words expreffing acting or fuffering. Thefe Ch. 1.
relations, it is evident, must be very
various and numerous; and they must be
expreffed either by feparate words, or by
fome change of one or other, or both, of
the words betwixt which the connection is
to be expreffed.

The last requifite of language I mentioned, refpects the found: as to which three things may be observed, ist, That the words, in order to exprefs fuch a prodigious variety of things, fhould be very much varied in the found. It is therefore neceffary that they fhould not confist of vowels only, or a few confonants, like the

words of the barbarous languages, but be
diftinguished and articulated by as many
confonants as poffible, but fo as not to
render the found harth and difagreeable.
For, 2do, A language fuch as we are speak-
ing of, must be of eafy pronunciation;
and not only the facility of the operation

of the speaker must be studied, but also
the pleasure of the ear of the hearer. 3tio,
The words must be of a moderate length,
not, like thofe of the barbarous languages,
enormously long; and particularly the ra-
dical words must be fhort, otherwise there
VOL. II.
C

cannot

Ch. 1. cannot be either compofition or derivation without enlarging the words immoderately. The declinable words alfo, in order to admit a fufficient variety of inflection, fhould be of a moderate length; and not only for these reasons, but for the fake of the eafy pronunciation of the language, the words ought not to be exceffively long.

Having premised thefe general obfervavations, which will ferve to explain what is to follow, I will proceed to confider the feveral parts of which language is compofed.

Ch. 2.

CHA P. II.

The works of art prior to the art itself.-The analytical method followed in this inquiry.— The formal part of language to be firft analyfed. Both the form and matter must have been analyfed before the writing art was invented. -The nature of that discovery.

A

LL the works, both of nature and of art, are compounds, which the fenfe prefents to the mind. These it is the bufinefs of science to analyfe, and refolve into their

principles,

principles, or conftituent parts. But not Ch. 2. only the works of nature exifted long before any fuch analysis was made, but even those of art, at least to a certain degree. For we are not to imagine, that arts were invented a priori, by discovering the principles firft, and from thence deducing the confequences; on the contrary, men began by practifing; very rudely and imperfectly no doubt at first; but as they improved the practice, they began to discover the principles, and at last acquired fcience enough to analyse the art, and deduce it from its principles. In this manner all arts have been invented, and among others the art of language. But the progress was very flow from practice to principles; and accordingly men had the use of language long before they knew any thing of the grammatical art, by which language is analyfed; and in like manner men fung, and played on inftruments, while yet no scale of mufic was known, nor any art invented by which a tune could be refolved into the feveral notes of which it is compofed. And in many nations of the earth at this day, these and many other operations of art are performed, without knowledge of the

Ch. 2. the art itself; that is, of its principles. As therefore the compound is first in order of time, at least with respect to us and our perceptions; fo compofition in the several arts, and particularly in language, is much easier than the analyfis: for by imitation merely we can compofe, or by natural fagacity without imitation, which was the cafe of the firft inventors of arts; but we cannot analyfe without fcience. Accordingly, how many people do we fee, that have not only the ready ufe of language, but speak very properly, without the least knowledge of the grammatical art? But though compofition be fo much easier and more obvious than analysis; yet this laft is the method of fcience, being that which conftitutes the very nature and effence of science; for nothing is fcientifically treated of, that is not refolved into its elements, or first principles. So that fcience does not follow the order of our perceptions, which begin with the compound, but the order of nature, accor-, ding to which the elements or principles of all things are first. As therefore we profefs to treat of language fcientifically, we fhall begin with analyfing it, and then we

fhall

fhall proceed to the composition of it; and Ch. 2. in this method we think the most proper, because the compound, in this matter of language, is well known to every one, as the composition is practised by every one; whereas, if the compound were not fufficiently known, it might be proper to proceed in a different method, and begin with

it.

All the works of art, as they are compofed by man, fo they can be analysed by him* And the best method of attaining

a

Every analyfis is a divifion, but every division is not an analysis; that is to fay, in logical language, divifion is the genus, and analyfis the fpecies. Any whole may be divided into the parts which compofe it, as a body may be divided into its different members; but fuch a divifion is not that which we call analysis, because the members of a body are parts of the body, when it is conftituted or formed, but they are not the principles which conftitute or form it, or, in other words, they are not the principles or elements of the body. Now it is the divifion into these last that I call analyfis; which therefore differs from the other divifion in this, that it divides the fubjects into parts more minute and fubtile, not obvious to fenfe or common apprehenfion; but which being dif covered, fhew the nature of the fubject, because they thew the principles which constitute it. And it is for this reason that analyfis is the method of fcience. The me thod of divifion, or diæretic method, and the analytical method, being two ways of investigating the idea or de

finition