192 The Loom of Language
Though we know little about early culture-contacts, common sense
tells us that what has happened in histoiioil times must also have
happened before
It has also been said that the primitive Ai van-speaking tribes could
count at least as far as one hundred This does not necessarily follow
from the fact that names for 2 or for 3 o. for 10, etc, are alike You
cannot exchange goods without being able to count Jt is therefore
quite possible* that Aryan-speaking tribes boirowcd the art of counting
from an outside source, or that it diflused from one branch of the family
to its neighbours Indeed, numerals are the most indefatigable wan-
derers among words, as indefatigable as alphabets In the language of the
Gypsies^ an Indie tribe., the namCvS for 7, 8, and 9 are modern Greek,
whereas those for 5 and 10 aic Indie. In the Fmno-Ugnan group, the
word for 100 is borrowed from Iranian, and Hebrew uhesh (6) and
scheba (7) are supposed to be derived from Aryan, while the Hebrew
name for 8 is assumed to be Egyptian But there is no need to go so far
back The English dozen and million have been taken ovex in compara-
tively recent times from the Romance languages
German philologists have not been content to draw encouraging
conclusions from words winch are alike and have the same meaning in
all the Aryan languages They have also speculated about the signifi-
cance of words which do not exist Of itself, the fact that the Aryan
family has no common term for the tiger does not indicate that the
pro to-Aryans jnhabited a region where there were no ugcrs. Once the
hypothetical Urvolk staited to move, mbcs which went into colder
regions would no longer need to preserve the word for it- If we are
enutled to deduce that the Hast did not use salt because the Western
Aryan word for the mineral docs not occur in the Indo-Iranian tongues,
the absence of a common Aryan word for milk must force us to con-
clude that proto-Aryan babies used to feed on something else.
LANGUAGE FAMILIES OF THE WORLD
In a modern classification of the animal kingdom taxonotrusts unite
many small groups, such as fishes., birds and mammals, or Crustacea.,
insects and arachmda (spiders and scorpions) m larger ones such as
vertebrates and arthropods Beyond that point we can only speculate
* Philologists sometimes justify emphasis oa similarity of number-word**
on the ground that they also shai c general phonetic features characteristic oJt a
language as a whole. This is also true of words which have undoubtedly been
borrowed;, and is easily explained by the phonetic habits of a people