30:30 Vocabulary Challenge - July 2023

[untrucdeouf]

My favourite words from the challenge aren't nearly as fun as @coldrainwater's -- as a beginner, most of my list is comprised of words that are prosaic, useful, and dull -- but here are a few I especially enjoyed learning:

paboreal - peacock (from pabo [turkey] + real, lit. 'royal turkey')
tumba-tumba - rocking chair (from the root word ‘tumba’, to fall down)
tiktilaok - cock-a-doodle-doo
ngiyaw - meow
isip talangka - crab mentality

It's always endlessly fascinating to me to learn what onomatopoetic animal sounds are in other languages.

Would be curious to see other's favourites from this month as well!

[rdearman]

Tomorrow, the 31st is scoring day! Don't forget to take your test and post the results here! Also, because August is another 31 day month, and to keep this thread from getting too cluttered, I opened a new thread for people to sign up for next month.

https://forum.language-learners.org/vie ... 21&t=19316

[rdearman]

Scoring day.

A very disappointing 18 out of 900 for me. It should have been closer to 50, but my spelling let me down, for example I could remember how to spell 월 (month) and so that let me down for 12 words, Jan, Feb, etc. But given that I gave up about 1/2 way through, I'm surprised I remembered (and spelled correctly) as many as I did.

[Anya]

146 out of 383 (TL-»NL)

[raikiro]

The time has finally come that my soul is free from this burden.
I will say I liked the challenge conceptually, but getting done with it was a test of my discipline. But hey, at least I got some statistics now. Beware, long post ahead.

What I did:
Iversen-style word lists, with two repetitions: one day and then three days after the initial list. The lists were all written by hand.
Three languages (Japanese, Spanish, Russian). For the first three days or so I tried to learn all the corresponding kanji as part of the word, but it took too long and there is no real benefit, as I never have to write kanji by heart anyway. On the recognition part I still had them and tested myself on them, but for production I counted it as a pass when I could remember the reading and meaning.
On the last day, all three languages were randomized and then split into two subgroups for testing: recognition and recall. Initially I asked on this thread if it was okay to just do recognition, since I am only interested in reading and not conversation, but I decided to test myself on both to compare the results. So I had 300 words per language, split into 2x150 for both directions.
I did not use pre-made lists, all the words were taken from intensively reading books. For Japanese and Spanish I took the novels I was reading already extensively anyways, and continued with them. For Russian, which is not good enough to read native-level novels with some degree of understanding, I used a graded reader with short stories, so the methodology was pretty much the same for all.
I did not do any re-reading of the text. Under normal circumstances I would have re-read the Russian stories at least, but it would have made it difficult to compare to the other two. There were no other means of repetition outside of the scheduled two rounds, and words that were repeated naturally during my reading. I did not look through the lists again.

Time-related statistics:
I recorded my times for reading + making the lists, studying the word lists, and review phases 1 and 2, for each language individually.
All together I spent 41:33 hours on the first 30 days, and another 104 minutes for the testing today.
Of those 41 hours, 18.5 hours were reading: 8 for Japanese, 5 for Spanish, 5.5 for Russian.
12 hours were spent on the initial study of the word lists: 4:38 for Japanese, 3:15 for Spanish, 4:11 for Russian.
Another 11 hours I needed for reviews 1+2. For review phase 1 I took about 1:20 for both Spanish and Russian, and 1:40 for Japanese. In phase 2, Japanese and Russian were quite close with 2:32 and 2:24 respectively, and Spanish only taking 1:43.
Phase 1 was recognition, with phase 2 being production. So it makes sense that phase 2 took longer.

So if we disregard the times for reading, because many worked with pre-made lists, it would have been about 23 hours total. However, it could be argued that learning words in context might have made a significant enough difference in retention to justify the additional time spent.

Testing results:
As stated before, I split into blocks of 150 words to compare languages and recognition vs. production.
Japanese: production 35/150 (23.3%), recognition 69/150 (46%)
Spanish: production 17/150 (11.3%), recognition 65/150 (43.3%)
Russian: both 40/150 (26.6% each)
It was to be expected that recognition would be easier, and 46% and 43% are actually not bad, given that I only reviewed twice and then left them alone.
Unexpected is that Russian had the exact same amount of words retained both ways - it was the best language in terms of production, but the worst by a mile in terms of recognition. Not quite sure if I should be happy or sad about this
Another unexpected result is that Spanish was pretty bad for production. I would have expected it to be much better, given that it is the easiest of the bunch. Then again, a big reason why I perceive it as easy is because of the large amount of shared vocabulary with known languages, but those words I did not include in my lists because I already understood them while reading. That probably played at least a part in it.

This brings us to a total of 266 out of 900 words learned. Learned as in: I can write down the corresponding translation, and for Japanese pronounce the characters correctly. Whether or not I can actually spontaneously use them, understand them in a bunch of different contexts, etc. is another question entirely.
Just in terms of numbers, 266 words learned within one month seems nice. It's most likely more than I would have learned otherwise. If I consider the time spent though, 266 words in 41.5 hours is a mere 6.4 words per hour. Hypothetically, if I disregard the time spent on reading and assume using a pre-made list, it would have been 23 hours, or 11.5 words per hour. Better, but still not a particularly efficient way to use my language learning time.

Conclusions:
The retention for my recognition in Spanish and Japanese is quite good, given that I just threw some words at my brain, reviewed them twice, and then ignored them for the rest of the month. Production definitely needs more work than what I did, which is unsurprising.
For Russian, I am not quite sure. I am still at a point were especially longer Russian words just don't really 'make sense' to me, so they are very hard to remember because they all sound the same and feel kind of random. Another strategy would likely have served me better.
Because I spent so much of my language time studying, I read way less than I usually would have. Had I read more, I also would have seen the words repeated more frequently within that time. I already expected this, and it is the reason why I do not enjoy intensive reading very much.
Alternatively, it might have been beneficial to study the words first with word lists and then put them into Anki. The problem is, I do not like Anki and try to avoid it nowadays.
Will I do the same again? Absolutely not. It was not a very enjoyable experience overall. I liked the reading, though I prefer extensive reading where I only look up words occasionally. The only reason I pushed through is because I was committed to the challenge and I wanted to have some actual results and statistics to analyze. The review rounds were very demoralizing, some were really good and other days I remembered only very few, despite my efforts to keep my strategy, time and circumstances the same.
I might try another 30:30 challenge in the future with a different strategy that includes something like re-reading short stories or something similar that is more in line with what I usually do. Unfortunately it is more difficult to quantify the amount of words learned 'passively' through reading. But for the next challenge in August I'm out - I really need a break from words and lists for now

[Mista]

Test result: 471/900 (tested for production of TL)

I don't feel I got enough time for repetition, so I want to keep on practicing these words and test again in two weeks.

[coldrainwater]

My results show 828 out of 900 for Latin.

Thanks raikiro for posting so much detail and analysis. Very interesting read. I don't have study times myself, but will try to provide a moderate amount of detail otherwise to the best of my recollection. I re-sorted the final exam tab here with missed items shown at the top. A primary/foreign key is included to reference the initial list and can be found in column A.

Testing Technique
I tested passive recognition only TL->NL and typed quite a few of the answers directly as planned using the comma-delimited list that I had massaged ahead of time for that purpose. For the remainder, I locked in my best guess mentally, checked the answer and marked it either right or wrong right then based on whether what I had was sufficiently synonymous with the stated answer. For the majority of incorrect answers, I just drew a complete blank and no amount of coaxing seemed to bring back a definition in the moment. There were many that I had to struggle with for upwards of a minute or more before the right answer dawned on me. The test took multiple hours to complete and was quite fatiguing.

Initial Learning
I researched the etymology of each term as a primary means of driving the initial word into memory, mainly using Wiktionary. When that wasn't available, I added minimal Latin context such as a short phrase or word surrounding it as a strong hint with mixed results. I made a few ad hoc, transient SRS lists to assist with this process. Those lists tended to vanish within a few short days due to my own tendency to delete terms that I felt confident around.

Review
For the most part, I used the wordlist itself with hidden answers, randomized. I removed from the list in an iterative fashion any term I thought I could remember on test day to avoid wasting review time. I typed out answers occasionally/initially but quickly learned that this is very fatiguing and that I could more quickly subvocalize the answer than keyboard it. I ended up focusing mostly on leeches, which amounted to 200+ terms at any given time. For no particular reason, I did more reviews around the 10, 20 and 30-day mark, with short, stringent SRS intervals (sometimes a mere 3-6 hours between reviews), focused entirely on the terms I had trouble with and ignoring as many others as possible. Two days before the exam, I did a leisurely Saturday review, just looking at words and definitions and spending time with the list. I then slept on it, did the 30 new words on Sunday and took the test on Monday. As a rule, I almost never study heavily the day prior to an exam.

Other notes
The long comma-delimited lists were less helpful than items with short, succinct definitions, so brevity will be my future focus. Lesson learned. The process of making the list itself before the challenge was also of high value and involved reviewing a full Latin dictionary. I really like the overall challenge and it seems like a very good way for me to introduce new languages, either as serious study candidates or for broad exposure. It had significant value for me as a stand-alone exercise.

[untrucdeouf]

Congratulations everyone for making it through! I've really enjoyed reading through everyone's experiences with the challenge.

My result: 555/900 (production of target language, Tagalog)

By far the most challenging part of this challenge was today: I'm out of town and realized I'd forgotten my laptop charger and wouldn't have enough battery left to get through the test. Luckily I had access to a printer so was able to throw all the English-language definitions into Excel, shuffle them into random order, print out all 900 definitions, and do the test on paper. Very much unplanned, but by the end of the endurance test I'd kind of stockholm syndrome'd myself into believing it was all for the best since it was a chance to test myself in a different medium than the one I'd studied in (honestly there probably isn't that much value in it, but sometimes I do occasionally get the feeling when doing SRS that I'm just recognizing the card rather than truly connecting the answer to the meaning).

What I ended up doing was going through the list with a pink pen and answering only the words that I knew immediately without having to pause and think, and then doing a second pass with a blue pen for words I had to think about or was kind of guessing.

Pink word score: 512/555
Blue word score 43/85
Unattempted total: 260

For the pink words I got wrong, most were pretty close -- they were typically wrong due to small spelling errors or leaving off an affix. Some of the blue words I got wrong were close, but others were totally in a different universe. I consider the 512 score to be the most accurate representation of words I "know" since I was able to produce them accurately without hesitation. (This is not to say I could do so in conversation though!)

Marking the test was hell -- without a laptop I had to look up each word separately via my Anki phone app to check it. The terms were all in random order so it had to be done one by one. I really do not recommend this.

I wanted to do more extensive analysis of success rate based on days of the month, but I'll have to wait for that and throw it in my personal log. Anecdotally I definitely struggled more with words from the last 10 days and felt like words from the first 20 days of the challenge were much easier to recall. I still think I'd have better success if I tried to learn 900 words in a month with no specific daily minimum or maximum. I also wanted to test myself on passive recall but I just can't face it today! Maybe I'll do it in a week or so to check post-challenge retention.

My ambitious goal was 700 terms and I fell short by 145. I still think 700 is totally within my reach. I faced a few challenges and made a few errors this month that I think limited my success:

1) My preferred vocabulary learning app of choice (Ulangi), which I've had a lot of success with in the past, went offline literally on day 1 of this challenge. The developer archived it on GitHub on July 1st. Incredible timing. So I wasted a bunch of time hoping it would come back online, finally accepting it wouldn't, and having to set up a new SRS system that worked for me (this ended up being Anki with some JS modifications). The time would have of course been better spent on the actual challenge.

2) I made my word lists from scratch each day. Ultimately I think this was not a good idea, because it's a time-consuming process, and on some of my busier days I spent almost all my study time just making the list. This time around I plan to have it done entirely within the first few days of the month.

3) I decided to switch my goal from passive recognition to active production partway through the challenge. I'll be planning to do a TL production challenge from the jump next month.

4) I was initially trying to memorize all of the words with their stresses marked. On the one hand this was probably good for ingraining correct pronunciation, but I noticed I was having a little trouble recognizing words in a written context because I'd learned, say, 'bahagyâ' but it would appear as 'bahagya' in a text. I think I am going to add a field in my SRS app that indicates the word stress but not waste time producing it.

I had a blast doing this and am looking forward to refining my strategy for month 2. Personally I actually like doing SRS and find it relaxing (maybe there's something wrong with me).

[tomhal]

Thank you all for such a cool ride! It was quite a struggle for me. However I have made it till the end.
My results are 541 out of 900. TL ( German ) to my native Czech.