[Lingtyp] interrogative verbal paradigms

Mon Oct 16 08:10:39 UTC 2017

This exchange again shows the need to define one's comparative concepts 
carefully – and it illustrates the common situation that we cannot 
easily transfer notions from familiar languages to all other languages.

One way to define "interrogative pronoun" might be as "a simple form 
that occurs in a content question and that occupies the position of the 
open parameter at issue".

This definition would comprise forms such as Italian "cosa" (which means 
'what?' or 'thing'), as well as French "que" (which is a bound form, in 
that it cannot occur by itself). Perhaps these are not "genuine" 
interrogative pronouns, in the sense that they are free forms (like 
Latin "quid?"), or in the sense that they only occur in an interrogative 
context (again like Latin "quid?"). But since even the forms of 
well-known languages may occur in non-interrogative contexts (e.g. most 
English wh-pronouns, used in relative clauses), it does not seem 
reasonable to define "interrogative pronoun" in a narrow sense.

I do share the feeling that Abkhaz and Abaza are unusual in some way, 
but I'm not sure how best to express this with clear comparative 
concepts. These elements may have been described as part of "verbal 
morphology", but this is more a claim about the orthography than about 
any clear property of the language.

Martin

On 15.10.17 10:49, Peter Arkadiev wrote:
> Dear Dmitry,
>
> many thanks, this is very helpful! I will have a close look at your work.
> When I said that Abaza and Abkhaz lack genuine interrogative words I meant interrogative word based on interrogative roots. Perhaps I am naive, since I have never worked closely on interrogative constructions before, but in all languages I know interrogative words, even if morphologically complex, are based on roots with interrogative meaning. In Abaza and Abkhaz, in contrast to the other NW Caucasian languages, there don't seem to be any such roots. In order to form a content question you need to embed whatever lexical root into an appropriate interrogative morphological pattern. True, if you consult a dictionary or a traditional grammar, you will find words for "who" and "what", but on closer inspection they turn out to be regular interrogative forms based on semantically general roots such as "possession". The only minor exception to this pattern in one of the words for "who" which has undergone some phonological erosion, but it still carries its regular origin on its sleeves in quite a transparent way.
>
> Many thanks again and best regards,
>
> Peter
>
> -- 
> Peter Arkadiev, PhD
> Institute of Slavic Studies
> Russian Academy of Sciences
> Leninsky prospekt 32-A 119991 Moscow
> peterarkadiev at yandex.ru
> http://inslav.ru/people/arkadev-petr-mihaylovich-peter-arkadiev
>
>
> 15.10.2017, 01:10, "Idiatov Dmitry" <honohiiri at yandex.ru>:
>> Dear Peter,
>>
>> When you say “there are no genuine interrogative pronouns in these languages; those elements that are described as such are in fact just interrogative verbal forms of the type just described”, you necessarily imply that interrogative pronominals must be nouns, and moreover, words without any morphological structure. That’s a matter of definition and everybody is free to use their definitions, but I think this one makes languages look unnecessarily more different than they really are and complicates things when you want to compare across languages.
>>
>> I suggest your question should better be formulated as follows:
>>
>> 1. Are there languages where interrogative pronominals are based on a bound interrogative root?
>>
>> 2. Are there languages where such interrogative pronominals based on a bound interrogative root obligatory function as predicates or clauses? (like they are in Abaza and Abkhaz)
>>
>> The answer to question #1 is yes. Such languages are relatively numerous. The two common cases are (i) when the interrogative root must be additionally marked for gender, number or any other nominal category to be able to function as a pronominal and (ii) when interrogative pronominals are expressed with conventionalized noun phrases not based on nominal interrogative pronominals (such as ‘which person?’ for ‘who?’). I guess in most languages with such complex interrogative pronominals, the latter can at least function as nominal predicates. Abaza and Abkhaz belong to another type of languages with interrogative pronominals based on bound roots. In this type, such interrogative pronominals with a complex internal structure obligatory function as predicates or clauses. In other words, in such languages, interrogative pronominals are clausal constructions.
>>
>> My impression is that this type is not very common. However, I can only provide examples (some synchronic and some reconstructed cases) where the fact that interrogative pronominals are clausal constructions leads to the lack of differentiation between ‘who?’ and ‘what?’ interrogative meanings. You can find them in my PhD (downloadable at http://idiatov.mardi.myds.me/PhD.html): a number of Mayan languages (pp. 492-509), some Arawakan languages (pp. 523-529), possibly also some Tacanan (pp. 543-545) languages and Urarina (pp. 536-537). I actually also discuss the Abaza and Abkhaz pattern (pp. 271-277).
>>
>> Also see pp. 3-5 for some definitional issues as to what is best considered as an interrogative pronominal.
>>
>> Hope this helps!
>>
>> --
>> Dmitry Idiatov
>> ALT treasurer & membership manager
>>
>> LLACAN (CNRS - Inalco)
>> http://idiatov.mardi.myds.me/
>>
>> 14.10.2017, 22:59, "Peter Arkadiev" <peterarkadiev at yandex.ru>:
>>>   Dear typologists,
>>>
>>>   in Abkhaz and Abaza, two closely related Northwest Caucasian languages, content questions (a.k.a. wh-questions) can be encoded by verbal morphology alone, without any separate interrogative words (see Hewitt 1979a: 10-23 for a description for Abkhaz). This is achieved by adding an appropriate interrogative suffix or prefix to an appropriate relative verbal form (on relativization in Abkhaz, see Hewitt 1979b; on Abaza, see O'Herin 2002, chapter 8). Cf. the following three characteristic examples from Abaza (my own fieldwork data):
>>>
>>>   (1) j-ʕa-ḳa-ŝá-ja? rel.abs-dir-loc-fall-what 'What fell?'
>>>
>>>   (2) w-ʕa-z-rə-há-da? 2sg.m.abs-dir-rel.erg-caus-fear-who 'Who frightened you?'
>>>
>>>   (3) w-ʔa-bá-nχa-wa? 2sg.m.abs-rel.loc-qadv-work-ipfv 'Where do you work?'
>>>
>>>   -ja and -da are interrogative suffixes with non-human resp. human reference (conspicuously featuring the j- and d- prefixes, which index non-human resp. human 3rd person singular absolutive arguments), while -ba- is an interrogative prefix for adverbial questions. Relativization is expressed by means of prefixes that indicate the role of the relativized or questioned element. This looks pretty much similar to a familiar (pseudo-)cleft strategy of forming content questions (e.g. Who is it who frightened you?). However, there are no genuine interrogative pronouns in these languages; those elements that are described as such are in fact just interrogative verbal forms of the type just described. The interrogative elements in Abaza and Abkhaz do not look cognate with the interrogative words attested in the other languages of the family.
>>>   I am wondering whether anything of this kind is attested in any other languages. I have looked at several reference works on questions (e.g. Siemund's article in HSK on typology and universals) and did not find there any mention of the Abkhaz pattern, neither is it mentioned in WALS. I would be grateful for any suggestions.
>>>
>>>   Thanks in advance and best regards,
>>>
>>>   Peter
>>>
>>>   --
>>>   Peter Arkadiev, PhD
>>>   Institute of Slavic Studies
>>>   Russian Academy of Sciences
>>>   Leninsky prospekt 32-A 119991 Moscow
>>>   peterarkadiev at yandex.ru
>>>   http://inslav.ru/people/arkadev-petr-mihaylovich-peter-arkadiev
>>>   _______________________________________________
>>>   Lingtyp mailing list
>>>   Lingtyp at listserv.linguistlist.org
>>>   http://listserv.linguistlist.org/mailman/listinfo/lingtyp
> _______________________________________________
> Lingtyp mailing list
> Lingtyp at listserv.linguistlist.org
> http://listserv.linguistlist.org/mailman/listinfo/lingtyp

-- 
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