ust necessarily be
for ever itself only, and not an other.

OBS. 2.--Our common words, then, are the symbols neither of external
particulars, nor merely of the sensible ideas which external particulars
excite in our minds, but mainly of those general or universal ideas which
belong rather to the intellect than to the senses. For intellection differs
from sensation, somewhat as the understanding of a man differs from the
perceptive faculty of a brute; and language, being framed for the
reciprocal commerce of human minds, whose perceptions include both, is made
to consist of signs of ideas both general and particular, yet without
placing them on equal ground. Our general ideas--that is, our ideas
conceived as common to many individuals, existing in any part of time,
past, present, or future--such, for example, as belong to the words _man,
horse, tree, cedar, wave, motion, strength, resist_--such ideas, I say,
constitute that most excellent significance which belongs to words
primarily, essentially, and immediately; whereas, our particular ideas,
such as are conceived only of individual objects, which arc infinite in
number and ever fleeting, constitute a significance which belongs to
language only secondarily, accidentally, and mediately. If we express the
latter at all, we do it either by proper names, of which but very few ever
become generally known, or by means of certain changeable limitations which
are added to our general terms; whereby language, as Harris observes,
"without wandering into infinitude, contrives how to denote things
infinite."--_Hermes_, p. 345. The particular manner in which this is done,
I shall show hereafter, in Etymology, when I come to treat of articles and
definitives.

OBS. 3.--If we examine the structure of proper names, we shall find that
most of them are compounds, the parts of which have, in very many
instances, some general signification. Now a complete phrase commonly
conveys some particular notion or conception of the mind; but, in this
case, the signification of the general terms is restricted by the other
words which are added to them. Thus _smith_ is a more general term than
_goldsmith_; and _goldsmith_ is more general than a _goldsmith_; _a
goldsmith_, than _the goldsmith_; _the goldsmith_, than _one Goldsmith_;
_one Goldsmith_, than _Mr. Goldsmith_; _Mr. Goldsmith_, than _Oliver
Goldsmith_. Thus we see that the simplest mode of designating particular
persons or objects, is that