[PapuanLanguages] 'do/say' verbs

[Andrew Pawley]

It seems that do/say polysemy crops up in many 
language families.  I'll avoid the temptation to 
scour lots of dictionaries just now but I know it 
occurs in

1. Some Polynesian languages, e.g.

Samoan & Tongan fai 'do, make, perform, say, tell'
Maori mea 'do, say'

and in some Australian languages, e.g.

Groote Eylandt yama 'be, do, say'

Jaminjung yu(nggu)- 'say, emit sound, do, be, 
move, become/turn into'  (Eva Schultze-Berndt 
Simple and Complex Verbs in Jaminjung  p349 ff)

However Eva argues tentatively that yu(nggu)- is 
a general performance verb.  It  is not a 
prototypical transitive verb in that it is never 
used to describe events with an affected or even 
a highly individuated second participant.  
yu(nggu)- has a single basic meaning, roughly

'x performs an event E'

but is more precisely defined as

'x internally causes, and gives immediate evidence of, an event E'.

The specific readings of yu(nggu)- are derived 
from other material in the construction.  This 
verb does not form a predicate by itself but 
occurs with a quotation, a cognate object ('word, 
speech'), a coverb of internal motion, e.g. 
'rotate', a covers of general manner 'how, do 
what, do like this'), a coverb of bodily or 
emotional condition, e.g. 'itch', 'angry') or a 
predicate nominal ('big', 'night) -- these 
elements represent the 'performed event'. The 
reading 'say/emit' is associated with quotations 
and cognate objects, 'do' or 'perform' with 
coverbs of manner,  'become' with predicate 
nominals, and so on.

This raises the question of whether a 
single-meaning, construction-based analysis can 
be entertained for the do/say verb in other 
languages. There are indications in other 
summaries that this may be the case, e.g. in 
Claudia's account of Savosavo, where different 
'senses' of tei seem to be in complementary 
distribution.

Andy Pawley





>Hi Stuart,
>
>Savosavo (East Papuan, Solomons) has a verb tei 
>which is used to express 'say', 'be thus', 'do 
>thus', 'want to X' (in this case it follows a 
>verb meaning 'to X'), or, if used with an object 
>suffix, 'to do sth. to so.' (the object is the 
>thing or person sth. is done to).
>
>Best,
>Claudia
>
>
>Stuart Robinson wrote:
>
>>In relation to Alex Francois' post from a few months ago, I wanted to ask
>>about another type of 'colexification'/polysemy--namely, between 'say' and
>>'make/do'. I have found that in Rotokas (East Papuan, Bougainville) there
>>is a single verb stem (pura) which means either 'to say' when it is
>>intransitive or 'to make' or 'to do' when it is transitive. I am wondering
>>how common this sort of things is in the region. Can people steer me to
>>some examples of other languages in the region that also have a single
>>term covering both meaning?
>>
>>Thanks in advance,
>>Stuart Robinson
>>
>
>--
>Claudia Wegener
>Insitut für Sprachwissenschaft
>Karl-Franzens-Universität Graz
>Merangasse 70
>8010 Graz
>Austria
>
>Tel.: +43-316-380-2423
>Fax: +43-316-380-9783
>Email: claudia.wegener at uni-graz.at
>
>_______________________________________________
>PapuanLanguages mailing list
>PapuanLanguages at anu.edu.au
>http://mailman.anu.edu.au/mailman/listinfo/papuanlanguages
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.anu.edu.au/pipermail/papuanlanguages/attachments/20061215/7a7fb11e/attachment.html