202 The Loom of Language
pnatc for soup, and there is no difficulty about recognizing what a
spoon isa because all the spoons are produced accoiding to a standard
pattern The table mannerb of an amalgamating language are largely
moulded by a code of gentlemanly usclcssness You have a laige assort-
ment of tools before you* Whether you use a loik \v*th or without a
knife or a spoon depends on conventions ol social class without regard
to the texture ol the food
lo all the intimsic difficulties of learning a language such as Lauo>
old-fashioned grammarians and schoolmasteis have added the dis-
tracting pretence that such table manneib have a rational basis This is
false line grammar of an agglutinating language such as Finnish (or
Esperanto) is mainly concxrned with meaning rlhe giammai of an
amalgamating language such as Latin is mainly concerned with social
ritual 11 you hope to master a language such as Latin^ the question you
have to ask is not what any one of half-a-do/en diileient alUxes which
grammarians describe as trade-marks oi the ablative case signify They
have no imiqm meanmg Each case-alla of a Latin noun is the trade-
maik of a shelf of diversely assorted idioms Ihe business of the learner
who succeeds m emeigmg horn the fog of false rationality in text-books
of classical grammar is to find out in what .situations Latin or Greek
authors use these uihxc\s '1 he use of I aim case-forms is a social habit,
like eating asparagus with tbc fingeis. rl he only reason for making an
exception of asparagus is that the people with money do ?.o
Like the boundary between oil and water m a test-tube, the diflcicncc
between amalgamation and agglutination is not clear-cut, It would be
difficult to give good reasons ioi descubing the personal sullixcs of the
Celtic veib (or the verb ol some Indian vernacular) as amalgamating m
tonuadistintuon to agglutinating Flexions of this kind pass through
the stage of agglutination to amalgamation rl hey then propagate them-
selves by analogy, as when we stick the -a on the pu*k in lie pwk$ /m
car hew* Couvuiih'iis of scupi may greatly exaggerate 01 hide icguto-
ties or irregularities of the spoken language. The literary language of
Germany preserves a luxuriance of flexions which ore not cleaiiy audible
m the daily xntcicoui.sc of many Ganhms The ;>anne is> mote true of
French French script conceals a wealth of contractions which would
make a faithful transcription of French speech recall the charactciistics
oi some Amerindian dialects (p 215), Written English is more isolating
thaa Anglo-American as we speak it, because it frowus on joaany agglu*
tinativc contxactions of the pronoun or negative paruek (e.g.
*t) wath helper vo:b&»