Andrew Carstairs-McCarthy on The Origins of Complex Language
Crossing the Cognitive Rubicon
Most linguists regard the origin of language as inaccessible to serious
investigation. Recent advances in biological anthropology, primate
ethology, and neuroscience make this attitude seem too pessimistic. But
specifically linguistic evidence can be brought to bear also, of a kind
analogous to the 'starred' sentences that have figured so prominently in
syntactic theory for the last forty years. In this book, I use linguistic
evidence to build the case for an evolutionary scenario with clearcut,
testable implications for language as it is today — a scenario that
reconciles evidence from grammar with what is known about the evolution of
the brain and the vocal tract, and about the communicative abilities of
other primates. The evidence suggests that the fundamental neural
development underlying language was for control of the vocal tract, not for
conceptual structure or predicate-argument structure — inverting what is
usually assumed in studies of human evolution.
A starred sentence is a sentence in an imaginary language (pseudo-English,
pseudo-Warlpiri or whatever): a conceivable alternative to actual English
etc. whose nonexistence is typically attributed to violation of some
principle of Universal Grammar. (Similar use is made of starred sentences
even in frameworks that reject UG as an autonomous mental 'organ'.) But
what about conceivable alternatives to UG itself? We cannot truly claim to
understand language until we have comapred actual UG (grammar-as-it-is)
with conceivable alternative UGs (grammar-as-it-might-have-been), and
suggested reasons why UG does not conform to one of these alternative
patterns. In this book I invite readers to consider various alternatives,
with a view to weakening the understandable tendency to assume that
grammar-as-it-is is, in fundamental respects, the only way that grammar can
be. Even the distinction between sentences and noun phrases turns out to
be not inevitable nor even particularly 'natural', from the point of view
of either communication or knowledge representation. Why does it exist,
then?
Evolution does not arrive at perfect solutions to design problems; rather,
it tinkers with what is available. For this reason, an organism may have
characteristics that are not 'functional' in its current environment, but
which for that very reason are particularly revealing about its
evolutionary history. Some central but puzzling characteristics of syntax
can be made sense of in this fashion: they fall into place as residues of a
kind of architecture that originally evolved in the neural control of
syllabically organized vocalization, with regular alternations in sonority.
Seen in this light, certain classic puzzles of syntactic theory dissolve;
for example, the lack of a satisfactory crosslinguistic definition of
'subject' is explained by the fact that subjects represent one of the
syntactic residues of the syllabic onset, for which no fundamental
cognitive motivation (semantic or pragmatic) is to be expected.
Within linguistics, the book's conclusions are supported by evidence not
only from syntax but also from vocabulary acquisition and inflectional
morphology. Outside linguistics, the book has startling implications for
logic and the philosophy of language. The distinction between reference
and truth is closely tied to the distinction between noun phrases and
sentences. Could it be a mere byproduct of that distinction, then? I
argue that the answer is yes. Frege, Wittgenstein, and Strawson, who might
have been expected to supply a solid nongrammatical motivation for the
truth-reference distinction, fail to do so. This distinction thus seems
likely to be a mere residue of syllable structure, via syllable-derived
syntax. Consequently the distinction between 'declarative' knowledge
('knowledge-that', expressed in sentences) and 'procedural' knowledge
('knowledge-how') dwindles in significance; access to declarative
knowledge, attained only by humans, can no longer be regarded as major
Rubicon in cognitive evolution.
These conclusions are likely to be controversial. But, however the debate
on them develops, I hope to have shown that fruitful inquiries can result
from comparing grammar-as-it-is with other alternatives in the space of
possible grammars.
|
VIEW BASKET
