JULY — SEPT. 185T.] The Study of Licing Languages. 235 
reject of course all words wliicli will never be required, such as 
those used only in learned works : next, reject all those that hard- 
ly ever occur even in books. Next let alone those that are chiefly 
in use only in certain particular lines of life, and are more or less 
technical. Again, have nothing to do with any words that are not 
commonly used in the ordinary matters of life. What can be gain- 
ed but clear loss by burthening the young beginner with a multi- 
tude of words, by far the greater part of those in the language, that 
will be thus rejected, when they have nothing whatever to do with 
his acquiring a useful knowledge of the language, and when, if re- 
quired, they can* afterwards be added in a tenth part of the time 
that they would reqfiire at first. Probably out of twenty thousand 
words in a language, the knowledge of 5,000 would set him so per- 
fectly at liberty in all ordinary conversation that neither he him- 
self nor those he converses with would be reminded that he does 
not know all. And if occasionally a person used one of the re- 
maining words, probably he could not mistake the meaning of it in 
the midst of so many known ones. And if he could not perceive 
what it must mean, he could have no difficulty in asking the mean- 
ing or understanding the explanation. Having thus relieved our 
student from such a mass of useless labor, let us next divide these 
5,000 words, or whatever the number is, into several portions, tak- 
ing out first one thousand and then another of the least common 
and least immediately necessary words, till we-have only a thousand 
left. Out of these we again take 250 three times and then 150 in 
the same way, so that to begin we have only 100 of the commonest 
words in the language, but consisting of all the different parts of 
speech so that little sentences may be formed out of them. The 
learner then deals with only one of these batches of words at a time, 
not troubling himself with the others till he knows the first batch 
as well as he does so many words of his own language. This is 
one of the great essentials of the system proposed. Words sh®uld 
never be partially learnt and forgotten again, nor imperfectly, that 
is, so that their true value and use are not thoroughly known. 
When once a word is taken up, it should of course be thoroughly 
secured both as respects the meaning of it and its use, and it must 
t 
