[Lingtyp] language typology, linguistic typology, comparative linguistics

[Martin Haspelmath]

On 28.02.18 13:40, Balthasar Bickel wrote:
> In terms of daily research and teaching we don’t see much difference 
> anymore between historical/evolutionary linguistics and typology, so 
> we don’t really care about the traditional associations that the label 
> “comparative linguistics” has. 
Yes, and I would even argue that the activities traditionally subsumed 
under "language typology" can only be defined in such a way that they 
comprise historical-genealogical studies as a special case. Dan Slobin 
cited the following definition of typology, which actually sounds much 
more like a definition of "(implicational) universals research":
> /The essence of typology lies in structural traits – ranging from 
> sound and grammar to lexicon and discourse – that could vary 
> independently from language to language but actually do vary together, 
> setting limits to cross-linguistic variation and defining the 
> groundplans on which languages are constructed. The discovery and the 
> explanation of such interdependencies.../

But a lot of actual research in our community (and research published in 
LT) has been concerned with setting up types, i.e. defining comparative 
concepts and pointing to previously unsuspected connections, rather than 
with discovering new universals or explaining older ones. As Randy 
LaPolla points out, by describing a new language, one can also make an 
important contribution to this endeavor (Dixon's description of Dyirbal 
helped us understand comparative concepts "accusative/ergative" better, 
but it didn't contain any universals).

So "comparative linguistics" simply consists in systematic comparison of 
language systems (and also of social embeddings, in comparative 
sociolinguistics), and in a second step, in finding explanations for any 
generalizations found in this way.

In many cases, similarities between languages (including cases of 
covariation of structural traits!) turn out to be due to shared 
inheritance from a common ancestor, and these cases have long been 
particularly well understood (and studied under the rubric of 
historical-genealogical linguistics). But how exactly to draw the 
boundary between shared-inheritance explanations and other explanations 
(or lack of explainability) is often unclear, and is the object of 
current research.

In the past I have sometimes defined "typology" as "non-genealogical 
comparison of languages", but this delimitation works only if one 
already knows what kind of explanation is correct.

One might also compare "comparative linguistics" and 
"historical-genealogical linguistics" to the contrast between 
"primatology" and "anthropology" – strictly speaking, the former 
subsumes the latter, but it is natural that anthropologists have their 
own conferences (and their own specialized subfields such as economics 
or sociology) because so much more is known about Homo sapiens than 
about the other primates.

Martin

-- 
Martin Haspelmath (haspelmath at shh.mpg.de)
Max Planck Institute for the Science of Human History
Kahlaische Strasse 10	
D-07745 Jena
&
Leipzig University
IPF 141199
Nikolaistrasse 6-10
D-04109 Leipzig





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20180228/c16c72a7/attachment.htm>