[Lingtyp] comparative concepts

Fri Jan 22 15:33:35 UTC 2016

Dear Eitan (and All),

Thank you for your thoughts. There is a lot in your message to think about but let me just get back to the question of whether, say, French adjectives and Italian

ones do or - as you say- do not form the same descriptive category. By what definition of the concept "descriptive category" is this so? What would French and Italian adjectives have to look like in order for us to be able to class them as members of the same descriptive category? Is this an empirical issue, or should we adopt a principle

by which constructions of different languages can never be members of the same descriptive category? If the latter, why should we adopt such a principle?

Best,

Edith

________________________________
From: Lingtyp <lingtyp-bounces at listserv.linguistlist.org> on behalf of Eitan Grossman <eitan.grossman at mail.huji.ac.il>
Sent: Friday, January 22, 2016 4:31 AM
To: Martin Haspelmath
Cc: LINGTYP
Subject: Re: [Lingtyp] comparative concepts

Dear all,

Thanks for this fascinating discussion!

I've always found the distinction between descriptive categories and comparative concepts to be extremely helpful, and have thought a bit about some of the issues that Edith Moravcsik raises. If one takes the distinction seriously, then yes, adjectives in French and Italian are different descriptive categories, and so are the categories of two different stages of a language. In fact, types of historical changes (assimilation, pathways of grammaticalisation, etc.) are also comparative concepts, in a sense, since the individual processes of change in actual languages are usually distinct.

The distinction between descriptive categories and comparative concepts also helps to clarify the ways in which borrowings are integrated into a target language, as I've argued in a recent paper (https://www.academia.edu/13522922/Language-specific_transitivities_in_contact). Unlike Martin Haspelmath's definition of comparative concepts, I wonder if they might nonetheless be cognitively real in a sense, at least for multilingual speakers who often match structurally distinct descriptive categories due to a common function (as in Yaron Matras' notion of 'pivot').
[http://a.academia-assets.com/images/open-graph-icons/fb-paper.gif]<https://www.academia.edu/13522922/Language-specific_transitivities_in_contact>

Language-specific transitivities in contact<https://www.academia.edu/13522922/Language-specific_transitivities_in_contact>
www.academia.edu
This paper argues that transitivities are language-specific descriptive categories, and the comparison of donor-language transitivity with target-language transitivity reveals fine-grained degrees of loan-verb integration. Based on a comparison of

An additional advantage of the distinction between descriptive categories and comparative concepts that (I think) hasn't yet been pointed out here is that it can help typologists to maximize variables for cross-linguistic comparison. If two language-specific categories are taken to be instantiations of the same cross-linguistic category, the differences between them are usually downplayed. On the other hand, discussions often focus on whether something really 'is' an instance of Crosslinguistic Category X (say, noun incorporation). If it's similar enough but not identical to a category identified in another language, it can be dubbed 'Pseudo-Category X' (say, pseudo noun incorporation) and then we have a new term in linguistics.

However, if one works with a broad comparative concept, everything that is not part of the definition can be turned into a variable. This seems to be especially valuable in frameworks like the one proposed by Balthasar Bickel in his 2010 and 2011 papers on multivariate analysis (e.g., http://www.zora.uzh.ch/57331/).

Although he is careful to distinguish his approach from those of Martin Haspelmath and Gilbert Lazard in these papers, I think that they are compatible, in that comparative concepts (e.g., 'clause linkage')  can be starting points for multivariate analyses, and multivariate analyses tell us what properties actually cluster together. This kind of bottom-up perspective can turn up really new insights.

Best wishes,
Eitan

Eitan Grossman
Lecturer, Department of Linguistics/School of Language Sciences
Hebrew University of Jerusalem
Tel: +972 2 588 3809
Fax: +972 2 588 1224

On Fri, Jan 22, 2016 at 11:55 AM, Martin Haspelmath <haspelmath at shh.mpg.de<mailto:haspelmath at shh.mpg.de>> wrote:
On 21.01.16 19:18, Edith A. Moravcsik wrote:
3/   IS THE ISSUE EMPIRICAL OR LOGICAL?

 As Östen Dahl has noted, it is important to clarify whether some or all other scientific inquiries in various fields also distinguish between descriptive categories and comparative concepts. How about cross-cultural studies, comparative literature, comparative religion, and the various fields of natural science? It seems implausible that the distinction would be linguistics-specific. If it is not, how is the distinction defined and utilized in other fields?

Comparative concepts are widely used in other disciplines when a comparative approach is adopted (I talked about this briefly in my 2010 paper, §9). I even found a paper published in a law journal that uses the term "comparative concept" (http://www.ejcl.org/22/art22-1.html).

Another example is the comparative study of folktales, where researchers use the "Aarne-Thompson classification index<https://en.wikipedia.org/wiki/Aarne%E2%80%93Thompson_classification_systems>". For biology and anthropology, I recommend Charles Nunn's book "The comparative approach in evolutionary anthropology and biology<http://www.press.uchicago.edu/ucp/books/book/chicago/C/bo11462152.html>". And for astronomy, the discussion about the classification of Pluto (is it a planet or not?) shows that the comparative concepts in other disciplines can also be pretty arbitrary. In comparative religion, terms like "clergy" are clearly useful for comparison, even though Catholic priests and Protestant ministers play very different roles within the system of the religion (and thus the fact that different denomination-specific terms are used for them is not an accident).

What may be special in linguistics is that the task of analyzing/describing an individual language is so challenging and absorbing. 90% of all linguists only ever study a single language, I think, and even those that adopt a comparative approach are usually very knowledgeable and concerned about analytical issues. Thus, our analytical terms are very prominent, and for many centuries, people have simply carried them over from one language (such as Latin) to another one (such as French, Russian, Persian and so on). That there is a problem with this dawned on them only in the early 20th century - and it required deep concern with Native American languages to understand it (those linguists who mostly focused on the bigger languages blissfully ignored the Boasian insights, including the generativists). I think in other fields, the confusion between analytical and comparative concepts was not so much of a problem, because the distinction was obvious - perhaps also because the diversity is more obvious in other fields.

(There my also be fields where separate comparative concepts may not be needed, because comparison in terms of universal analytical notions is sufficient. I'm thinking of chemistry - it's surely no accident that Mark Baker compared linguistics to chemistry in his utopian 2001 book "The atoms of language". It seems that chemists have indeed been successful in figuring out the universal blueprints of stuff, and they can compare kinds of stuff via the blueprints.)

Best,
Martin

--
Martin Haspelmath (haspelmath at shh.mpg.de<mailto:haspelmath at shh.mpg.de>)
Max Planck Institute for the Science of Human History
Kahlaische Strasse 10
D-07745 Jena
&
Leipzig University
Beethovenstrasse 15
D-04107 Leipzig

_______________________________________________
Lingtyp mailing list
Lingtyp at listserv.linguistlist.org<mailto:Lingtyp at listserv.linguistlist.org>
http://listserv.linguistlist.org/mailman/listinfo/lingtyp

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20160122/e9c976ea/attachment.htm>