bject, writes thus: "The practical
instruction of the _countinghouse_ imparts a more thorough knowledge of
_bookkeeping_, than all the fictitious transactions of a mere _schoolbook_,
however carefully constructed to suit particular purposes."--_New Gram._,
p. vii. But _counting-house_, having more stress on the last syllable than
on the middle one, is usually written with the hyphen; and _book-keeping_
and _school-book_, though they may not need it, are oftener so formed than
otherwise.

OBSERVATIONS.

OBS. 1.--Words are the least parts of significant language; that is, of
language significant in each part; for, to syllables, taken merely as
syllables, no meaning belongs. But, to a word, signification of some sort
or other, is essential; there can be no word without it; for a sign or
symbol must needs represent or signify something. And as I cannot suppose
words to represent external things, I have said "A _Word_ is one or more
syllables spoken or written as the sign of some _idea_." But of _what_
ideas are the words of our language significant? Are we to say, "Of _all_
ideas;" and to recognize as an English word every syllable, or combination
of syllables, to which we know a meaning is attached? No. For this, in the
first place, would confound one language with an other; and destroy a
distinction which must ever be practically recognized, till all men shall
again speak one language. In the next place, it would compel us to embrace
among our words an infinitude of terms that are significant only of _local_
ideas, such as men any where or at any time may have had concerning any of
the individuals they have known, whether persons, places, or things. But,
however important they may be in the eyes of men, the names of particular
persons, places, or things, because they convey only particular ideas, do
not properly belong to what we call _our language_. Lexicographers do not
collect and define proper names, because they are beyond the limits of
their art, and can be explained only from history. I do not say that proper
names are to be excluded from grammar; but I would show wherein consists
the superiority of general terms over these. For if our common words did
not differ essentially from proper names, we could demonstrate nothing in
science: we could not frame from them any general or affirmative
proposition at all; because all our terms would be particular, and not
general; and because every individual thing in nature m