LINGUIST List 4.397

Mon 24 May 1993

Disc: The number universal

Editor for this issue: <>

Directory

1. , relevance of number universal

In a recent LINGUIST posting, Paul Kershaw questions the "utility" of
the universal that the plural is never consistently more marked than the
singular.
 Since judgments of "utility" depend on one's goals, there cannot be a general
answer to this question, but what bothers me is Kershaw's suggestion that
work in linguistics should be strictly divided among semanticists/morphologists
on the one hand and "socio- and anthropological" linguists on the other.
 The problem is not only that it is never clear at the beginning what will
turn out to motivate a linguistic generalization -- strictly system-internal
factors or more general cognitive/sociological/anthropological factors.
The problem is also that very little may be left to semanticists and morphol-
ogists if they decide to disregard all facts that may have an extra-linguistic
motivation. Take another example, the suffixing preference, discussed in
Christopher Hall's 1992 book "Morphology and mind". Hall shows that only an
approach that combines structural, diachronic and psychological considerations
can account for the strong cross-linguistic prefrence for suffixes over
prefixes. His explanation is ultimately in terms of system-external factors,
but in my view an impoverished morphology whose interests are so narrow that
it does not worry about such things cannot claim the right to represent the
whole field of morphology.
 So clearly, yes, the number universal is highly relevant for both
morphology and semantics, and linguists who work on language structure should
cooperate with socio- and anthropological linguists, rather than divide
language up among them.

Martin Haspelmath (Free University of Berlin)