ur things to be necessary. In citing these, I
vary the language, but not the substance or the order of his positions.
_First_, That we should speak and write words according to the
significations which belong to them: the teaching of which now pertains to
lexicography, and not to grammar, except incidentally. "_Secondly_, That we
should observe _the relations_ that words have one to another in sentences,
and represent those relations by such variations, and particles, as are
usual with authors in that language." _Thirdly_, That we should acquire a
knowledge of the proper sounds of the letters, and pay a due regard to
accent in pronunciation. _Fourthly_, That we should learn to write words
with their proper letters, spelling them as literary men generally do.

10. From these positions, (though he sets aside the first, as pertaining to
lexicography, and not now to grammar, as it formerly did,) the learned
critic deduces first his four parts of the subject, and then his definition
of grammar. "Hence," says he, "there arise Four Parts of Grammar;
_Analogy_, which treats of the several parts of speech, their definitions,
accidents, and formations; _Syntax_, which treats of the use of those
things in construction, according to their relations; _Orthography_, which
treats of spelling; and _Prosody_, which treats of accenting in
pronunciation. So, then, the true definition of Grammar is this: Grammar is
the art of _expressing the relations_ of things in construction, with due
accent in speaking, and orthography in writing, according to the custom of
those whose language we learn." Again he adds: "The word _relation_ has
other senses, taken by itself; but yet the _relation of words one to
another in a sentence_, has no other signification than what I intend by
it, namely, of cause, effect, means, end, manner, instrument, object,
adjunct, and the like; which are names given by logicians to those
relations under which the mind comprehends things, and therefore the most
proper words to explain them to others. And if such things are too hard for
children, then grammar is too hard; for there neither is, nor can be, any
grammar without them. And a little experience will satisfy any man, that
the young will as easily apprehend them, as _gender, number, declension_,
and other grammar-terms." See _R. Johnson's Grammatical Commentaries_, p.
4.

11. It is true, that the _relation of words_--by which I mean that
connexion between them, whic