The Classification of Languages 199
contribute nothing to the meaning of a statement. Thus grammatical
gender (p. 113) is completely absent
Where we draw the hue between a language which is piedominantly
agglutinating or isolating depends on where we draw the line between
a word and an affix. If we do not know the history of a language, it is
not easy to do so We do not recognize words such as except or but as
separate entities because they are names of things at which we can
point or because they stand for actions we can mimic We distinguish
them from affixes such as mis- or anti-> because we can move them about
in the sentence. Now this test is straightforward because of the charac-
teristics of English word-order For example, we put prepositions on
the one hand, and pointer-words or adjectives on the other, in front of
a noun A pointer-word with two or more adjectives, adverbs and
conjunctions can separate a pieposition fiora a noun When the adjec-
tive comes after the noun, as it usually does in French, the distinction
is not so sharp, and it is less sharp in some Indie vernaculars The
Hindustani (p 412) adjective precedes and the directive follow** the
noun. If these postpositions—we cannot rightly call them prepositions
—never strayed further afield, there would be nothing to distinguish
them from case-affixes like those of Finnish.
Even the status of a pronoun as an independent element of living
speech is difficult to assess by any other criterion The reader who
knows some French will realize that the pronouns je> me, tu, te> il3 etc,
never stand by themselves When a Frenchman answers a question with
a single word, he replaces them by mois tois lui, etc We recognize them
as words by their mobility in the sentence That je or il do not always
stand immediately in front of the verb is due to three accidents of the
French language, i e. the fact that the pronoun object and the negative
particle ne precede the verb, and the use of inversion for question
formation By the same token (p 198) we ought to call the personal
suffixes of the Finnish verb3 pronouns
Thus the distinction between an affix and a particle is clear-cut only
when the conventions of word-order permit the independent mobility
of the latter. We are entitled to speak of a language as isolating when,
as in Chinese vernaculars, great mobility of unchangeable elements is
characteristic of it. When we speak of a language as agglutinating, we
usually mean that a clear-cut distinction between particle and affix is
impossible because any of the formal elements described by either of
these names occurs in a small range of combinations with recognizably
separate words, e g those we call nouns, adjectives, or verbs Some
grammarians apply the epithet agglutinative to any language with a