37§ 



POPULAR SCIENCE MONTHLY 



SHORTER ARTICLES. 



A VOCABULARY TEST 



Professor Kirkpatrick's article in 

 a recent number of the Popular Sci- 

 ence Monthly leads me to present the 

 results of an investigation on practi- 

 cally the same lines, extending over sev- 

 eral years when I was engaged in teach- 

 ing college students to read German. 

 I used a dictionary test, a little dif- 

 ferent in detail, but practically the 

 same as Professor Kirkpatrick's, to 

 find the number of German words 

 which could be defined by students 

 when they entered the second year's 

 work in the subject in college. Some 

 of them had had one year's college in- 

 struction, and others were admitted on 

 examination. 



I found that the vocabulary of those 

 who could pass such an examination 

 was never less than 2,00D words, and 

 went from that up to 5,000. The mark 

 received on the examination was in 

 close relation to the extent of the vo- 

 cabulary. Those who had more than 

 5,000 words were generally fit to go 

 into a higher course. 



The test was repeated at the end of 

 the year. The result then was from 

 5,000 to 12,000 words. The marks on 

 the final examination of the second 

 year's course were also in close rela- 

 tion to the extent of the vocabulary. 

 I tried this with classes for several 

 years, getting sufficiently uniform re- 

 sults to prove conclusively to my mind 

 that these were, the normal figures. 



I was then interested to extend the 

 investigation to English, and had sev- 

 eral classes make the same experiment 

 for their own language, but with the 

 very important feature that I used an 

 unabridged dictionary, containing over 

 100,000 words, instead of one contain- 

 ing only 28,000. I found that most of 

 the college sophomores reported from 



50,000 to 60,000 words. Of course, if 

 they had had only 28,000 to select 

 from, it would not be surprising if they 

 had reported only 20,000; and I think 

 that Professor Kirkpatrick made a mis- 

 take in using so small a book. J found 

 that students who had not studied 

 Greek regularly reported from 10,000 

 to 15,000 words less than those who had. 

 I also experimented with a number 

 of people who had never been to col- 

 lege, but, with an ordinary common 

 school education, were regular readers 

 of books and periodicals. These re- 

 ported generally from 25,000 to 35,000 

 words, though some of them went high- 

 er, even as high as the lower figures of 

 the college students. 



I then took a few cases of the work- 

 ing vocabulary in foreign languages of 

 those really proficient in them, chiefly 

 among modern language teachers. The 

 results are probably fairly typified by 

 my own case, which could, no doubt, 

 be matched by almost any one who has 

 made a life study of different lan- 

 guages. I found that my English vo- 

 cabulary was about 65,000 words ; Ger- 

 man (counting all compounds given in 

 the dictionary), 58,000; Danish (large- 

 ly the same roots as German ) , 52,000 ; 

 French, 30,000; Italian, 22,000; Latin, 

 18,000; Spanish, 16,000; Greek, 13,000, 

 and Old Norse, 11,000. 



I should guess that these figures, 

 which are for languages belonging to 

 only two general families, could be re- 

 duced to 20,000 or 30,000 actual roots, 

 or perhaps even less ; but to verify such 

 a guess would require an investigation 

 with a system of slips, for which I 

 probably shall never have time. I 

 leave the interpretation of these facts 

 to the reader, who can be assured that 

 they are facts. 



E. H. Babbitt. . 



