Time depth in comparative linguistics

"In my paper to the aforementioned symposium (Renfrew 2000b), I argued
that it seems reasonable to request any scholar who rejects the
reasoning offered here, associating *PIE with the early spread of
farming from Anatolia, on the grounds that it is all "too early," to
set out clearly the rationale of the alternative (and later)
chronology proposed. References to very general perceptions of rate
of language change and of time-depth in language families will not
do, since many of these are circular, being themselves based upon a
priori assumptions about the dating of *PIE."

"The American linguist Terrence Kaufman, addressing this general
problem (Kauffman and Golla 2000: 47), has a carefully formulated
expression in which he suggests the chronological limits of the
comparative approach:

"The possibility of establishing a genetic grouping requires (a) the
availability of relevant data from the languages being compared, which
in turn usually requires (b) that the relationship is no older than
8000-10,000 years before the earliest date at which the languages are
documented."

"Following this rule of thumb, and dating the earliest Hittite records
to c. 1400 B.C., with the earliest Mycenaean Greek only a little
later, would give an earliest date for *PIE between c. 11, 400 and c.
9, 400 B.C. (Renfrew 2001, p. 39)."

Renfrew, Colin (2001), "The Anatolian Origins of Proto-Indo-European
and the Autochthony of the Hittites," in Greater Anatolia and the
Indo-Hittite Language Family, Robert Drews (ed.), Journal of
Indo-European Studies Monograph Number 38.

M. kelkar