Learning and Maintaining 10+ Languages

My first conversation class went really well yesterday. Obviously, I didn’t speak very well, since it was my first time. But even though I had less preparation than any of my previous languages, this was the best first conversation ever. I entered a lot of items in anki – 38. That may be the most ever, and it’s a rate I can’t maintain for long. That number needs to drop below 20. Maybe by the end on the week...

Also, after the lesson I immediately decided to stop Pimsleur. No need to even finish the first level – my time is better spent doing other activities now. In the past, I would have bought CDs for all three levels, total cost of over $300. This time, because of the subscription format and fewer lessons, I only spent $20!

Today, my second class also went well; 31 new items in anki. The teacher was not as good at conversation; I had to drive it most of the time, which is hard for a beginner. Some teachers just know how to do random conversations, and some don’t.

Understanding the language is the least of my challenges – I probably understood 90% of what she said today. But I yearn for the days when I will be able to speak comfortably, without wondering if I can just use a Spanish word for the Portuguese. I’m thinking about visiting Brazil a couple months from now. If I can get 50 hours of conversation under my belt by then, I should be fairly comfortable. I plan on just staying in one city, meeting language exchange partners and doing day trips for 1 month.

I hit a landmark a few days ago; I passed the 10 hours of conversation mark, so I decided to make a new entry here. This is usually the point where I feel comfortable saying I’m A1 (Beginner). However, I believe I’m already A2 (pre-intermediate). It was my 7th lesson where I first felt this way, and I have not fallen back to a lower level since then, which sometimes happens in the early stages of conversing. In fact, I seem to notice some improvement every time. I feel like I’m trying to “catch up” to my true level in the language. This is a good thing – whenever I feel this way, I tend to make steady progress at a decent rate.

It is possible that I’ll reach a level that I feel is roughly B1 (lower intermediate) before going to Brazil; I’m sure going to try to make that happen. Previously, I stated that I wanted to complete 50 hours of conversation before going there, which would require me to wait until late November to go. I may decide to go as soon as early November though, because that is my traditional time for heading out on my annual travels. So 40 may be more realistic.

Another clear sign of progress is that the number of items to put in Anki after my conversation has dropped dramatically. These are the numbers for all 12 classes I’ve had so far, in chronological order: 38, 31, 23, 24, 21, 16, 26, 13, 13, 11, 10, 14. Ime, when these numbers are under 20, my Anki work does not overwhelm me, so there is no need to delete or limit anything.

Reading has been going well. I continue to read half a passage from LC Conversations per day. Although the percentage of known words has been consistent (85-90), I am reading faster, and I am finally confident enough in my pronunciation now to read without playing back every single sentence to make sure I didn’t mess up. Apparently the early pronunciation work, including the guide I put together, paid off. Also, although they talk quite fast and colloquially, I am definitely understanding more than I did before. I will soon ramp up to one passage per day.

I type a few lines every day that I study, and post it in LC Write and Correct. This process has helped me figure out the correct way of saying certain things that I was wondering about. I also do 5 lines of hand written scriptorium, which I think gives my brain yet another way to connect with the language.

The biggest challenge in language learning is getting to the point where you can understand movies, TV, natives talking to each other, etc. This is a level of listening skill that is far above merely holding a one-on-one conversation. I am still far from understand at this level. After Brazil, I hope that I will understand over 50%, on a sentence level, of most series that I watch. I haven’t really checked my level, but I’m probably 10-20% right now. Lots of work to do!

Today I completed Teach Yourself Brazilian Portuguese. I say “completed”, but I mean passively. I worked through it quickly, making sure I read and understood everything. Tomorrow I will start it again, but this time I will work through it actively. I will make sure I can translate from English to Portuguese, memorize all the vocabulary and grammar points. This will take a while. Ime, the vocabulary sticks really well when I do this, but grammar just so-so. It is still worth the effort though.

Today I also completed my 20th hour of conversation. I had a teacher cancel at the last minute, but found a replacement – the first man I talked to. He started out talking very colloquially, which threw me a bit. For example, he said tá instead of está. I commented on this, and he stopped doing it, which is sort of a shame. I need to eventually be able to understand natives speaking colloquially like this. My reading is all colloquial, which is good, but I need to get there in all skills.

I continue to be amazed at my progress in this language. I started less that 2 months ago, and feel like I’m close to B1. In most of my previous languages, I wasn’t even A1 after 2 months. I have been telling my teachers that I started 3 months ago, and they are amazed. I decided to do this for simplicity sake – I can say “3 months” up until I’m in Brazil, and it will still be sort of accurate. Everyone is telling me my Portuguese is amazing, especially considering the time I’ve put into it. These types of complements are really common though; I will try not to get a big head, lol.

I think I can now safely claim that the shortcuts I decided to take with this language made sense. I’m talking about stopping Pimsleur after only 20 lessons, and starting conversation after only 3 weeks of preparation. Another shortcut, which I haven’t talked about, is limiting my “explicit” vocabulary studies. I still put items from my conversations into anki, but not every little thing as I did with previous languages. And I haven’t loaded anything from my textbook, reading, etc. I feel that the vocabulary is soaking in at a much greater rate than with previous languages. I almost feel like one of those people who can get by without doing any explicit vocabulary study. Most of those people are students of European languages, and now I believe it can work in that case. Actually, I have stated this before, but I repeat it now. I firmly believe that we assimilate most of our vocabulary through exposure and practice regardless of whether we explicitly study it or not. Massive exposure and practice are mandatory. Explicit study isn’t, but it increases our efficiency, especially when it comes to dissimilar languages. Portuguese is about as similar as they get, so explicit study is less of an aid.

Anyway, things are very encouraging! I look forward to my trip to Brazil next month.

I’m happy today because my tutor said “Why does your profile still say you are A1? Your level is B1!” That was good enough for me. Right after our class, I updated my level. I started studying on August 16, and it is now October 15, so approximately 2 months. Very cool.

I have been reading posts about how people are learning Portuguese. So many are using inferior methods, it was shocking at first. But now I understand that just about anything works for this language. Now I better understand why people push these methods, and think they are “good” in general. They do not realize how weak they are and that they fail miserably for languages that are not similar to their L1s.

Anyway, I will continue to use my hard core method that works for all languages. Today was my 29th 1hr conversation. I have 14 more scheduled before my trip. That makes a total of 43 hrs of conversation before arriving in Rio. At first glance, that doesn’t seem like very much. But when I first went to Russia, although it was after 6 months of study, I only had 30 hours of conversation in. Also, my level was only A2. So I feel I am better prepared. My new goal is to reach B2 by the end of my trip. That’s about 2 months from now, so I think it’s possible. Actually, regardless what my level is, if I can review it only once a week without a big drop in skill level at that time, I will be happy.

I’m happy today because my tutor said “Why does your profile still say you are A1? Your level is B1!” That was good enough for me. Right after our class, I updated my level. I started studying on August 16, and it is now October 15, so approximately 2 months. Very cool.

I have been reading posts about how people are learning Portuguese. So many are using inferior methods, it was shocking at first. But now I understand that just about anything works for this language. Now I better understand why people push these methods, and think they are “good” in general. They do not realize how weak they are and that they fail miserably for languages that are not similar to their L1s.

Anyway, I will continue to use my hard core method that works for all languages. Today was my 29th 1hr conversation. I have 14 more scheduled before my trip. That makes a total of 43 hrs of conversation before arriving in Rio. At first glance, that doesn’t seem like very much. But when I first went to Russia, although it was after 6 months of study, I only had 30 hours of conversation in. Also, my level was only A2. So I feel I am better prepared. My new goal is to reach B2 by the end of my trip. That’s about 2 months from now, so I think it’s possible. Actually, regardless what my level is, if I can review it only once a week without a big drop in skill level at that time, I will be happy.

I just wanted to post a quick log entry to close out this Portuguese spurt. At the end of November, while still in Rio, one of my teachers implied that I was no longer lower-intermediate. I interpreted that to mean I am now upper intermediate, or B2. And I’ve improved since then, so I feel good about saying that I made it! From zero to B2 in 4 months.

That’s really about it. I finished up my month in Rio, and now I’m relaxing in Medellin Colombia. I’m not really doing anything special language-wise, except try to re-capture my French level, which has suffered dearly due to interference from Portuguese. I’m not sure whether I’m going to Thailand or Cambodia next. If it’s Cambodia, I might actually start that language. If it’s 3 months in Thailand, I’ll probably get back to revising Tagalog Lite.

It’s been 8 months since the last post; about time for an update. I finished the 4th draft of Tagalog Lite last month. Although it was a major change from the 3rd one, it only took 2 months to complete; far less that I expected. I don’t expect nearly as many corrections in editing this time.

I went to hand it off to the editor, and was told that he would not be available until September. So I decided to write a set of drills for each lesson. This was always part of the plan, although I intended to do it after editing. They consist of simple sentences that focus only on the specific lesson. They differ from the existing sample sentences, which try to maintain grammar from all lessons. These drills were requested by several reviewers after the 1st draft was posted online.

I finished writing the drills yesterday, still 2 weeks before the editor becomes available. Due to the increased size of the book, I predict it will take him a month to edit. So conservatively, I have a month and a half before I need to do anything else regarding the book.

So what to do now? I definitely want to do some language learning, and here are some ideas.

1)  Start a new language from my hitlist. The possibilities are German, Italian, Vietnamese, Cantonese, Malay, Cambodian, Cebuano, Arabic, Hindi.

2)  Improve existing languages.

a)   Improve conversation in all languages by doing 3 lessons per day and using my conversation improvement technique for all of them.

b)  Work on my biggest weakness – reading Chinese characters. Specifically, improve Mandarin and Japanese reading by reading a ton and memorizing definitions and pronunciations of unknown characters and compounds.

3)  Start a YouTube channel for answering common questions about language learning that come up frequently.

I could add more to the list, but I think that is enough to think about. I’ll try to make a decision today.

In closing, I want to share a positive experience that I had last night. After a long bicycle ride with a friend, I convinced him that we should watch “Beck”, a Japanese movie about a new band competing in a battle of the bands. He is a guitarist, and open minded about different cultures, so I thought he would like it. The sad thing about Beck is that it keeps getting taken down from YouTube. We found a version, but it was only dubbed in Portuguese, no English. He only speaks English, so he asked if I could translate, and I did. Half the time I was translating from Japanese, and half the time from Portuguese. But I did that simultaneous translating for the entire movie, and he said he understood it. It’s nice to feel my studies have value.

I decided to go with 2) b above. According to the stats in the reading tool, here are my approximate average percentages of unknown unique words after I finish reading a passage:

Spanish 5%

French 5%

Portuguese 7%

Tagalog 10%

Swahili 15%

Russian 15%

Thai 15%

Korean 15%

Japanese 25%

Mandarin 35%

Keep in mind that these stats would look much better if they were for non-unique words. That is because a few hundred very common words, which I already know, make up a majority of the words in a passage, and are repeated many times. But the numbers above clearly show the problem I have with Japanese and Mandarin reading. To make it worse, without tools, it is often impossible to even pronounce an unknown word in those languages, since Chinese characters may give no clues.

So I am working on my greatest weakness – reading Mandarin. I’ve decided to focus on simplified characters, since that is the most common. I will need to do traditional later. I will also need to do Japanese eventually. My goal is to get the unknown words under 10%. I don’t know how long this will take, but I will dedicate at least an hour a day to this until it gets resolved.

My first day at it was yesterday. I read some of passage 1 in Mandarin Conversations. I read until I saw that I had accumulated too many unknown words, and stopped. Then I put the words into a word list and memorized them. L2 to L1, then L1 to L2. Finally, I put them in Anki. Today I did the Anki reviews, then I reread yesterday’s reading. It was much easier. Next, I continued reading until I had about 20 unknown words, and repeated yesterday’s exercise. I think I’ll limit it to about 20 words per day. I hope that after a few weeks I’ll be able to read a whole passage before I hit 20 unknown words. (I am listening to the passages too – I didn’t add that here to avoid cluttering up the reading exercise)

It took me 5 days to get through the first 6-minute conversation. Number of new words were as follows:

day 1 = 26

day 2 = 20

day 3 = 20

day 4 = 17

day 5 = 12

I started the second conversation today, and added 21 new words. I wanted to get past 1/4th of the way through the passage, and I did. I expect this passage will take no more than 4 days, the next one 3 days, and the next one 2 days. I also expect to be reading whole passages before I finish 10 total, and I plan to only read one a day from then until I finish them. Anki reviews are building up, but not bad so far; maybe 10 minutes. I’m going to document my procedure here, for possible future use:

1)  do Anki reviews

2)  listen to audio for yesterday and today’s reading, reading along casually

3)  read yesterday and today’s reading

4)  memorize the unknown vocabulary, whether it is truly unknown or I merely couldn’t read it, and put in Anki

5)  listen to today’s reading, reading along casually

6)  read today’s reading

7)  listen to today’s reading without looking at text, trying hard to pick up everything

I’m not saying this is the right way to do it, but it seems to be working quite well so far, so I’ll stick with it.

I’m about to start my travels again. To make a long story short, there was extensive water damage to my condo and it will take weeks to properly repair, so I decided to avoid the demolition/construction. I will go the Philippines first because I’m finishing up the book (still in editing) and I want a nice, low-pressure place to stay for a couple months to continue with this Chinese reading exercise in the afternoons. China and Taiwan are still closed due to the pandemic. I would have liked to have gone to Brazil for a month to practice my Portuguese, but that would mean I’d be doing spurts in two languages at once, which can be stressful.

I leave in a three days, I just finished the last day (day 20) of this exercise before travelling, and wanted to report on it. It may take longer that I thought. I’m on the fifth conversation, and it will probably wind up being a 5 day passage. I thought I’d be down to 3 or even 2 days per passage by now. I was hoping that by the end of the first month I’d be doing 1 passage per day.

conversation 1: 5 days   

conversation 2: 4 days   

conversation 3: 4 days   

conversation 4: 4 days 

There is some good news though. First, unknown (unique) words is in the mid 20’s for these passages. It was in the mid 30’s when I started doing this, which may have been because I was looking at harder materials. But regardless, it’s nice to know that it’s in the same difficulty range as Japanese.

The other good news is that my Anki reps aren’t spinning out of control. I’ve done 20 days in a row, except for one day off, and my reps are taking about 40 min/day. I wanted it be below 1 hr/day by the time I finished 1 month, and it looks like that would have happened even without this forced break coming up due to travel. I’ll be working on this for 2 months in the Philippines, then 2.5 months in Thailand, so hopefully that will be enough to finish it.  

I have a confession to make – I was wrong about something that I’ve believed for a long time. I believed that if you are going to systematically memorize and review vocabulary and grammar from content that you consume, it is always better to do L1 to L2 reviews. Actually, I have always done both L1 to L2 and L2 to L1 reviews just to be safe, but I believed that L2 to L1 were not very important, and the ability to understand the L2 items would happen naturally, provided I did L1 to L2. In my defense, if you are only interested in conversation and writing, I still believe this is correct.

But I consider myself to be a well-rounded language learner, and it turns out that L2 to L1 reviews are very important for reading. You might be thinking “so what – you are doing both anyway, so you are covered”. Reviewing both ways isn’t necessarily a bad idea, unless you are trying to learn to read as quickly as possible. L1 to L2 reviews require much greater recall effort than L2 to L1, imo at least twice as much, so doing reviews both ways takes at least 3 times as much time as just L2 to L1. And this hardcore recall work is something that you should limit when you study – doing too much makes people want to quit.

I may have never figured this out, since most languages have not required me to do intensive reading. With most languages, I was able to just read extensively in the reading tool, noting pop-up definitions of unknown words, and became comfortable reading in a few dozen hours. But recently I decided to fix the biggest hole in my proficiency of all 10 foreign languages. I decided to keep reading Mandarin intensively until I know more than 90% of new texts at first glance. When I started, I was at an appalling 65%, so extensive reading by itself was not cutting it.

When I started this exercise, I decided to try to memorize/review every new word that I came across, since I was using a very trustworthy source. Mandarin Conversations are 100% natural conversations, with very little if any strange vocabulary/grammar, so I am not concerned about usefulness. I tried to do 20 new words per day, both L1 to L2 and L2 to L1, and it very quickly became hard to stay on top of.

I have read many posts of learners disregarding L1 to L2 reviews when learning to read, but never believed it would work for me. Seeing over a month of my precious time slip away with very slow progress, I decided to make the leap and give it a try. I ditched the L1 to L2 reviews, and increased the load to 40 words per day. Surprisingly, even after doubling the word count, after a couple weeks of this, review time is significantly less than the previous method, and it feels less intense.

Bottom line, if you are just trying to improve you reading in a language that uses Chinese characters, doing L2 to L1 reviews only is an efficient way to maximize the number of new words you can assimilate per day.  

So why does this make me feel so good? Shouldn’t I be embarrassed about telling people to never skip L1 to L2 reviews? Sure, but it’s not worth dwelling on such things, and the big picture reveals something much more important. My poor reading skills in Mandarin and Japanese were the main reasons why I thought I’d never be able to reach advanced levels in all my languages simultaneously. I thought it would take thousands of hours, and wasn’t willing to sacrifice that time. Now I’m looking at a few hundred hours, so the dream of C1+ in all my languages is alive again!  

Time for an update. I have been in Thailand now for two weeks. I had a terrible case of amoebas (something I ate probably) the last couple weeks in the Philippines, and the swimming pool pump broke, which closed it down. After being sick and without exercise for two weeks, I was quite happy to leave Manila.

Although my goal was to finish it in Manila, the editor of Tagalog Lite was so busy that he couldn’t check it consistently during that time. But he has scheduled some time for it each week, and we are about halfway through the final check. It’s really looking good. He has now figured out a lot of complicated grammar issues that neither of us could explain very well before. When I say complicated, I mean it’s hard to explain, but still very high frequency.

I reached a bit of a milestone in my Mandarin reading spurt. I’m spending 2-3 hours a day on it, including reviews. Up until a week ago I was adding exactly 40 words per day, and only reading up to that point (plus re-reading the previous day’s segment). I would start in one passage, and finish in another. But a week ago I noticed the unknown words had dropped, and I made the decision to read a complete new passage every day, regardless of the unknown words. The first day was under 40, then I had 3 days over 40 (one was a whopping 54), so I wondered if I’d made the right decision. However, the last 4 passages were under 40, so I’m now sure things have settled down. My routine has simplified. Every day I do my anki reviews, read the previous day’s passage, read the new passage, memorize the new words and export them to anki.

Today was day 75 (not counting skipped days), and I finished passage 40. There are 100 passages total, so I will finish in 60 days at this pace. My reading has improved dramatically. My goal is to get the unknown words below 10%. I started in the mid 30’s, now I’m in the low 20’s.

I’ve also become interested in reading speed. In a Chinese Forums post, I saw people comparing reading speeds. Natives speak about 250 cpm (characters per minute), and this matches our Chinese Conversations. So in addition the 10% unknown words, I’m setting a goal of half native speed. Passages are 6 minutes long, so my goal is 12 min, first read. I tested it a while back, and I did 18 min, but that was my third read. I will do more testing as I go along.

Here's an update, because my italki Korean teacher decided to record our lesson. It was kind of a nice surprise – I don’t know why she did it, but I saw she was recording it and asked if I could have a copy. It turns out that videos are downloadable by all on Skype, so I downloaded it, cut out a 5 min chunk that wasn’t 100% about languages, and posted it for fun. I’m speaking quite slowly these days it seems. To summarize, I studied Korean really hard for 1 year, spent 1 month in Seoul, the following year spent 1 month in Pusan, and the third year spent one month in Seoul again. Then came the pandemic, so I missed 3 years of travel, and I only maintained it sporadically. I review it about twice per month now, and it’s my worst language. Even after all that though, I speak it at a slow intermediate level, so I won’t complain. I hope to do a spurt in it next year, and hopefully never fall down to this level again.

In other news, today was day 121 (active days), and I finished passage 85 of Mandarin reading. I have made huge improvements since the beginning, but I don’t think I’m going to hit the 10% unknown mark by passage 100, because I'm probably at around 20% now. I will probably ease off after passage 100 though, because it's been a long hard pull.

Today I reached two important milestones. First, I finished passage 100 of Mandarin Conversations, officially ending this spurt. Some quick stats – 100 passages in 137 (active) days; 4017 vocabulary items; about 500 hours (including Anki reviews) total study time. My goal was to drop my Mandarin unknowns from 35% down to below 10%. I did not reach that goal. If I look at my stats, they average out in the high teens, but that’s because I haven’t gone back to change all the words I didn’t know at the beginning to “known” now. I’m guessing my true unknown level is about 15%. I’m really happy with this, and it means that I read Mandarin about as well as I read Swahili/Russian/Thai/Korean. I have mixed feelings about stopping the spurt. I figure I could reach my goal and really hit it out of the ball park in another 200 to 300 hours. And there is the fear of losing my level too. I think I’m going to take a few days off, then keep reading new passages twice a week for a while to see how that goes. I’ll keep doing all the anki reviews for the time being, but will eventually delete all cards over one month old at some point. The ultimate goal is to be able to read, consistently and comfortable, only as often as I maintain the language, meaning once every 10 or so days. But the interim plan above should safeguard my level, and give me the time I need to do the stuff required for the milestone below.

The second milestone I reached is even bigger, since it marks the beginning of the end of a 3 year project. As of today, we have finally finished going back and forth, checking and perfecting Tagalog Lite. The ball is now back in my court, and I can work on it full time. I need to update all the tables, rewrite the intro, hire a voice actress, and post the book online. I also plan on creating downloadable anki decks, so I have a busy month or so ahead of me!

leosmith wrote:

Today I reached two important milestones. First, I finished passage 100 of Mandarin Conversations, officially ending this spurt. Some quick stats – 100 passages in 137 (active) days; 4017 vocabulary items; about 500 hours (including Anki reviews) total study time. My goal was to drop my Mandarin unknowns from 35% down to below 10%. I did not reach that goal. If I look at my stats, they average out in the high teens, but that’s because I haven’t gone back to change all the words I didn’t know at the beginning to “known” now. I’m guessing my true unknown level is about 15%. I’m really happy with this, and it means that I read Mandarin about as well as I read Swahili/Russian/Thai/Korean. I have mixed feelings about stopping the spurt. I figure I could reach my goal and really hit it out of the ball park in another 200 to 300 hours. And there is the fear of losing my level too. I think I’m going to take a few days off, then keep reading new passages twice a week for a while to see how that goes. I’ll keep doing all the anki reviews for the time being, but will eventually delete all cards over one month old at some point. The ultimate goal is to be able to read, consistently and comfortable, only as often as I maintain the language, meaning once every 10 or so days. But the interim plan above should safeguard my level, and give me the time I need to do the stuff required for the milestone below.
The second milestone I reached is even bigger, since it marks the beginning of the end of a 3 year project. As of today, we have finally finished going back and forth, checking and perfecting Tagalog Lite. The ball is now back in my court, and I can work on it full time. I need to update all the tables, rewrite the intro, hire a voice actress, and post the book online. I also plan on creating downloadable anki decks, so I have a busy month or so ahead of me!

Congrats on all your efforts!