[Lingtyp] terminology

Martin Haspelmath haspelmath at shh.mpg.de
Tue Jul 31 16:47:34 UTC 2018

On 26.07.18 11:52, Johann-Mattis List wrote:
> There's a common practice in programming and mathematics to collect
> "problems" in the field. E.g., problems of graph theory, problems of
> algebra, etc. People know that these may be extremely hard to resolve,
> maybe even impossible, but they agree on the nature of the problem, and
> then try to build up solutions for them, which also produce new
> terminology, etc. In linguistics, I have the impression that we are often so overwhelmed by the feeling of how impossible it is to fully
> describe something in explanatory terms, that we forget to even state
> what the problem is, that we even forget to *describe* what we want to
> explain. We can definitely learn from our colleagues in other branches
> of science here in being a bit more pragmatic, less obsessed with the
> detail (at least at first), but also more systematic in applying our
> methods and techniques.

I fully agree with these sentiments, though I wouldn't say that it has 
to with a lack of "pragmatism" on the parts of linguists.

Rather, many linguists don't even *want* to make a difference between 
explanation and description, for principled reasons: They think that 
finding the right universal categories is the main research task of 
linguistics, and it's only after we have found them that we can describe 
languages properly. So linguistics is like chemistry: We first need to 
find the elements (and ideally the entire Periodic Table) before we can 
describe each compound properly.

This is of course the research programme of (traditional, 20th century) 
generative linguistics, but if one adopts a different research 
programme, where description and explanation are logically separate, 
then we can aspire to a more cumulative approach (I have a recent 
blogpots about the difference between the two approaches: 

So in order to agree on a common terminology, we don't need to agree 
that the concepts have a "deep reality" (= are established research 
results). We can just agree that they are useful for some purposes and 
that they have a clear meaning – they are not more than "units of 
measurements" (e.g. the SI units in physics, which are not research 
results either).


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

More information about the Lingtyp mailing list