zenmonkey's multilingual adventures of a traveller

Continue or start your personal language log here, including logs for challenge participants
User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2528
Joined: Sun Jul 26, 2015 7:21 pm
Location: California, Germany and France
Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 7030
Contact:

zenmonkey's multilingual adventures of a traveller

Postby zenmonkey » Mon Jul 27, 2015 8:14 am

Intro

So - I’m once again uncertain about logging and getting lost in the time-suck that a forum can be. But, I also see this as immensely valuable, motivating and a good companion to language learning. I’ve kept logs on paper, HTLAL, a personal blog (my wordpress blog died an ugly death) and other places I’ve forgotten …

I hope this blog will help me think about method and language learning. As such I will try to remind myself not to just check off the lessons done and what not but to actually work on the ‘meta’ of the learning - the why, hows and ways.

As an intro post, I’ll keep it brief before I go and dig out some of the things I need from HTLAL, etc.

Language relevant background:
  • I’m an old fart in the language learning arena - I started at an early age and was on HTLAL since 2006.
  • I live in between Germany and France and travel almost weekly to one or the other.
  • Born in France, I grew up in Mexico and California and spent many years in the US where I studied at the university level. I moved to France, did graduate work there and lived for over twenty years. I've spent 8+ years in Germany or traveling between both.
  • My family background and travel has brought me into contact to different levels with a lot of languages - Italian, Portuguese, Polish, Yiddish, Hebrew, Chinese, Arabic, Ladakhi to name a few.

Current focus - 2022
Learning: Persian/Farsi A0 - started Dec 20, 2021 from nothing.

German: Would like to maintain and strengthen from my B2 level.
French, Spanish: Maintenance and practice of core native-level.

Dabbling in Hebrew, Yiddish and others...

Main Language Tools - Currently Using - updated 01.2022

Actively working only on Farsi/Persian with:
  • John Mac - Teach Yourself Books - Modern Persian
  • D. Halbout, M. H. Karimi - Assimil - Le persan
  • An Introduction to Persian - Wheeler M. Thackston
  • Daily flashcards with Anki

Language Tools - old - from 2019

  • Hebrew - HebrewPod101 also using italki with a teacher, Hebrew from Scratch, Anki, Memrise, FSI, ASSIMIL and Living Language, Shtisel
  • German - duolingo, variety of books, DW site, ASSIMIL (the advanced module) and various test prep books. Now at level that I'm working with a lot of native material
  • Tibetan - On and off, Italki teacher
  • Setswana
    Memrise:
    -- Intro to Setswana completed
    -- There are 2-3 other courses but I'm not yet actively working with them
    Anki:
    -- Setswana - unisa Deck (of my making) w/sound
    -- Setswana - Peace Corps Deck - rdearman - w/sound
    -- Fluent Forever Deck (meh, can't recommend, hard memorisation of animals you don't need)
    Books:
    A Course in Tswana
    Peace Corps: Intro to Spoken Setswana
    Peace Corp : There is no word for grammar in Setswana
    Kauderwelsch Setswana
    Apps:
    uTalk modules
  • Yiddish - only incidentally (2019) - primarily books, ASSIMIL and Memrise course

Lyric Translation Site - not currently using

Fast Link to My Log Posts
Hebrew Log Entries
German Log Entries
Setswana Log Entries
My Hebrew Resource List
My Anki subs2srs guide

Thanks for reading.
Comments, corrections are always welcome.


Now, to go sort some things ...
Last edited by zenmonkey on Mon Jan 17, 2022 11:59 pm, edited 40 times in total.
3 x
I am a leaf on the wind, watch how I soar

User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2528
Joined: Sun Jul 26, 2015 7:21 pm
Location: California, Germany and France
Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 7030
Contact:

German Objectives and Current Material

Postby zenmonkey » Mon Jul 27, 2015 8:49 am

Somehow I can't seem to find my old post on the core German material I had. Oh, well, probably wasn't that well organized or relevant.
Here is an online of my objective and tasks for German - The objective hasn't changed that much in years.

GERMAN
- OBJ: Move from Intermediate to Advanced German. For me the level of fluency I want can be summed into a couple simple phrases.

I want:
- Reading - comfortable reading novels, e-mail and newspapers with a level that I only need to look up 2 to 3 words per page.
- Writing - capable of expressing myself clearly and without faults in an eMail or letter or forum post.
- Listening and comprehesion - understand most films, radio and conversation in a group.
- Speaking - clearly understood in everyday interactions, capable to give a presentation in German, hold a meeting, give directions.

Current material and tasks
  • Anki flash cards - Continue with my old list, assure that I use them at least 5 days a week.
  • Duolingo - Continue using this every day.
  • Reading - Currently Vier kurze Krimis by Doyle in a bilingual edition - at least 20 minutes each day.
  • Listening - A podcast, radio, movie or song a day.
  • Production - Right now, just going to use Duolingo output to cover this - I recognize it is weak but I'm giving myself permission on that. For now.

Too much? :shock:
1 x
I am a leaf on the wind, watch how I soar

User avatar
Jar-Ptitsa
Brown Belt
Posts: 1000
Joined: Sun Jul 19, 2015 1:13 pm
Location: London
Languages: Belgian French (N)

I can speak: Dutch, German, English, Spanish and understand Italian, Portuguese, Wallonian, Afrikaans, but not always correctly.
x 652

Re: zenmonkey's multilingual adventures of a traveller

Postby Jar-Ptitsa » Mon Jul 27, 2015 9:43 am

Hi zenmonkey :)

I will read your log.

Viel Erfolg mit deinem Deutsch. ich denke, dass du damals damit angefangen bist als wir auf htlal geschrieben haben, oder? Wie findest du es in Deutschland zu arbeiten? Wie du weißt, sind wir hier in England mit der Arbeit meines Vaters, und es ist irgendwie toll im Ausland zu sein, aber in London wohnen auch ganz viele Franzosen.
1 x
-w- I am Jar-ptitsa and my Hawaiian name is ʻā ʻaia. Please correct my mistakes in all the languages. Thank you very much.
: 1 / 50 Spanish grammar
: 5 / 50 Spanish vocabulary

User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2528
Joined: Sun Jul 26, 2015 7:21 pm
Location: California, Germany and France
Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 7030
Contact:

Re: zenmonkey's multilingual adventures of a traveller

Postby zenmonkey » Mon Jul 27, 2015 2:01 pm

Danke Vogltje,

Während bin ich nicht mehr ein Anfänger auf Deutsch, leider meine schriftliches Art stinkt immer. Ich lebe und arbeite in der Nähe von Frankfurt. Ich war in München vor und liebte die Region.
1 x
I am a leaf on the wind, watch how I soar

User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2528
Joined: Sun Jul 26, 2015 7:21 pm
Location: California, Germany and France
Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 7030
Contact:

Cloze Karaoke - Anki Song Bits

Postby zenmonkey » Mon Jul 27, 2015 2:51 pm

I've been wanting to use music and spoken poetry as study tools for German because this is one area where I feel that understanding the subtleties of the language is necessary - it will push me further and I greatly enjoy word play.

I may have bitten more than I can chew.

Influenced by emk's post on subs2srs here and what AJATT calls Massive Context Cloze Deletions (MCD*) I decided to look a little look at how I could make learning cards from music - either videos with lyrics or songs. And I looked at few tools that are out there and stumbled across something that might work for me.

First, quick background definitions so as to avoid confusion.

- MCDs - Are just longer sentences or even paragraphs placed in a flash cards system with missing words (Cloze deletions) where context can help create the understanding of what is missing. For example: El joven ____ el agua ahora. (The boy now ____ the water) It becomes clear what needs to be put in there and one can use it to learn words or verbal forms, etc... (Honestly I don't see the need for a new term MCD, I'll stick to Cloze)

But reading on the MCDs and what was being done with short subtitled movies got me interested. Can I do this with music?
(Obviously yes, but is it fun and useful?)

I ended up using Audacity to take a song and slice up into small MP3 files so that I could put the song lyrics in an Anki deck. I'll try this for several songs. Each card is a long cloze test. With Audio!

So what I have is an Anki Card that is build like this:

Screen Shot 2015-07-27 at 1.33.40 PM.png


With holes - each time the card shows I hear the bit of music and need to fill in what is missing as well as review it for understanding.

Screen Shot 2015-07-27 at 1.35.24 PM.png


Quick note - In order to create a Cloze deletion in Anki just select the text to be deleted and press CTR-Shift C (or cmd-shift C on a Mac).

I pulled lyrics off of either YouTube comments or lyricstranslate.com.

My first attempt was with the Fantastischen Vier - Sie ist Weg.



Possibly this is where I bit off too much. The text is great but as a 'rap' it is a bit hard to get, certainly pushing my envelope.
Perhaps to paraphrase the song "schätze bin ein bißchen hochgeflogen ungelogen und hab dabei {mit mir} mich selbst betrogen"

(we need a strikeout text format here)
You do not have the required permissions to view the files attached to this post.
4 x
I am a leaf on the wind, watch how I soar

User avatar
emk
Black Belt - 1st Dan
Posts: 1620
Joined: Sat Jul 18, 2015 12:07 pm
Location: Vermont, USA
Languages: English (N), French (B2+)
Badly neglected "just for fun" languages: Middle Egyptian, Spanish.
Language Log: viewtopic.php?f=15&t=723
x 6330
Contact:

Re: Cloze Karaoke - Anki Song Bits

Postby emk » Mon Jul 27, 2015 3:20 pm

zenmonkey wrote:Possibly this is where I bit off too much. The text is great but as a 'rap' it is a bit hard to get, certainly pushing my envelope.
Perhaps to paraphrase the song "schätze bin ein bißchen hochgeflogen ungelogen und hab dabei {mit mir} mich selbst betrogen"

Cool idea! I've actually made quite a few sound cards from MC Solaar's rap songs, and here's how I did it (more or less; the technical details have changed over the years):

  1. I used a tool similar to Subtitle Edit to make a separate subtitle for each line of the song, and aligned them.
  2. I asked subs2srs to generate a deck with 3 subtitles per card: one line before, the current line, and one line after. Basically what I did with Avatar for Spanish, but with music instead of video.
  3. I imported the cards into Anki, putting the song name and the audio clip on the front of the card, and the subtitles on the back of the card.
  4. When reviewing the cards, I passed them if I could passively understand the middle line of the three.
I find the trick with these audio cards is to keep them (1) short enough that I can deal with every little section of audio separately, and (2) as easy to make as humanly possible. As a general rule, I prefer very easy cards—they seem to provide 95% as much learning benefit as hard cards, for 10% of the effort, so you can just do some more and come out way ahead. If you're like, "Ugh, this card is too hard," I've personally found that it's basically always safe to make it easier, or to just delete it.

This format seemed to work very well: Within a month, I had very solid listening comprehension of my favorite songs, and I could sing along with almost all of the lyrics, even the fast ones. It turns out that really fast "spoken word" songs are an excellent warmup drill for speaking, as long as the accent isn't too far off from your own! This is the closest I've ever managed to come to "shadowing," because I can't seem to do it using the normal techniques for some reason.
3 x

User avatar
Jar-Ptitsa
Brown Belt
Posts: 1000
Joined: Sun Jul 19, 2015 1:13 pm
Location: London
Languages: Belgian French (N)

I can speak: Dutch, German, English, Spanish and understand Italian, Portuguese, Wallonian, Afrikaans, but not always correctly.
x 652

Re: zenmonkey's multilingual adventures of a traveller

Postby Jar-Ptitsa » Mon Jul 27, 2015 3:28 pm

zenmonkey wrote:Danke Vogltje,

Während bin ich nicht mehr ein Anfänger auf Deutsch, leider meine schriftliches Art stinkt immer. Ich lebe und arbeite in der Nähe von Frankfurt. Ich war in München vor und liebte die Region.


ich glaube nicht, dass es stinkt :lol:

Ich habe gehört, dass München sehr schön ist. Frankfurt ist bekannt weil es die Flughafen Stadt ist, auch Finanzzentrum, aber sonst weiß ich nichts davon.

Fehler machst du schon auf Deutsch siehe ich, aber besser wenn ein(e) Muttersprachler(in) etwas mehr dazu sagt. Josquin ist dabei, er ist Deutsch und sehr nett. wenn du Verbesserungen möchtest, meine ich dann.

Toll, dass deine Tochter Architektur macht. Möchte sie es danach studieren? ich habe als Gärtner-assistentin gearbeitet, und ich interessiere mich viel in die Architektur der Häuser, nicht nur der Gärten. es kombiniert so wunderbar Kunst mit Mathe, ästhetische und funktionelle auch zB Geologie und die Geschichte.
1 x
-w- I am Jar-ptitsa and my Hawaiian name is ʻā ʻaia. Please correct my mistakes in all the languages. Thank you very much.
: 1 / 50 Spanish grammar
: 5 / 50 Spanish vocabulary

User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2528
Joined: Sun Jul 26, 2015 7:21 pm
Location: California, Germany and France
Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 7030
Contact:

Re: zenmonkey's multilingual adventures of a traveller

Postby zenmonkey » Mon Jul 27, 2015 10:10 pm

After some pain in using Subtitle Edit (it would not open my video and did not accept the version of VLC, etc... I looked to see if an Mac based program existed - low and behold - a massive improvement, in my view, over the Subtitle Edit program - I was able to subtitle a 5-minute music video in about 15 minutes with exact start/stops on audio.

Moved that and the video file over to the PC and there generated the extracts using subs2srs (my settings included a line before each and after each text to be studied and the audio (whoops, a mistake) and voila - working German song cards. Thanks EMK for the hints. Finally got something working.

However, I think that because I did not move the media into the Anki User media folder BEFORE importing the cards - it imported the file names as text and I've had to manually correct that - for fifty cards. Meh. Waste of time.

And I think that might be an issue. In general this is is very labor intensive for me. Even if I think I can cut this by 80%. Perhaps if I did this with a video and L1 subtitles too - or if I can find a few subtitle files that I don't have to generate then maybe this will be part of my tool chest.

Anyway, the quality of the audio cutting and the cards themselves is excellent. This time, to save space, I only did audio. But I can see how much FUN video cards might be.

Screen Shot 2015-07-28 at 12.06.37 AM.png


Looking back on my effort with the first song, there is a world of difference. I think I'm going to toss those cards out - they are confusing and next to useless.

Oh, and today's song to Anki counts for at least 3 days. Sheeez!
You do not have the required permissions to view the files attached to this post.
3 x
I am a leaf on the wind, watch how I soar

User avatar
zenmonkey
Black Belt - 2nd Dan
Posts: 2528
Joined: Sun Jul 26, 2015 7:21 pm
Location: California, Germany and France
Languages: Spanish, English, French trilingual - German (B2/C1) on/off study: Persian, Hebrew, Tibetan, Setswana.
Some knowledge of Italian, Portuguese, Ladino, Yiddish ...
Want to tackle Tzotzil, Nahuatl
Language Log: viewtopic.php?f=15&t=859
x 7030
Contact:

Anki song cards - my short guide

Postby zenmonkey » Tue Jul 28, 2015 8:58 am

Finished correcting the imports this morning and I think I need to post this as a short guide on how I built these cards as a reminder for myself on how to avoid some time consuming pitfalls - hopefully it might help others.

1) get video of a song off youtube (btw - fair use, personal and educational of music I already own) using clipconverter

2) get song lyrics and translation from lyricstranslate.com - note that for some songs, even transliterations are available.

3) convert to subtitle file using Aegisub (on the Mac) or Subtitle Edit (on the PC). Personally I found Aegis MUCH easier to use - but there is definitely a learning curve there.

4) use subs2srs to slice and dice the video and subtitle files to media and an import file for Anki.

The time consuming parts:

1) subtitle edits - this just takes a long time, but it is faster when I can do it in a single pass - hence my preference for Aegisub - it just maintained sync better and was easier to create the timing cues. I only imported one language - I hope importing the next one is just a copy paste over the existing text files ... otherwise. Argh. Might paste both into Excel first to assure lines match up before creating edits. For languages that need transliteration - like Hebrew for me - I create several subtitle files (2 or 3) and using Aegisub it requires one pass, then add the other languages by copying the next text into my first sub.

1b) Make sure that the subtitle that you are using is long enough to be of interest. Cards with one/two words of dialogue aren't very useful. (See EMK's post and the links on making longer dialogues/merging subs)

2) creating srs data - the settings are important here - particularly adjusting if you want the lines leading/trailing the dialog you are keeping. Click on the "Advanced Subtitle Options" and select the number of lines leading/trailing. My settings for the first good attempt were 1 leading and 1 trailing. However I think this just clutters the answer side of the card. No need to include additional audio clip, I think, what are you going to do with it? In future attempts I will only keep the lead line and use the remaining space of the card for the L1 translation.

For files with L1, L2 and the transliteration - I open the srs output file in Excel and paste the missing column for the transliteration. So I have L1, L2, transliteration. It would be very easy to have an L3 or L4 import.

3) importing - Each import needs to configured just right in Anki, matching the import fields to the proper Anki data field. BUT first move all the media to the right media folder - it's in Documents/Anki/User 1/collection.media. If you don't move the media first the import will consider the snapshot data as text and NOT a pointer to a media file.

READ the Import guide - I lost so much time fiddling on with this. It can be found at Importing Into Anki

I opened my import file in Excel to review the columns as I set this up. My order, more or less: Tags, Sequence Marker, Audio, Snapshot (or video in the future - check to make sure video is iPhone compatible when exporting), Expression field (for the L2* sub), Meaning (for L1 answer), Reading (for the prior sub line in L2). I feel that having a trailing song line does not add anything for me but clutter. I prefer to have a translation there. I'll be able to give the card an OK when I understand and know the meaning of the line.

Given that Excel exports an UTF16 file, when needed I also convert the file to UTF8 using http://www.fileformat.info before importing into ANKI.

I played with the look of the feel of the import card format before importing to create something than met my needs - larger and more standard fonts and I deleted all the funky font choices.

After importing the template card in Anki, I went to 'Browser' found the card and edited it:

Screen Shot 2015-07-28 at 10.53.08 AM.png


Once I felt the look and feel matched what I wanted then I did the import.

Then go back to following the steps in import guide. Do make sure HTML is on during the import.

The result is just perfect for me.
Screen Shot 2015-07-28 at 9.30.46 AM.png


or

Screen Shot 2015-08-30 at 22.13.45.png



*L1/L2 - I alway mean source language (L1) and target or learning language (L2)
You do not have the required permissions to view the files attached to this post.
Last edited by zenmonkey on Fri Jun 09, 2017 3:23 am, edited 3 times in total.
5 x
I am a leaf on the wind, watch how I soar

User avatar
emk
Black Belt - 1st Dan
Posts: 1620
Joined: Sat Jul 18, 2015 12:07 pm
Location: Vermont, USA
Languages: English (N), French (B2+)
Badly neglected "just for fun" languages: Middle Egyptian, Spanish.
Language Log: viewtopic.php?f=15&t=723
x 6330
Contact:

Re: Anki song cards - my short guide

Postby emk » Tue Jul 28, 2015 9:20 am

Great guide! It definitely helps, I think, to work out the kinks in your process.

zenmonkey wrote:2) creating srs data - the settings are important here - particularly adjusting if you want the lines leading/trailing the dialog you are keeping. Click on the "Advanced Subtitle Options" and select the number of lines leading/trailing. My settings for the first good attempt were 1 leading and 1 trailing. However I think this just clutters the answer side of the card. No need to include additional audio clip, I think, what are you going to do with it? In future attempts I will only keep the lead line and use the remaining space of the card for the L1 translation.

Yeah, I wasn't clear in explaining why I usually included one line before and after. It's actually one of my speed optimizations for card creation. For me, the most time-consuming part of cranking these cards out is getting really precise alignment for subtitles to audio. For speed, especially when dealing with entire TV episodes or movies, I work with much sloppier alignment, and then patch it up as follows:

  • Add ~1.5 seconds of audio padding before and after each clip.
  • Add an extra line of dialog before and after each clip, but grey, and in a smaller font.
My rationale: Even if I do I really quick and dirty subtitle alignment, this will allow most cards to be salvaged, with minimum labor—the padding means that I probably have the entire line there somewhere, and the extra context helps make sense of the padding, and of lines that are just weird grammatical fragments.

But if you have really accurate audio alignment, you can reduce the padding and even drop the context lines if you want, and nearly all cards will still be usable. It's just that I've optimized my process for dealing with TV episodes and films, where there are hundreds of cards, and I where only want to clean up the worst alignment errors.

Also, I've experimented with both video and audio + an image, and decided that video really added nothing for me personally, except much longer download times.

Anyway, once I got the kinks out importing 900 cards at a go, the next lesson I had to learn was aggressive card deletion! Even a few bad cards can really destroy the whole Anki experience (because they keep getting failed and recycling), and hey, subs2srs cards are cheap—I can always make another 900 or so with a few hours work. :-) So if I look at a card and groan, "Oh, no, not again," it's gone.

This should cover most of the non-obvious tricks and traps I had to figure out the hard way, and that I've seen other people stress over. :-)
4 x


Return to “Language logs”

Who is online

Users browsing this forum: guyome and 2 guests