[FUNKNET] grammaticalization of negatives/interrogatives

Wed Mar 16 05:44:25 UTC 2005

Dear Wolfgang,

Thanks for your thoughtful reply. I think you are on the right track in looking for cognitive operations that lead to this interesting grammatical phenomenon. I would only comment on your proposal that I think the locational construct is not X is THERE but X is HERE, because it is the absence from a proximate speech situation that I think would be more typical than from a distal one.

With regards,
Matthew

Mr Matthew Anstey
Charles Sturt University, School of Theology, Academic Associate
Free University, Amsterdam, PhD candidate 

St Mark's National Theological Centre
15 Blackall St
Barton ACT 2600
Australia

Ph:  +61 (0)2 6273 1572
Fax: +61 (0)2 6273 4067
Email: manstey at csu.edu.au 
Web: http://www.stmarksntc.org.au/html/staff/anstey.html

  _____  

	From: Discussion List for ALT [mailto:LINGTYP at LISTSERV.LINGUISTLIST.ORG] On Behalf Of Wolfgang Schulze
	Sent: Thursday, March 10, 2005 8:40 PM
	To: LINGTYP at LISTSERV.LINGUISTLIST.ORG
	Subject: Re: [FUNKNET] grammaticalization of negatives/interrogatives

	Dear Matthew,

	many thanks for having drawn our attention to the Semitic data. I do not want to comment upon them here (I leave it to comparative semitologists, as you did). Still, let me stress that the strong correlation between negation and interrogation strategies you have mentioned seems to be crucial for the understanding of either of them in many languages. 

		Many semitic languages show a probable development from an interrogative particle of place "where is ...?" to an negative existential "there is not ...". Similar to English, "Where's Pete?" that implies "Pete is not here".

	Note that in your example you describe the 'emergence' of 'constituent questions', based on a negated (locational) existential construction. Hence, we can state a common 'relational' (verbal) concept {BE=NOT=THERE ~ BE=WHERE}. Some folks have claimed that the underlying strategy is 'verificational': Accordingly, cognition gets into a state of hypothesizing that the applicability of a {X is THERE} pattern is justified: The actual input however does not stimulate the activiation of this pattern, leading to a 'negation' {it is NOT that X is THERE}. The 'tension' between these two cognitive 'states' provokes a verificational strategy, in case the 'pre-input 'hypothesis {X is THERE} is strong enough. Cognition now 'expresses' its hypothesis (where ever it may have come from) more than it simply 'asserts' the 'negative state'. From this we can assume that in case both strategies are linguistically encoded with the help of a common strategy, the 'verificational' version should include additional 'markers' that refer to the underlying 'hypothesis' {X is THERE}. 

	Therefore, we can draw the following picture (VER:FOC = Verificational Focus)

	{X is NOT THERE}    >   ASSERTION
	{X is NOT THERE} x VER:FOC    >   QUESTION [e.g. Intonation, Q-particles etc.]

	Naturally, it also can go the other way round (ASS = Assertion)

	{X is NOT THERE} x ASS >   ASSERTION [e.g. assertive particles]
	{X is NOT THERE} > QUESTION [+ speech act related strategies]

	Likewise, both strategies can be combined, or both focal strategies are lacking. Nevertheless, this correlation obviously is restricted to Constituent Questions that focus on location (or its metaphorization). Another option seems to be based on the concept of WHAT: Compare the following examples from Arabic:

	mâ         katab-tu                 risâlat-a-n
	not:perf  write:perf.1sg:perf  letter-acc-ndef
	'I did not write a letter.'

	mâ     huwa  sabab-u        sm-i-hi                        l-gharîb-i ?
	what  he/it    reason-nom  name-gen-3sg:poss:m  art-strange-gen
	'What's the reason for its strange name?' 

	Let's assume that the two mâ's are synonymous (which is from being clear from a diachronic point of view): Can we claim that the negation particle mâ stems from WHAT? Maybe, that here, the story goes the other way round: If we interpret *mâ as some kind of 'it is NOT that', the concept WHAT may have emerged from something like *'isn't(?)' [a tag]. Hence, the second phrase would read: *'Isn't (there) a reason for its strange name' > '[yes, there is], ......'. I know, this analysis is more a guess than anything else; it goes against the standard assumption that terms for WHAT often are derived from indefinite (dummy) nouns ('thing' etc.), or from deictic terms marked for interrogation. Still, at least for Arabic, none of these two grammaticalizations paths holds. A superficially parallel type is given e.g. in Udi (and Eastern Caucasian language (which couriously makes use of an element ma, too): 

	s^uk'al-ax      yaq'-al       ma     tad-a-nan        salam
	anyone-dat2  way-super  proh  give-mod-2pl   greeting
	'Do not greet anyone on the road!'

	ma-q'un     lax-e        s^o-t'-ux?
	where-3pl  lay-perf   he-sa-dat2
	'Where did they lay him down?'

	However, note that in Udi, things are more complicated because the 'negative' ma is used with prohibitives only (which reminds us of the Indoeuropean prohibitive *mê ). But whereas the prohibitve base is nicely documented in a number of sister languages of Udi, Udi ma = where does not have convincing cognates. Hence, we *may* assume that it reflects a concept 'is is/should be NOT [there]' taken from the term now used to encode the prohibitive.

	Finally, what to do with polar question (sentence focus)? Naturally, the integration of negated tag-question is a very common option. Still, this does not help to illuminate the status of 'negation' itself, because the 'negation' is already present in the tag. A typical example is German (others have given much better examples):

	Geht sie [nicht] in die Stadt?
	'Does[n't] she go to town?'

	Sie geht in die Stadt, nicht [wahr]'?
	'She goes to town, doesn't she?'

	The description of the corresponding cognitive strategies heavily depends from the syntactic and semantic 'nature' of the tag construction. Nevertheless, it can be hypothesized that here, it is not the negation itself that conditions the 'question' construction, but the fact that a *tag* is present. This can be seen again from German:

	Sie geht in die Stadt, ja?
	'She goes to town, 'yes'?'

	It's simply a matter of conventionalization which type (emphatic assertion, emphatic negation) is 'selected'. Hence, it is (in my eyes) difficult to assume that the negation itself grammaticalizes as a question marker. Rather, we have to deal with some kind of piggybacking: In a tag, negative as well as assertive constructions become  processed as 'question markers' just *because* they are embedded into a (often intonational) pattern of question marking (or: the tag itself is a question). I assume that if such a negative construction or parts of it are tranferred into the 'matrix' clause, they also add their intonational etc. pattern to this clause. In other words: The grammaticalization of negative segments of a tag as Q-markers in polar questions is a secondary effect, not the primary grammaticalization path.

	Best,
	Wolfgang

	--
	Prof. Dr. Wolfgang Schulze
	Institut für Allgemeine und Typologische Sprachwissenschaft
	Department 'Kommunikation und Sprachen' (Dep. II) - F 13/14
	Ludwig-Maximilians-Universität München
	Geschwister-Scholl-Platz 1
	D-80539 München
	Tel.: ++49(0)89-2180-2486 (Sekr.) / -5343 (Büro)
	Fax: ++49(0)89-2180-5345
	Email: W.Schulze at lrz.uni-muenchen.de
	Web: http://www.ats.lmu.de/wschulze.html
	New Version: http://www.ats.lmu.de/index.php

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20050316/af7483b9/attachment.htm>