Beli Tsar's log - Ancient Greek, Farsi, Latin, Hebrew

[Beli Tsar]

It’s been a disappointing week, languages wise. Apparently being ill while working 12 hour days and arranging small children’s birthdays isn’t helpful for maximising study time.

Greek:
- Reading at about half my normal rate
- kept anki reviews up to date
- added new Anki vocab for Classical (rather than Koine) Greek
- A few pages of Athenaze.
- A few podcasts from dia nuktos dialogos.

Farsi:
- read 3500 words on LingQ + 1.5 hours of listening.
- hit about 150 reviews/day on Anki (max)
- kept memrise up to date, no new cards
- reviewed past tense (very simple!) in lesson
- watched a Mehran Modi comedy (https://youtu.be/g_Y8xHl6-2U) - good fun, and so rare to find accessible subtitled content other than arthouse films. This one is particularly good for all the many elaborate greetings Iranians use.
- A few hours mostly incomprehensible input, chatting, drinking tea and playing Backgammon. All good fun.

French:
I’m not supposed to be doing French. But I was tired, and watching TV in a foreign language was all I was up to most nights. And there is a real lack of good programming in Ancient Greek, for some reason, and decent Farsi TV (not films!) with subtitles are rare. So I had some fun in French.

Observations:
My thoughts have been revolving around SRS and Farsi this week. The more I go on, the more I think that I get more profit out of this than anything else at this (very early) stage. If I was starting over, I’d have worked harder to get a frequency list earlier, and have done a thousand words before anything else. But I was a bit fed up of SRS after having just done the same thing for Hebrew. Now it’s catchup.

I’ve also been comparing, informally, Anki and Memrise. I’ve recently been using Anki more - it’s compelling how convenient it is to adjust and add on the fly, and it is much better long-term in avoiding massive card build up. Key for me was the greater flexibility when it comes to testing grammar and sentences, for which Memrise isn’t much use.

But the one Memrise beats Anki solidly on is actually teaching you cards. The Anki Manual makes that clear - it’s designed to help you remember what you have learned, not to teach it. It shares that with Supermemo, of course.

But Memrise shows that it is possible to use SRS to learn lists of words you have never seen before at great speed. I routinely learn words in Memrise at 4* the speed Paul Nation’s estimates it takes with flashcards. And that’s with audio, which at least doubles the time it takes, and in a new script, Farsi. In Greek without audio it is much, much faster - and a completely familiar script even better.

And words I learn in Memrise are wired thoroughly into my head. People often talk about SRS words only being useful passively. I haven’t found that. Sure, I need to speak them a bit to get them really ready, but they are often ready to go as it is.

So, this week, I did some messing around with Anki - can adding more card views at short intervals replicate this? The answer, I think, is ‘not very well at all’. I’d love to know if anyone has done better on this.

In addition to Memrise’s excellent pattern of showing you each card 6 times, in 2 directions, and requiring typing once, I suspect that the much-maligned multiple choice is actually very helpful. There is an early stage of memorisation where actually producing a term is too difficult - and yet some kind of testing and recognition strengthens the memory better than just viewing it again. Memrise does that - so that you can absorb a large number of cards, slightly more slowly, but in the long run much more deeply.

Rather than abandoning Anki immediately, though, I am wondering about doing some more rigorous experiments - 100 words memorised with Anki, Memrise and some kind of control, and then testing recall. If I can get enough useful stats out of each program it should be quite helpful to compare their relative utility in each situation.

[Beli Tsar]

I've had a bit of a Christmas break, and as you can tell from the gap between this and my last post, it's been rather too long...

Greek:
This is going swimmingly. I'm really enjoying it - sometimes I'm really unsure what I am actually doing that's making me better, because I don't think I'm putting that many hours in. But it's certainly getting better.
- Basic reading - besides my regular reading, I finished the Italian Athenaze, which I'd been using as a reader for Attic, and started using the JACT series for the same purpose.
- SRS - I learned a few hundred words of Attic vocabulary from JACT, but got a bit burnt out from SRSing and let it slip otherwise.
- Listening - Both Alexander Veroniensis' Barbarismos and Seumas MacDonald's dia nuktos dialogos have been brilliant. They use very different accents (erasmian and koine, for those who care) and initially I found the Koine especially very tough. But it's amazing how quickly listening has become easy.
- Grammar - I picked up the rest of the JACT books with the intention of reviewing grammar properly, since mine is really ropey.

Greek keeps hoovering up other time, because it's just so much fun. The literature is really gripping, and it's great being at a break-through stage in a language when you can suddenly access loads more texts.

I'm not too surprised that Koine is getting pretty easy, in general, so that I've been enjoying reading the Apostolic Fathers very much. But Attic is coming along nicely too, though it is famously harder. My parents-in-law, who love languages, gave me a Loeb edition of Plato for Christmas, which will come after JACT.

Best of all, I picked up an edition of the Odessey yesterday second-hand, and was delighted to find that I can make pretty good sense of it already - enough to want to keep turning the pages, because it's beautiful and exciting.

Listening seems to have a pretty big impact on my reading skills - whether it is the speed of processing necessary, or just extra input, I really notice a jump in my reading skills on days I listen.

My grammar skills are appallingly weak, and I notice that people with fewer reading hours can pick up details I can't because of it - even if I can read better than they can overall. So I'm going to do a review of grammar, using the JACT textbooks, finally getting those optatives and -mi verbs sorted out. After that I'll move on to do some composition to get the grammar really cemented, with Eleanor Dickey's excellent textbook. At least that's the plan...

I am keen to use SRS to help me internalise the grammar well, and that's been a big part of my SRS-burnout. It's just too easy to make too many really hard cards - declensions and so on. But, on the other hand, adding sentences with review of lots of difficult forms seems to be both effective and fun. It's time-consuming, but probably what's necessary.

Farsi:
This is where Christmas hurt. I kept up on SRS reviews, but little else. I meant to have a couple of weeks off, but it stretched too far, and work's been so busy since Christmas... You can tell that I am not motivated enough to learn the language really well.

But seeing some more Iranians this week, and doing some basic reading bits and pieces for work, was an encouragement. My speaking is really bad, but the reading skills are doing just fine. I think the trouble is being in the no-mans land between real beginner and being able to consume proper native content. Beginner content is dull... Perhaps time to try some listening-reading?

I've also got going on the Teach Yourself textbook, because I've been so unsystematic so far. But it's quite odd - should the second chapter really be all about numbers and counting? Is that really the thing to master at this stage? And it is so short, and there's so little serious content so far...

Plans
I want to commit to reaching basic reading skills in Hebrew this year, so there's a limit to how much I can commit to Farsi. I reckon I'll do six months of focus on Farsi this year and six months on Hebrew, meanwhile keeping the other ticking with a little reading/listening/SRS reviews. I'm not sure how to divide the time, though - big blocks or month by month...

[Beli Tsar]

Seeing Iranian friends again has restored my enthusiasm for Farsi. Spending too much time on this site has kicked that into mad determined obsession.

Aims for the year ahead
Dissatisfaction with my production and listening has led to scouring the forums this week on efficient, fast ways to learn - even more so than usual. In addition to absorbing again the wisdom of Iversen, Reineke and Iguanamon, among many others, I have been looking for evidence of Smallwhite’s methods. It seems to be quite hard to find much detail.... But the results seem very impressive. It all seems logical, focused, effective, and mildly scary. It makes me think I’ve been a bit lazy and passive in my approach to learning.

So I’m doubling down on vocab and grammar for now, and aiming to hit B1 this year in Persian. That’s optimistic, but better that than lazing around.

While I could sit back and enjoy the process more, I’m sure I’ll get more enjoyment still out of it once I can read and speak properly.

B1 in Farsi
I really want to hit this level in Farsi this year. To make this worthwhile, I need to be able to communicate usefully.

I intend to spend 6 months focusing intensively on Farsi, probably split up into two blocks of 3 months. The first started on the 1st of February. I’ll reexamine goals after the first block.

In that first block, I intend to:
- SRS 1250 words of vocab by frequency. I’ll pick up plenty other bits and pieces reading and listening, but this will be the core. I already know a good number of these, and a lot of others, but need to make sure this core is solid
- Finish TY Complete Persian
- Start Routledge’s Basic Farsi Grammar
- Hit my original 100,000 words of Farsi read target
- Listen 30 minutes a day hours (45 hours)

I’m not tracking everything - e.g. sentences added to SRS, grammar added to SRS, speaking etc.

I ought to add daily writing, but I’m struggling to see where to fit a daily writing habit in my day. Perhaps once the TY book is done I can add it in in some evenings. Certainly once I’ve got the core grammar down I want to work out a way of doing good, rigorous drills.

Hebrew
After the first 3-month block of Farsi, I’ll put that on the back burner. I’ll keep up SRS reviews, a bit of reading, listening, and speaking. But real energy will go to Biblical Hebrew, up until the summer. My goal here is more modest: basic slow reading capacity. My goals will be:
- Vocab to 1500+ (from 600)
- Finish Kittel & Hoffer, Biblical Hebrew
- Get reading (and listening)
- Extra goals can be added!

[Beli Tsar]

[Reserved for later!]

[Beli Tsar]

Week 1 Review
- SRS - up to 129. got off to a rather bad start, as I didn’t have the frequency list set up. But I substantially ramped up other SRS in the background.
- TY Complete Persian - 7 Chapters (of 21); a great start.
- Words read - 5200 - on target. I expect this will get faster with time.
- Time listened - 4 hrs 50 min. Well ahead of target

Currently I’m being fuelled by mad obsession. It’s not at all sustainable, so we’ll see how I do once that wears off in a week or two.

I’m utterly amazed how easy it is to work through the first sections of TY Persian with a bit of reading and basic grammar knowledge under your belt. My previous experience was with much more hardcore ancient Greek textbooks. And I came at that with no knowledge of languages, no understanding of English grammar, no vocab, and no SRS. This, comparatively, is a walk in the park.

I’ve been drilling conjugations with enthusiasm.

The book’s also a refreshing change from entirely travel-centric language books, which, if you aren’t travelling, can feel utterly irrelevant. Instead, there are subtle hints of well-off 1990’s expat Iranians in London - characters in dialogues are all photographers and graphic artists bumping into each other at the London film festival, or spending time at their second homes on the Caspian Sea.

That makes a big contrast to the frequency list, which, though it apparently includes conversations, seems skewed towards newspapers. The result is a picture of official Iran: ‘America’ comes at number 107, well before, for instance, ‘no’. The example sentences are even clearer - ‘The main goal of the enemies of the system is stopping the revolution’

SRS will be the big challenge. Lack of vocabulary has been holding me back, so it’s important. I know I can learn at this rate, and more, but whether I can keep it up for three months is a different matter. The data-entry side is even worse

[Beli Tsar]

Week 2 Review
- SRS - Target to 275 - actual 368
- TY Complete Persian - 4 chapters, to 11 of 21
- Words read - Target 5000 - actual 5200
- Time listened - Target 3.5 hrs - actual 5.5 hrs

I still feel like I’m flailing around somewhat, hungry for effective methods, but not quite satisfied, especially with listening. And yet, big improvements this week.

Right at the beginning of the week, I saw a big jump in listening comprehension, mostly from improved vocab. I am reminded of an ancient quote of Cavesa’s on htlal -something like ‘Many listening problems are vocabulary problems in disguise.’

That’s true of real, live Persians too. And Persian friends are also noticing a sudden jump in confidence and capacity to speak.

SRS targets are more than manageable, but I’m finding unfamiliar words in a script that’s still not natural to me much harder to absorb than I expected. It shows how narrow my reading has been, in part. But it's also the script.

I’d love an extra half-hour a day, or thereabouts, and real, focused time, so I could work through FSI or something similar. Nearly everything I do is in little snatches of time...

[Beli Tsar]

Week 3/12 Review
: 3 / 12 Weeks

: 411 / 1275 Words (+71 words)
: 12 / 21 TY Modern Persian (+2 Chapters)
: 55424 / 100000 Words read (+2280)
: 12 / 42 Time listened (+4.2 hours)

Ill this week - not ill enough to stop working, but ill enough to wreck language study. I’m glad I was ahead on words, but this week only listening was itself ok.

This new week has started very poorly, too.

Fortunately, at just the point where Memrise reviews were about to become unbearable, words suddenly gelled and it’s much easier - after a day off at the weekend, as is often the case.

I was finding the range of new words, in tough script, with pronunciations that are hard to recall, just too much - progress was quite painful. But two weeks to let the first lot really sink into my mind, and it’s all coming so much more naturally.

TY Complete Persian really is poorly structured. The idea of a verb isn’t introduced till half way through the book in multiple chapters, and then you get two new tenses, the participle needed to form them, and some extra material, all in one chapter. It’s well explained, so not hard to understand - but there’s no chance to internalise it at all. The chapter explains the pluperfect, but doesn’t even require it for the exercises - though it’s a common tense in Persian.

Flicking through the old TY Persian by Mace, it’s so much better. It’s unashamedly a grammar textbook, but if you like that sort of thing, it’s well done. If it wasn’t for the lack of audio it would be a great place to start. So I have to decide between that and Routledge for my next book.

[Beli Tsar]

Week 4/12 Review
: 4 / 12 Weeks

: 547 / 1275 Words (+137 words)
: 14 / 21 TY Modern Persian (+1 Chapters)
: 57814 / 100000 Words read (+4431)
: 16 / 42 Time listened (+4.4 hours)

It’s been a bad week for free time, so only one chapter of TY done. But I’ve done a bit on the next two, and ought to be able to wrap them up tonight. Reading, likewise, is somewhat behind - but I don't care much, as I am getting faster, and should be able to make up lost time there more easily than elsewhere.

Finally getting on top of the vocab. As I’ve said before, I’ve been finding it much harder than expected to absorb new words. Partly the language is still new. But partly the crossover of my frequency list with stuff I’ve learned or read before is far less than I expected - about one in ten, which when I thought I’d learned a thousand odd relatively high frequency words is disappointing. The frequency list really is drawn primarily from newspapers, and it shows - and I haven’t read any.

I experimented with both memrise and anki, with using pictures, with using sound, and with mems. In the end, mems blow all the other options out of the water - just far more effective, and quicker. I’d been resisting - it’s a lot to create - but I’m managing to create something for 90% of words within 30 seconds.

The results are good, and the strain of the SRS is beginning to diminish as a result.

Attacking the vocab problem from the other side, I’ve been reading some newspaper articles from LingQ, choosing topics like ‘Mullahs who protested against the revolution’ as that seems like it will cover the frequency list words best…

Planning
I’m one-third through this three-month burst, and am still planning to tackle Hebrew when I’m done. It’s too early to tell if I should be pleased with results so far, but I am beginning to think about how I keep up Persian while I do Hebrew. I’m already feeling my Greek begin to suffer as it is.

My plan to keep up Persian is:
Keep on top of SRS reviews (no new cards)
Listen daily - 10-15 minutes minimum
Read something in LingQ every day.

Over the next two months I want to get to a stage where this is really viable. I’m running out of beginner content, and I don’t want to be reviewing that forever, so I need to be able to use native content.

I have two obvious options. One is to both read and listen to the New Testament - which is familiar, easier, and interesting to me. The other is to tackle News. That feels like it would be useful for getting to know Persian culture also - especially with such a politically-driven nation.

News would be harder. But the vocab frequency list I am using is tailored for it. It would also be helpful long term - I want to read the Bible in Greek and Hebrew, not Farsi, in the long term! Whereas I’d be very happy to switch all my news consumption to Farsi - listen to NHK world news, and read BBC Persian, Radio Farda, Al-Jazeera, etc. etc. It’s a good slot for language maintenance long-term.

It would be nice to do both. But I’d like news to be possible, so I’m switching my reading to that now, with the hope that in a couple of months NHK is roughly comprehensible.

Difficulties
I’ve had a number of little frustrations lately that set limits on what I can do.

The first is that it seems to be impossible for me to buy Persian ebooks or audiobooks - sites won’t accept my payment, presumably because of sanctions.

The second is that though I can get hold of a few ebooks for free, I cannot convert them into a useful format for LingQ/Readlang/LWT, which at my stage I need to use them properly. Even when I have books as pdfs, copying and pasting page by page still doesn’t work, as I can’t get the words to divide properly. Computers just don’t like Farsi script very much… Though perhaps unusual DRM is playing a role too...

So I am limited to web content at the moment, just when I’d love to get into a longer book. And easy content (Harry Potter!) would be so helpful.

So focus on learning to read news makes sense for now, unless anyone else has a solution to these issues!

[MorkTheFiddle]

The fact that you are adhering to a very ambitious language learning plan is very impressive.

[Beli Tsar]

The fact that you are adhering to a very ambitious language learning plan is very impressive.

Thanks for the encouragement! We'll see whether it is a wise plan in the next few weeks...

And certainly this week, for various reasons, has been a bit more disappointing.