416 EEPOET — 1891. 



of the relative is highly commended as an idiom, and we actually 

 imitate the grammatical blunders of classical authors pro- 

 vided that they occur often enough to constitute an idiomatic 

 usage. 



If we want merely to teach clearness and precision of 

 thought we ought rather to go direct to the study of the laws 

 of thought and inference as taught in logic. I am afraid 

 schoolmasters are not without blame in this matter ; we have 

 been trained to teach Greek and Latin grammar in a certaia 

 way, and we find it easiest to go on in the same groove. Boys 

 and girls, my experience tells me (strange as the statement 

 may seem), like to learn by heart in preference to being made 

 to think ; and the learning of the lists of gender rules and the 

 syntax is not altogether distasteful to the majority. Un- 

 doubtedly there is no lesson so easy for an ordinary school- 

 master as the old-fashioned grammar and construing lesson — 

 none that requires less originality and less preparation ; and 

 therefore our prejudices are all in favour of the retention of 

 this method of teaching. Personally, I believe in the study of 

 the thought rather than of the rules for the expression of the 

 thought; and, though I know it is a question how far young 

 minds can grasp the full meaning of beautiful thoughts in 

 language, still we ought to remember that the ancients them- 

 selves knew little of grammar rules ; they read their Homer 

 for the sake of the story and the poetry ; they learnt accuracy 

 of syntax from constant usage rather than from conscious ap- 

 plication of rule : and I believe that we are too much inclined 

 to spend our time in teaching how to express thought before 

 we furnish our young minds with thoughts to express ; we pay 

 too much attention to grammar and analysis, instead of realis- 

 ing that the best means of attaining accuracy of expression in 

 our own language is constant practice in the use of it. 



Again, in the matter of vocabulary, Latin and Greek are 

 the key to the scientific terms of all Aryan languages, not of 

 English only ; and a knowledge of classics no doubt helps to 

 save us from slight inaccuracies in the employment of words, 

 as well as from their gross misuse after the manner of Mrs. 

 Malaprop. Still, though we may know something more about 

 a word if we can trace its derivation, we do not therefore under- 

 stand its present meaning better. We assign, for instance, 

 just as definite a meaning to the word "sleep " as we do to 

 "somnolence," even though the derivation of "sleep" may 

 be unknown to us, while we can trace "somnolence" back 

 through Latin, Greek, and possibly Sanskrit. It is undeniable 

 that too close an adherence to the etymology will often lead 

 us astray in the pronunciation of words like " theatre," 

 "orator," "senator," "relative," "ablative;" and in the 

 meaning of others like "security," " indifferent," "censure," 



