PROGRESS UPDATE, WEEK 17

Summary

RUSSIAN

Active study: 5.2 hours (study-and-click: Эхо Москвы, “Культурный шок” [podcasts]; Евгений Замятин, “Мы” [novel])
Total active study this year: 85.2 hours
Current estimated reading level: 80% (B1)

Other activities this week:
– 1.5 hours of Russian music in the background
Total for other activities this year: 35.6 hours

Wanderlust confessional

My main vice this month – too much programming and not enough clicking. Although I must say in my defence, programming still involves languages. 😉

Teango’s word of the week

“заусенец” (hangnail) – those occasional pesky distractions on the end of your fingers that try and tempt you to start nibbling.

Notes

My main goal at this stage is to boost my vocabulary to the point where I can comfortably read books like “Ночной дозор” and ideally pick up any average Russian novel and start reading without too much difficulty (i.e. understanding at least 94%). Sometimes this seems like an impossible ambition without dedicating years and years of blood, sweat and tears. However I know it can be done.

The reason I particularly mention it now is because I recently surpassed 6,000 words and tested my reading level using an unseen selection of 300 words taken from Дневной Дозор. I was initially worried that I’d left my Russian studies for too long, having probably forgotten a huge chunk of vocabulary, and considered skipping the test altogether. However curiosity got the better of me, and I discovered that my fears were indeed justified. I dropped from my last score of 87% to a miserable 80%! At this point I seriously thought about giving up on Russian altogether (as much as I love the language) and moving over to another easier language where I know I could do all this in a mere fraction of the time and with far fewer headaches.

However, I stayed on track. I decided I must be a complete dimwit not to have reached at least 90% yet, but analysed the words I didn’t know in the test section anyway. And as it turns out, most of the words I didn’t know also didn’t occur in my study novel “Мы”! Hooray – not a complete nincompoop after all… 🙂 In fact, when I compared both texts with some software I wrote especially for the task, and incorporating a morphological analyser to determine lexemes, I got the following analysis:

x = unique word families in "Мы" listed in frequency order; y = % of words this covers in "Ночной Дозор"

“The study text contains 5,887 unique word families. This covers 75,731 (77.803%) of the 97,337 words in the test novel.”

This means that if I were able to learn every single word in “Мы”, and could conjugate all the verbs and decline all the nouns, pronouns and adjectives, I’d still understand less than 78% of the test section. This made me feel even more relieved at first, and it made sense as the two texts are almost a century apart and completely different genres. Yet it did make me wonder i) why could I reach 95% in Swedish after 55 hours and 4,000 clicks, but only 80% in Russian after 80 hours with 6,000 clicks, ii) how was I able to achieve more than 87% back in February and even 80% now after a long hiatus, and iii) how on earth would I ever be able to hit my target of 94% in a passage from this test novel?

So I took a quick look at the books I’d used for learning to read Spanish and Swedish last year, and found that the words had a much higher overall lexical similarity to English and other languages I know. That’s not to mention that they were in the Roman alphabet, the word order was relatively predictive and could be more easily aligned, and the passages bore easier grammar in comparison to the morphological complexities of Russian.

I then returned to the test section in “Дневной Дозор” and discovered that there were a handful of words I knew that weren’t in my study novel – i.e. names and loanwords (or at least words very close to the English translation). These accounted for 13% of the passage, with 6% covered by names, and the remaining 7% covered by similar words (minus 3 “false friends” in this case). The picture started to become clearer.

I quickly turned myself to the task of cross-indexing various Russian texts with some more home-baked apps and continued to examine their word families in order of frequency within the text. It soon became plain that texts could vary to a much greater extent in their vocabulary than previously anticipated, largely depending on features such as size, genre, age and general reading level (e.g. the range was 72-91% across a dozen chosen texts). This suggested that it’s essential to cover as wide a variety of texts as possible in order to be able to pick up a new book and start reading comfortably.

So I’d answered the first question: Russian is a tough language for English speakers, and my study and test texts being so very different in terms of vocabulary didn’t help matters at all. I’d also also answered the second question: I was able to boost my scores by an extra 6-13% on average through my knowledge of other languages and picking out names, as well as probably a touch of my own morphological analysis or contextual disambiguation along the way. But the third question was still left unanswered?

In search of a solution, I turned to my three familiar musketeers of language learning once again for help, “Materials, Methods, and Motivation”, and focused long and hard on how I could reach that infernal goal of 94% in “Ночной Дозор”. Scrutinising all the steps in my approach, I discerned that two key areas for improvement would be a) to improve upon overall depth and recall of vocabulary (and try not to abscond from study for so long in future 😉 ), and b) gain a broader vocabulary base (or just get an easier test novel! lol).

So here’s what I plan to do… To improve on review, I’ll begin each new session by reading and listening to the previous session to bring me up to speed after the break. To carry this through, I’ll need to review the previous day’s work at the start of each new day, as well as listen and read the previous week’s material over the weekend too. With any luck, I’ll retain more words in the long run, deepen what I can recall, and get to practice some extensive listening and reading without pauses earlier on.

To gain a broader set of vocabulary, I’d like to study each novel up to the point where I can successively understand 94% of a page (either new or not recently studied or reviewed), after which I’ll listen and read without pauses to the end of the book instead. Then it’s on to the next novel! The idea behind this is that I will quickly race over the exponentially long stretch of less frequently occurring words in that last remaining 6% (see graph above), which are likely to be more specific to the study novel and less likely to appear in the test novel anyway, and I’ll cover a broader base of vocabulary in less time.

I’m also going to try and simplify my approach further by ignoring the third step of study-and-click (that’s where I usually listen to the audio and click new words I recognise), as well as reducing my (paper) logbook notes to one line entries for recording the material used, hours spent studying, and the number of clicks achieved at the end of each day. I hope to catch up with listening skills and grammar a little later on in my language learning programme (and I have some interesting ideas about how to tackle this which I look forward to sharing), but for the time being, my focus is firmly on acquiring as much vocabulary within context as I can, and increasing my determination to reach this goal!

Advertisements