cod2 Groupie United Kingdom Joined 4356 days ago 48 posts - 69 votes
| Message 1 of 9 30 March 2015 at 10:37pm | IP Logged |
Question to 10,000 sentence miners who have been mining sentences seriously for at least 3 years:
I wanted to know where you get your sentences from. I extensive-read L2 books and newspapers, and I don't want to stop and mine sentences because that ruins the flow completely. I watch/listen to a lot of L2 material, but once again, my focus is on the listening/watching and not on mining. Also most of the A/V materials don't have subtitles because if I only watch subtitled videos then my choices get severely limited.
Sure, I'd usually note down words that really catch my attention, but I generally do no more when reading/listening/watching.
So that leaves me mining from only dictionaries. Which I really don't like doing any more. I feel I should be mining more from native L2 materials.
What do you do?
Edited by cod2 on 30 March 2015 at 10:37pm
1 person has voted this message useful
|
Michel1020 Tetraglot Senior Member Belgium Joined 4819 days ago 365 posts - 559 votes Speaks: French*, English, Spanish, Dutch
| Message 2 of 9 30 March 2015 at 11:27pm | IP Logged |
Are you trying to confirm your L2 skill by the number of sentences you mined ? because reading, listening and watching a lot of l2, you probably came accross more than 10K L2 sentences and in a more efficient way than any mining would have.
Go on reading, listening and watching - don't hesitate to read, listen and watch the same material many times with random intervals. Test yourself on new material.
I like AJATT a lot but he also write "Advice On How To Take Advice (Including Mine)".
So free your mind from the 10k sentences mining - it seems to me you are doing better already.
3 persons have voted this message useful
|
chaotic_thought Diglot Senior Member United States Joined 3344 days ago 129 posts - 274 votes Speaks: English*, German Studies: Dutch, French
| Message 3 of 9 31 March 2015 at 12:03am | IP Logged |
Do listen/reading on a novel and then after you're done, review a random sampling of head-sentences of paragraphs. These seem more likely to me to produce memorable gems from the work. Then cut out sentences that are too long or too short, and maybe just pick a few that you find interesting. For example here are a few I picked using this method from "20.000 Mijlen onder Zee". Each sentence stands pretty well on its own, and if you are familiar with the story they recall the scene and related events from the work.
1. In handelszaken is dit mogelijk Ned, maar niet in de wiskunde.
2. “Een ander die beter bij de hand is!” schreeuwde de kapitein, “500 dollars voor hem die het verwenschte beest raakt!”
3. Men berekent dat het water op den aardbol eene oppervlakte beslaat van 3.751.322.76 vierkante myriameter of meer dan 37½ millioen hectaren.
4. De zee was op een kilometer afstands rondom de Nautilus duidelijk zichtbaar.
5. “Nu heeft vriend Land gelijk.” zeide Koenraad, “en ik ben het met hem eens.”
...
For example, if you wanted to drill vocabulary, you could produce such a list of 100 sentences from each book you read and then use the "cloze" option in Anki and remove the difficult words from each sentence. Having already read the story, you should be able to fill them in with just this prompt. Note that if you mine 100 sentences from each book (100 would be a good upper limit for me - more than that and I'd rather just read the book again in its entirety) then you'll have to read 100 books in your target language to make it to your 10000 sentence goal.
4 persons have voted this message useful
|
day1 Groupie Latvia Joined 3694 days ago 93 posts - 158 votes Speaks: English
| Message 4 of 9 31 March 2015 at 7:51am | IP Logged |
Tatoeba
Also, many dictionaries give sample sentences for words with somewhat difficult usage
2 persons have voted this message useful
|
chokofingrz Pentaglot Senior Member England Joined 4991 days ago 241 posts - 430 votes Speaks: English*, French, Spanish, German, Italian Studies: Russian, Japanese, Catalan, Luxembourgish
| Message 5 of 9 31 March 2015 at 7:37pm | IP Logged |
If you like browsing sentences on Tatoeba in your chosen language, but would prefer to browse them all offline in txt or Excel, someone wrote a simple shell script to extract all the relevant sentences from the 200MB database download they freely offer.
However, if the language is a big one I think it is better to extract a subset and work with 1000-2000 sentences maximum (larger files slow my computer down). Even for relatively unpopular languages (e.g. Slovakian, 1700 sentences) it can create a nice starting point to work from.
If anybody wants to do that but isn't technically minded, I could provide a link to all the files I just extracted. By the way, in this method the sentences don't come with any translations.
2 persons have voted this message useful
|
cod2 Groupie United Kingdom Joined 4356 days ago 48 posts - 69 votes
| Message 6 of 9 01 April 2015 at 4:43pm | IP Logged |
Thanks for all the answers. Always useful to get different perspectives.
chokofingrz wrote:
If you like browsing sentences on Tatoeba in your chosen language, but would prefer to browse them all offline in txt or Excel, someone wrote a simple shell script to extract all the relevant sentences from the 200MB database download they freely offer.
However, if the language is a big one I think it is better to extract a subset and work with 1000-2000 sentences maximum (larger files slow my computer down). Even for relatively unpopular languages (e.g. Slovakian, 1700 sentences) it can create a nice starting point to work from.
If anybody wants to do that but isn't technically minded, I could provide a link to all the files I just extracted. By the way, in this method the sentences don't come with any translations. |
|
|
That's a great suggestion chokofingrz. I didn't know Tatoeba offers a full download. I tried to use their random sentence feature but it gets tedious after a while, as it mostly brings back useless sentences.
1 person has voted this message useful
|
garyb Triglot Senior Member ScotlandRegistered users can see my Skype Name Joined 5009 days ago 1468 posts - 2413 votes Speaks: English*, Italian, French Studies: Spanish
| Message 7 of 9 02 April 2015 at 11:20am | IP Logged |
I've been putting sentences into Anki cards for a few years now, although I still don't think I'm quite at 10000! My main sources are online articles/news/blogs/etc. (I especially seek out stuff that has a lot of everyday/colloquial language) and Kindle books (using the highlighting feature). I also sometimes take sentences from films/videos, although that can be a bit more awkward as I generally have to pause, maybe rewind to be sure I've got it correctly, and type it in manually. I also like example sentences from grammar books and dictionaries, but they're not a major source.
2 persons have voted this message useful
|
patrickwilken Senior Member Germany radiant-flux.net Joined 4335 days ago 1546 posts - 3200 votes Studies: German
| Message 8 of 9 02 April 2015 at 1:01pm | IP Logged |
If you use a plugin like Firelang for Firefox, or Readlang for Chrome, you can automatically save and export sentences into SRS for words you look up.
It makes the generation of relevant sentences for Anki really painless.
5 persons have voted this message useful
|