Evolution of Language
- From: "J.A. Legris" <jalegris@xxxxxxxxxxxx>
- Date: Tue, 24 Jul 2007 18:14:48 -0700
I'm getting excited! For a nice review of what's going on in language
evo, have a look at Zuidema's 2005 Ph.D thesis at:
http://staff.science.uva.nl/~jzuidema/thesis/zuidema05phdthesis-compact.pdf
WARNING: 177 pages long! (7.2 Mbytes)
Sample (pp. 3-4 ):
--------------------------------------------------------------------------------
1.2 How to Study the Evolution of Language
Language evolution is of course not the only field where it is
difficult to find empirical evidence: in cos-
mology, general relativity, paleontology, origins of life and many
other fields researchers have struggled
to find ways to test the coherence of their theories, and to test the
sometimes very indirect predictions that
follow from them. The solution in these fields has not been to abandon
the interesting questions, but to
formalise the theories, and to work out testableconsequences, even if
it requires many intermediate steps.
For the evolution of language this requires the development of
complete and formal scenarios that explain
the evolution of the unique features of human language (which are
testable in modern humans) from a
plausible precursor state in the human lineage that is not unique in
nature (and hence, open to empirical
investigation through comparative research).
Only when we have precise scenarios of the evolution of language and
worked out ways to test empir-
ically the plausibility of one scenario against another, can we
conclude - if that turns out to be the case
- that there are too many alternative scenarios consistent with the
available data. In my view, we have
certainly not reached this stage yet. In this thesis I work out a
number of formal requirements for theories
of language evolution, and argue that existing models and theories -
including models presented in this
thesis - do not yet meet all requirements. The thesis is complementary
to interesting work arguing that
much more empirical data can and should be gathered, and reporting
results from such studies (Hauser,
1996; Hauser etal., 2002; Fitch & Hauser, 2004).
Of course, many other researchers have emphasised the need for
scenarios of language evolution to be
(i) testable (for instance, Lieberman, 1984), (ii) complete (for
instance Botha, 2003; Bickerton, 2003b) and
(iii) formal (for instance, Batali, 1998; Steels, 1997; Nowak etal.,
2002). Of those features, formalisation is
perhaps most controversial, at least in the way this has been worked
out in current models. Derek Bickerton,
for instance, has been vocal in his criticism of the
oversimplifications in mathematical and computational
models (e.g. Bickerton, 2003b). There are two responses to such
criticism, formulated nicely by Cavalli-
Sforza & Feldman (1981) and Batali (1998). The first emphasises the
precision that comes with formal
models:
"Our position, however, is that a mathematical theory is always more
precise than a verbal
one, in that it must spell out precisely the variables and parameters
involved, and the relations
between them. Theories couched in nonmathematical language may
confound interactions
and gloss over subtle differences in meaning. They avoid the charge of
over simplification at
the expense of ambiguity." (Cavalli-Sforza&Feldman,1981,p. vi).
The second response emphasises the heuristic value of formal models,
that helps the researcher to
explore consequences of a set of assumptions that might me overlooked
in verbal theorising:
"Mathematical and computational models provide a way to explore
alternative accounts of the
emergence of systems of communication. If the consequences of a model
are consistent with
expectations based on intuitions or speculation, they might obtain a
small measure of support.
But more interestingly (and, as it happens, more often), the
consequences of a model may
deviate from expectations. In working out the reasons for the
differences, one can potentially
develop a richer set of intuitions. Models are thus valuable to the
degree that they explicitly
illustrate the consequences of the set of assumptions they embody.
This may be even more
important than whether those assumptions are correct." (Batali,1998,p.
406).
Both the precision and the exploration aspect of formal modelling will
play a role in this thesis.
--------------------------------------------------------------
--
Joe
.
- Follow-Ups:
- Re: Evolution of Language
- From: feedbackdroid
- Re: Evolution of Language
- Prev by Date: Re: A preliminary look at Spoonerisms
- Next by Date: Re: role of language in human though process
- Previous by thread: pandemonium: a paradigm for learning
- Next by thread: Re: Evolution of Language
- Index(es):
Relevant Pages
|
