[Lingtyp] Comparative notions

[Gilbert Lazard]

Sorry, Martin for duplicating.

Dear Martin,
I am now rereading your proposal (made in december 2016) of a workshop on comparative notions at the next ALT meeting in Zurich. I think it is a good idea. However I am a little sorry that in your bibliography you did not mention my own contribution to the question.
May I remind you that you discovered the idea of intuitive comparative notions in my 2006 book (La quête des invariants interlangues…)? In a review of that book, in 2008, speaking of what I said of the crosslanguage investigation of transitivity, you wrote : “Ultimately all these generalizations about transitivity are based on the concept of prototypical action, and this concept has no other foundation than the linguist’s intuitive feeling that it might be useful for comparison (as it turns out to be). When I first heard this idea, it struck me as rather strange, especially in the context of an approach that tries to make linguistics more rigorous (or  “scientific”, Lazard 1999a). However, once one recognizes that comparison cannot be based on categories, Lazard’s view of the tertia comparationis becomes cogent“ (LT 12 : 307). That was the very beginning of your own reflections and of all subsequent  and current discussions on the necessity of “intuitive conceptual frameworks“ (ICFs), as I call them, for language comparison. Was it not worthy of being recalled ? 
As for your questions, here are my positions :
a) All linguists know that all languages are different from one another. All linguists also feel that all languages in a sense resemble one another. But feeling is not knowledge. This is why languages must be compared in order to perceive more precisely what is common to different languages. If the investigation is successful, then it may result in the discovery of a common category, or, more probably, a “quasi-category“ approximately to be found in each of the languages compared. In such a case, it may be considered that there is a (possibly approximate) taxinomic relation between the common category and language-specific categories. — Anyway, each language must be described in its own terms. However, coding may be practically be based upon felt likeness, e.g., adjective in English and in Portuguese. It is only necessary to distinguish names designating language-specific categories (for instance, by writing it with an initial capital letter) and those only motivated by likeness.
b) As you know, I think that any notion may be used as an “intuitive conceptual framework“ (ICF), provided it is given a clear unambiguous definition. It is also necessary that the (parts of) languages to be compared have been described as systems, so that they can be checked against the ICF.
c) The notion chosen as ICF may be part of any region of the representation of the world, e.g., “prototypical action“. There is no reason for limiting it to the domain of grammar. But, of course, grammatical notions may also be used, if clearly defined, for our representation of languages is also part of our representation of the world. They may receive any kind of definition: there are no “hybrid categories“.
d) Generalities such as SOV, VOS, etc. belong to the domain of likeness. Such traditional notions as “subject” or “S”, etc., are fuzzy. I believe they are neither fully adequate nor entirely inadequate. They both reflect and hide deeper and more abstract realities that it is our task, for us typologists, to discover. For example, it is highly likely that the traditional notion of Subject include two different functions, which I have called “predication subject and “reference subject”: they appear separated both in ergative languages and in affective constructions (Lazard, Two possible universals…, LT 19, 2015).
e) Language description and typology are obviously different tasks. The descriptive linguist proceeds from form to meaning, along the semasiological path. The typologist, on the contrary, in language comparison cannot but proceed from concept (the ICF) to language facts, i.e. follow the onomasiologic path. However, comparison requires languages to be described as systems (cf. above b), which is a link between description and typology.
Best wishes

Gilbert Lazard,
49, av. de l'Observatoire,
F-75014 Paris

gilzard at orange.fr



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20170525/0759c6c5/attachment.htm>