Detailed analysis: How many words do we know?

[tommus]

A very comprehensive analysis was just published by researchers at Ghent University, Belgium on the number of lemmas known by native speakers of American English. They tested 221,268 people in 265,346 sessions.

http://journal.frontiersin.org/article/10.3389/fpsyg.2016.01116/full

Here is part of their abstract:

"Based on an analysis of the literature and a large scale crowdsourcing experiment, we estimate that an average 20-year-old native speaker of American English knows 42,000 lemmas and 4,200 non-transparent multiword expressions, derived from 11,100 word families. The numbers range from 27,000 lemmas for the lowest 5% to 52,000 for the highest 5%. Between the ages of 20 and 60, the average person learns 6,000 extra lemmas or about one new lemma every 2 days."

[qeadz]

I just learnt what a 'lemma' is. Now I can relax for 2 days knowing I've met my quota

[s_allard]

I really have to congratulate tommus for making this excellent article available to the ALLF community. This question of vocabulary size is a pet subject of mine and this article goes a long way to clarify many of the issues that inevitably crop up when we discuss this topic. Now we have a serious document that we can refer to and avoid much unnecessary and sometimes unpleasant argument.

I especially like that fact that the authors point out how difficult it is to define what a word is and that any discussion of vocabulary size must start with the definition of a word. I should also mention that when measuring vocabulary size these authors, like all authors, use a methodology based on estimating vocabulary size according to frequency intervals. No one has measured the actual vocabulary of people.

[aokoye]

S_allard you might want to take a look at papers pertaining to corpus linguistics and psycholinguistics. I know that the paper linked is by no means the first of it's kind (though maybe one of the first in a psychology journal...).

[s_allard]

S_allard you might want to take a look at papers pertaining to corpus linguistics and psycholinguistics. I know that the paper linked is by no means the first of it's kind (though maybe one of the first in a psychology journal...).

The advice is well taken. Although this is not the first paper on vocabulary size, the methodology used to ascertain vocabulary size of a large sample is probably state of the art. As for corpus linguistics in general, I'll have to admit that I'm more familiar with applications in sociolinguistics, having cut my teeth on labovian sociolinguistics as a graduate student.

[reineke]

Ghent University
Center for Reading Research
Word test

"How many English words do you know? With this test you get a valid estimate of your English vocabulary size within 4 minutes and you help scientific research."

http://vocabulary.ugent.be./

LEVELS TEST OF VOCABULARY - RECOGNITION
"This is a vocabulary test. It has 40 questions, ten at each of 14 thousand-levels. If you use this test carefully, it should give you a good idea of the number of English words you know. Alternatively, you can also just use the test as a practise activity."

http://www.lextutor.ca/tests/levels/recognition/1_14k/

[mcthulhu]

The Ghent University test was entertaining, thanks. It also reinforced the first question I had when I read this post - what does "know" mean? For that particular test, at least, it seems to mean "recognize as a real word." I would usually aim a bit higher (e.g., at least some idea of a word's general meaning), but I suppose too many distinctions would make a test impractical.

I think I've seen more words than I know.

[aaleks]

The Ghent University test was entertaining, thanks. It also reinforced the first question I had when I read this post - what does "know" mean? For that particular test, at least, it seems to mean "recognize as a real word." I would usually aim a bit higher (e.g., at least some idea of a word's general meaning), but I suppose too many distinctions would make a test impractical.

I think I've seen more words than I know.

I tried to answer 'yes' only if I knew the meaning of the word. But in case with English I'm a foreigner, so I have to do so.


There's another well known (at least among Russian) test http://testyourvocab.com/ But I have some doubts of its reliability.

[William Camden]

I just learnt what a 'lemma' is. Now I can relax for 2 days knowing I've met my quota

Lemmas are rodents which gather in large numbers and then jump off cliffs.