ON ETHNOLOGY. 
291 
second proved, that this phenomenon results from a general law; and in 
order to arrive at this law, we allowed ourselves no assumption, as far as we 
are aware, except that everything expressed in language, which is tlie ex- 
pressjon of rej^on, must originally have been reasonable, and therefore a 
irutli and ^lity. The question, Ciin a language be supposed to begin 
with mflexjons ? appears to as to iin])Iyr an absuniiiy. But «o docs the 
nret of all questions: why every wonl must be originally a true and 
adequate expression of the mind? And the examination of the facts 
shows us how- that law operates. First, iuHi xions, ns we have seen, resolve 
tUeraseives, whenever wc have the im.-ans of observing their formation, into 
worn-out prepositions or postpositions: but these again wc fouml, in the 
instances we examined, to have been in an earlier stage substantial words, 
nouns or verbs. We further found, that, when flexions are worn-out, and 
Mine event brings on u new secondary fonnatiou, worn-out flexions call 
forth the formation of a new affix or suffix from the class of particles. 
rims the line of progress runs in the tlirecUun of an increase in the 
nimiber of words formal, that U to say, of words serving for tbi* formative 
I'urpo^ of the mind. This coineidi-s with the nectasary purpose of all or¬ 
ganic language, to constitute and inork till the component parts of a sentence. 
Now It is clear that no word, which has once ceased to he full or substantial, 
can ever become so again; it has lost iu substantial, independent life, and its 
distinct substantial signification. It becomc-s an algebraic sign, and more or 
ess nnintelhgible in it.self'. The more substantial and iudependrtjt state is, 
oeoessarily tlie more anin’ent in any fine of devrlopmeiiL 
Thus much we can establmh by following out the logical prticess wc have 
underUken toexplnin. But this method alone cannot bring us further. Logi¬ 
cally, it is impossible to define the diflerent classes of this second great family 
of languages, otherwise than by establishing that the more the single 
words in a sentence are regaided as unchangeable, and their position in the 
sentence as the sign of the part they represent iu it, the nearer such a lan¬ 
guage must be to the Hrat chms. But whether, for instance, the system of 
agglutiriution or incorporation of the American ami the Basque languages is 
a proof of a backwardnens in tlm stage of development, compared with the 
use of affixes, must depenil upon concomitant circumsUincea. It certainly 
will be so, whenever tlie aftix-laiiguages are freer from the symnietric con- 
striiction ol a sentence, ami tlie isolation of the single words from each other. 
1 he great fact upon which we insist here, is this: every primitive language 
must be compoaml of words which are absolutely inorganic, because in this 
way alone the origin and the progress of word-forming, and the origin and 
Uevtlopment ol languages can be explained r.ilionally. 
Application of the prt^ding intfuiries to the Problem of the Classijicaiion 
of the. Egt/ptian Lunguage and of Languages in general. 
We shall rt'capitulatc briefly the results of the two preceding Investigations, 
have first examined some striking phenonieim in the formation and the 
wmponent parts ol laiignage, ami wp have then endeavoured to explain 
them by a general piiiloeopiiical indnetion. 
By the first we believe w« have fslabli«hnd the constancy of the following 
pheuomenu. The Jir.st is, that every language has in itself an element of 
become the eU-inent of death to the old, and 
of life in a new one. I he constant action of the mind upon the articulate 
exp^sioD of substantiality prevaili gradually, but necessarily, over the 
}wiuvcne9s of tills substantiality, and makes single words subservient to 
the expression of all that belongs to the mind ; of relation, outwardly of time 
u2 
