sential to
the study, are palpably erroneous. I attribute this to the carelessness
with which men have compiled or made up books of grammar; and that
carelessness to those various circumstances, already described, which have
left diligence in a grammarian no hope of praise or reward. Without
alluding here to my own books, no one being obliged to accuse himself, I
doubt whether we have any school grammar that is much less objectionable in
this respect, than Murray's; and yet I am greatly mistaken, if nine tenths
of all the definitions in Murray's system are not faulty. "It was this sort
of definitions, which made _Scaliger_ say, _'Nihil infelicius definitore
grammatico_.'"--See _Johnson's Gram. Com._, p. 351; also _Paragraph_ 5th,
above.

17. Nor can this objection be neutralized by saying, it is a mere matter of
opinion--a mere prejudice originating in rivalry. For, though we have ample
choice of terms, and may frequently assign to particular words a meaning
and an explanation which are in some degree arbitrary; yet whenever we
attempt to define things under the name which custom has positively fixed
upon them, we are no longer left to arbitrary explications; but are bound
to think and to say that only which shall commend itself to the
understanding of others, as being altogether true to nature. When a word is
well understood to denote a particular object or class of objects, the
definition of it ought to be in strict conformity to what is known of the
real being and properties of the thing or things contemplated. A definition
of this kind is a proposition susceptible of proof and illustration; and
therefore whatsoever is erroneously assumed to be the proper meaning of
such a term, may be refuted. But those persons who take every thing upon
trust, and choose both to learn and to teach mechanically, often become so
slavishly habituated to the peculiar phraseology of their text-books, that,
be the absurdity of a particular expression what it may, they can neither
discover nor suspect any inaccuracy in it. It is also very natural even for
minds more independent and acute, to regard with some reverence whatsoever
was gravely impressed upon them in childhood. Hence the necessity that all
school-books should proceed from skillful hands. Instruction should tell
things as they are, and never falter through negligence.

18. I have admitted that definitions are not the only means by which a
general knowledge of the import of lang