|
 |
Word freqs vs Word learning: Paradox? | ||
Tags: Vocabulary learning | ||
Share with:
| ||
 Language Learning Forum : General discussion |  | |
| Jeffers Senior Member United Kingdom Joined 4912 days ago 2151 posts - 3960 votes  Speaks: English* Studies: Hindi, Ancient Greek, French, Sanskrit, German | Message 9 of 19 19 November 2014 at 8:29am | IP Logged | |||||
I found that article quite interesting, although it is based on research that I am a bit familiar with from other similar articles. One correction to your post: the learners did not learn 2500 words in 6 months. They got to a rate of learning of 2500 words per year over 6 months, so they learned 1250 words in 6 months. The takeaway point of the article was that the learner should learn the 3000 most frequent words as an "immediate high priority", and then develop "strategies to comprehend and learn the low frequency words of the language". One very encouraging (and new) point for me was the table of vocabulary size and coverage in novels aimed at teenagers. In that type of literature, 2000 words gives 90% text coverage, and 2600 words gives 96% text coverage. It may be an obvious point that teen literature is easier than adult literature, but sometimes it's important to face the obvious. If you want to learn vocabulary from reading, you need high text coverage to be able to guess from context. If you start with teen lit (or easier), you can start reading and learning from native material sooner. It also seems to me that reading within a narrow topic of interest would be similarly beneficial. 2 persons have voted this message useful | ||||||
| patrickwilken Senior Member Germany radiant-flux.net Joined 4536 days ago 1546 posts - 3200 votes Studies: German | Message 10 of 19 19 November 2014 at 8:49am | IP Logged | |||||
|
Some really interesting ideas in this discussion, but I think I didn't make my point clear enough.
I find it fascinating that children learn year in, year out, about a 1000 word groups per year. As Iversen points out some of us resort to SRS or word lists to try to capture the lower frequency words. Children obviously don't do this. They just keep learning and don't seem worried about the fact that a certain word is only seen 1/100000 times or 1/1000. That's weird to me, and I would like to know how they do it.
But the papers by Nation (I can dig out a reference if you want) show that 8000-9000 words is enough for 98% coverage of adult text, and 2000-3000 words for 95% coverage of ADULT text. So it doesn't sound like these figures totally agree, as young adult literature should presumably have a somewhat easier vocabulary. EDIT: Sorry, I must have misread the Nation paper. The reference I was thinking of was: How Large a Vocabulary Is Needed For Reading and Listening? I.S.P. Nation (2006). According to this paper you need about 4000-5000 words for 95% coverage of adult novels. So young adult novels do offer a significant advantage. Edited by patrickwilken on 19 November 2014 at 10:11am 1 person has voted this message useful | ||||||
 Iversen Super Polyglot Moderator Denmark berejst.dk Joined 6706 days ago 9078 posts - 16473 votes Speaks: Danish*, French, English, German, Italian, Spanish, Portuguese, Dutch, Swedish, Esperanto, Romanian, Catalan Studies: Afrikaans, Greek, Norwegian, Russian, Serbian, Icelandic, Latin, Irish, Lowland Scots, Indonesian, Polish, Croatian Personal Language Map | Message 11 of 19 19 November 2014 at 10:20am | IP Logged | |||||
This is an important observation. Basically it means that you should treat fixed expressions and grammatical items like you treat single words: some are very common and should be learnt as early as possible, but luckily they are also so common that you will see them often enough to see a pattern. The rest are so rare that you should take care to notice them and do something to remember them while you have the chance -like writing them down and/or deliberately forming associations to nail them.
Parents who actually teach children may know more about this, but my suspicion is that the first 1000 words are used so often in a normal family that even a child can't avoid picking them up. And once you have learnt them you can't put them on your CV a second time. After that there is a period where slightly rarer words like doll, Ipad, icecream and pooh are overrepresented in the communications with the child. They may be rare in the statistics, but not rare in a family with children. The task changes when those first couple of thousand words are learnt because ALL remaining words in the language belong to the rare category. But hopefully the kids are picked up by the school system before their linguistic development peters out. Or they have environments where they are expected to hear and react to rare words/word combinations like boeuf Bourguignon, zebra and goalkeeper. Beside kids are supposed to have retained their curiosity. Curiosity may kill cats, but it helps humans to pick up words. Edited by Iversen on 19 November 2014 at 11:18am 2 persons have voted this message useful | ||||||
| s_allard Triglot Senior Member Canada Joined 5433 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 12 of 19 19 November 2014 at 1:04pm | IP Logged | |||||
|
In their well-known paper, Paul Nation and Robert Waring write:
"At present the best conservative rule of thumb that we have is that up to a vocabulary size of around 20,000 word families, we should expect that native speakers will add roughly 1000 word families a year to their vocabulary size. That means that a five year old beginning school will have a vocabulary of around 4000 to 5000 word families. " Vocabulary Size, Text Coverage and Word Lists I find it hard to believe that a five-year old who has not yet entered school can have a vocabulary of 4000 to 5000 word families. Considering that they have not yet started to read or write, where does the child get so many words? From listening to the people around them and watching a lot of television? I think there are clearly methodological issues of how one measures knowing a word in children. 1 person has voted this message useful | ||||||
| luke Diglot Senior Member United States Joined 7208 days ago 3133 posts - 4351 votes Speaks: English*, Spanish Studies: Esperanto, French | Message 13 of 19 19 November 2014 at 1:25pm | IP Logged | |||||
In addtion to those activities, the parents I know read to their children. Early books are aided by pictures. One parent was telling me this week he is reading C.S. Lewis' The Chronicles of Narnia to his girls, who are ages five and 2 1/2. The younger daughter of course understands less than her sister, but she is getting exposure to a real story dramatically delivered by her father. Dad recognized that even for the five year old The Chronicles of Narnia is a stretch, but he is pleased that his daughter enjoys the story and wants to find out what happens next. Another parent mentioned his son was graduating from university. He told me he read The Odyssey and The Iliad to his son when he was a boy. The father found it interesting that his young son was so interested in the story, particularly with it's graphic descriptions of combat. As far as how can one figure out if a 5 year old knows words like "dragon", "grass", "bush", "skyscraper", "helicopter" or "grizzly bear", one uses a similar approach to the one Iversen poked some holes in. I.E., for children, you show pictures and have them select the correct word, which could be spoken. E.G. Is this a bear, a zebra, a cat, or a fox? Is this a skyscraper, a house, a tent, or an igloo? 2 persons have voted this message useful | ||||||
| patrickwilken Senior Member Germany radiant-flux.net Joined 4536 days ago 1546 posts - 3200 votes Studies: German | Message 14 of 19 19 November 2014 at 1:31pm | IP Logged | |||||
That's crazy isn't it? Assuming a linear growth in vocabulary, that suggests babies surpass your 350 word core within the first four months of life. :) Though more likely within 18 months... Edited by patrickwilken on 19 November 2014 at 1:32pm 1 person has voted this message useful | ||||||
| s_allard Triglot Senior Member Canada Joined 5433 days ago 2704 posts - 5425 votes Speaks: French*, English, Spanish Studies: Polish | Message 15 of 19 19 November 2014 at 5:24pm | IP Logged | |||||
|
Now that somebody else has raised the wonderful subject of a small active core vocabulary, I just want
to mention that when talking here about "knowing" X number of words at a given age, we are talking about receptive or passive knowledge. In my opinion, they is usually tantamount to being able to recognize the word and have some element of meaning. Actually having used the words is a whole other issue. I happen to believe that most people actively use very small numbers of words. But I don't want to get into that debate. But there is a very important methodology issue here. All the studies of vocabulary size and text coverage are based on the idea of summing the vocabularies of individual texts or speakers and then counting the number of words to get the desired level of coverage. The actual vocabulary used by an specific text or speaker will be much smaller than that of the aggregate sample. For example, if your corpus of writing includes samples from the 19th century, it will contain plenty of vocabulary related to sailing ships, horse-drawn carriages, steam engines, etc. Similarly, modern contemporary writing will refer to mobile phones, tablets (of the computer kind), Internet, websites, etc. You need a huge vocabulary to cover both kinds of writing, but if you are interested only in one era, you can disregard most of the other vocabulary. So, in fact you may not need a 20,000-word vocabulary to read a selected range of novels. If you like reading Michael Chrichton, then maybe 5000 words will do fine. Add Harry Potter to the mix and maybe 6000 will be all that you need. Similarly, when it comes to speaking and writing, if you want to talk about a wide range of topics, a wide vocabulary is necessary. But if you limit your topics, a common core ranging from 350 to 1000 word families plus any specialized vocabulary may be all that you need. 1 person has voted this message useful | ||||||
| patrickwilken Senior Member Germany radiant-flux.net Joined 4536 days ago 1546 posts - 3200 votes Studies: German | Message 16 of 19 19 November 2014 at 7:21pm | IP Logged | |||||
|
I guess one possibility for why children learn at a fairly fixed rate could be that there is fairly strong cognitive bottleneck to learning new words. Perhaps it's fairly difficult for children to learn more than 3-4 words/day on average - even though they might be encountering many more.
I was pleased today to see that I've learnt something like 7000-8000 words in German over the last 2.5 years, mostly from reading, but that still only comes down to about 8-10 words/day. My experience with SRS is that some words are just very easy to learn, and others not. Perhaps the difficult ones are difficult because the semantic network that represents our L2 is more ready to accept certain words than others at any point in its development. Of course, learning the L2 is different, as we tend to use our very sophisticated L1 semantic network as a bridge to our L2 network. I can say what debt or banker or psychopath is in your L2 and you can learn this immediately because you already have these words in your L1. Children don't have this and so can only build from the ground up - they can't learn concepts (words) before they have learnt others. We can by mapping these directly onto our L1 - though it's presumably easier to learn a word if it fits naturally into our developing L2-semantic-network, rather than having to be forcibly grafted on from our L1. Edited by patrickwilken on 19 November 2014 at 7:23pm 1 person has voted this message useful |
|  | 
Printable version
|
| Copyright 2024 FX Micheloud - All rights reserved No part of this website may be copied by any means without my written authorization. |
