijing and picked up simplified, later spent a year in Taiwan and picked up traditional again. Then worked in Taiwan as a translator for about a year, working with traditional. I don't know if my trajectory is something you should plan for -- it was a process of many years -- so just offering it up as a data point. With traditional, I used to have some difficultly with reading vertical lines: they are longer than horizontal lines and I'd have trouble finding the next line after getting to the end of one. But I think that's gone away with practice.

[imron]

On 9/4/2021 at 4:33 AM, phills said:

Continue deep-diving in one form first, or keep flipping between the 2? 

I think it’s better to deep dive in one, and then pick up the other. 
 

Once you’ve read a dozen novels in your strongest, then switch to reading the other, and you’ll be fluent enough in the other after one or two books. 

[phills]

Thanks for the perspectives.  I think I'm going to keep focus on simplified chars for now, and not switch yet.

 

I thought I was "mostly satisfied" with my reading speed after 10 books, but I think I jumped the gun.  I still am not fast enough read TV subtitles in real time, when the show is in neither Chinese nor English.  

 

E.g. I can't watch a foreign language movie, subtitled in Chinese, and get all of the plot.  I think that requires 200-250 chars / minute. 

 

I think I'll try another 5 books in simplified chars and re-evaluate.

[pinion]

@phills As usual I'm late replying, but yes, thirding @Lu and @imron on getting good at one before diving into the other.

 

Reading vertical lines also really threw me for a loop at first--I'm pretty much used to it now, but I still occasionally get to the bottom of one column and then jump to the top of the wrong column, which never happens when reading horizontally.

 

I think my reading speed is still increasing, albeit slowly, but the biggest difference I've noticed in the year(!) since I made the original post is that reading has become much less taxing for me--it used to really require 100% of my attention, whereas now it only requires maybe 70%? I can now read in Chinese as a way to relax when I'm tired (though it probably helps that I tend to choose authors I'm more familiar with for this).

[phills]

@pinion please keep posting your progress charts & data

 

It's amazing how much my progress is tracking yours.  I'm at 3.2 million chars now, and I just measured a high speed of 188 chars per minute (good flow, familiar authors / subjects), and a low of about 120.  I dropped traditional chars for now, and got my old speed restored.

The progress is right along the lines of your chart.  I despair of ever getting to 15 million, but am targeting 200 cpm (on a good run) by 5 million.  

 

It's as close to "science" as I've found on acquiring Chinese reading ability.

 

I wish I could contribute more to some kind of wiki-speed project.  But I hate recording my speed each time, because that tends make me anxious while reading (feels like an exam with a ticking clock in the back).  So I check it every once in a while, just to see whether I'm staying within the band. 

 

The next checkpoint for the graph will be seeing if by 5 million, I can hit 200 cpm as a high, and consistently stay in the high 100s (150-200) on average.  If so, your graph will have matched my experience at 600k (when I "got" the knack of reading, <100), 2m (100-150), 3.2m (120-180) and 5m.

 

Maybe someone good at internet / user-contribution projects can come up a good way to combine the data from everyone's experiences together.

[大块头]

On 10/1/2021 at 6:24 AM, phills said:

Maybe someone good at internet / user-contribution projects can come up a good way to combine the data from everyone's experiences together.

 

My next project is a tool that predicts when you'll reach a near-native reading speed using a statistical technique called regression analysis. I'm still refining it in Python (animation below is my progress as of last night), but the goal is to release it as an easy-to-use spreadsheet. I think it'll be possible to get an accurate prediction with only ~5 books. The early-time predictions are currently all over the place, but incorporating a Bayesian prior informed by @pinion's dataset should smooth that out.

 

If anybody reading this has similar reading rate data they'd be willing to share, please get in touch via Chinese-Forums or email (elsey.jack@gmail.com).

 

[phills]

That's very interesting.  Shouldn't the X-axis be "number of characters read" rather than "days"?  

[realmayo]

On 10/1/2021 at 5:00 PM, 大块头 said:

I think it'll be possible to get an accurate prediction with only ~5 books.

 

Wouldn't the content/style/genre of those five books, as well as your state of mind, social life, and your Chinese reading ability, etc, have to be pretty similar/static throughout the whole process?

[大块头]

On 10/1/2021 at 12:24 PM, phills said:

Shouldn't the X-axis be "number of characters read" rather than "days"? 

 

It could be, but I think total reading time is a better choice of independent variable. The model I've come up with assumes that the rate at which your reading rate increases is linearly proportional to the difference between your current reading rate and the native reading rate your performance will asymptotically approach.

 

 

On 10/1/2021 at 12:29 PM, realmayo said:

Wouldn't the content/style/genre of those five books, as well as your state of mind, social life, and your Chinese reading ability, etc, have to be pretty similar/static throughout the whole process?

 

The model assumes your reading ability (measured as your reading rate) will increase as you read more books. The other factors should remain roughly static for best results, but they don't need to be similar, if that makes sense. For example, you can select books from a wide range of different authors and genres, but you should select them somewhat randomly. Certain statistical assumptions will be violated if you start out reading only wuxia novels and then suddenly switch to philosophy textbooks halfway through the process.

[alantin]

@大块头 Awesome！I'm a big fan of Bayesian statistics too and have done something similar to predict times for learning x numbers of characters in the past.
I also think a number of characters read would be a more informative value for the X axis but I don't know how this would fit to your model. I guess in any case you need to keep a consistent pace of reading in order to see the predicted results. I wonder how taking brakes or changing focus between reading, watching movies, talking with people, and writing would affect the pace of increasing reading speed.

[realmayo]

On 10/1/2021 at 5:57 PM, 大块头 said:

but you should select them somewhat randomly

Fair enough. However I think most people start with easy books and then increase difficulty over time.