Transcript KAMI - Wiki du CREDO

‘Reconstructing kinship in Australia: the role
of semantic change and system change
constraints’
Patrick McConvell
AIATSIS/ANU
THE SYMBIOSIS OF LINGUISTICS AND
ANTHROPOLOGY IN KINSHIP STUDIES
• Hage’s ‘linguistic turn’ in diachronic kinship studies - Clark’s
proto-Polynesian reconstructions showed developments to
be different to what anthropologists had proposed
• ‘Comparative linguistic evidence is obviously crucial for an
evaluation of Allen’s theory or for similar theories of
irreversibility in the evolution of kinship systems.’ (Hage
2001)
• ‘The separation of [anthropology and comparative
linguistics] has been …’much to the detriment of progress
with diachronic issues in ethnology’ (Hage 2001 citing Blust
1993).
PAMA-NYUNGAN KINSHIP
• Pama-Nyungan (Pny) is a language family which covers most of
the Australian continent except for the central northern tropics and
Tasmania.
• Amid much other evidence for Pama-Nyungan as a family is the
existence of ancestral kinship roots and suffixes reflexes of which
are found dotted all over the Pny area but only in one adjacent
Non-Pny area, where the forms are borrowed
• An example is the term * kami ‘mother’s mother’ with a distribution
of reflexes illustrated on the next slide
• The complex form *kami(ny)-jarr ‘woman’s daughter’s child’
(reciprocal of MM) is also reflected in various Pny languages
across a broad area. Because of its discontinuous distribution it is
highly unlikely to have been spread by diffusion.
pPNy *kami ‘mother’s mother’
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
*kami-ny-jarr ‘grandchild,mostly wDC’ <‘maternal
grandkin/grandchild dyad’
AUSTKIN DATABASE
http://austkin.pacific-credo.fr
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
*KAMI REFLEXES HAVE A NUMBER OF MEANINGS
INCLUDING ‘FATHER’S MOTHER’. WHY DO I SAY *KAMI
MEANT ‘MOTHER’S MOTHER’?
•There are more reflexes meaning MM than any other
meaning - not a convincing argument by itself.
•*kami(ny)jarr reflexes mean ‘woman’s daughter’s child’ the
reciprocal of ‘mother’s mother’ in languages without a *kami
reflex (except for Yolngu, which we will look at later).
•There is another root *papi- which is widespread and
reconstructable to pPNy which generally means FM. A
proto-system with two FM’s and no MM is implausible.
•If it is MM it forms a recognised type of type of kinship
system - ‘Kariera’ (more on this later)
•First, let’s look at where it has another meaning - FM - and
why.
MEANINGS OF *KAMI REFLEXES
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
*kami(ny)-jarr ‘grandchild’
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
In Gumbaynggirr, south of Yugambeh, FM is kami and MM is paapany. This
reverses the probable Proto-Pama-Nyungan meanings *kami MM; *papi FM.
The extension of the *papi- reflex to MM might have caused this.
In the Karnic sub-group of languages, a new term for MM comes (which is
found elsewhere) kanyi- and kami switches to FM - the order and causal
relations of these changes cannot be certain at this stage of research.
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
Whether the two changes MM>FM in the Lake Eyre and neighbouring
regions and north-eastern New South Wales are independent or
related has not yet been established.
YOLNGU (North-east Arnhem Land)
gaminyarr ‘man’s daughter’s child’ (reciprocal of FM)
not the same meaning as MM and its reciprocal
‘woman’s daughter’s child’
This is the same change MM>FM as in the Lake Eyre
region and Northern New South Wales.
CHANGE OF *KAMI in three regions via
polysemy
MM=> MM, FM =>FM
TYPES OF KIN EQUATION
• Merger within parallel and within cross
(Kariera)
• Merger of same gender (‘grandmothergrandfather’ )
• Alternate generation equivalence (sibling =
parallel grandparent; cross-cousin = cross
grandparent)
• Consanguineal-affinal (prescriptive)
• Skewing (adjacent generation: mother =
cross-cousin etc)
KARIERA SYSTEMS
QuickTime™ and a
TIFF (LZW) decompressor
are needed to see this picture.
The above are Kariera systems that do not
differentiate gender in the grandparent generations
but other systems do eg
Yolngu ngathi MF; FMB; momu FM; MFZ
Ayapathu
(Rigsby)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
kami
kamindhinhu
ngathi
ngathindhinhu
piipi
piinhayi
ngayunpa (poko)
paapa
kaali
thowe
muki
mukithu
wunhayi
karrki
yapi
RED=HAVE
YOLNGU
COGNATES
THOMSON (1972)
Kariera
Prescriptive
equations
Omaha
skewing
KINSHIP AND MARRIAGE SYSTEMS IN CAPE YORK
PENINSULA AND N.E.ARNHEM LAND
10
ARNHEM
LAND
3
GULF O F
CARPENTARIA
1
Ayapathu
7
11
4
CAPE YORK
PENINSULA
9
5
6
7
2
8
NORTHERN
TERRITORY
QUEENSLAND
MAN
MARRIES
FZDD
MBD
ΤKaradjeriΥ
MyBD
FZD
MBD/FZD
ΤKarieraΥ
A hypothesis about proto-Pama-Nyungan
grandparent terms
• *kami- MM;
• *papi- FM;
• *ngaji- MF
• (the root for FF remains less clear) ?mayi-li ?pula
Is the proto-Pama-Nyungan kinship system
Dravidian/Kariera?
Cross and parallel are distinguished. If in addition
there was a gender distinction, so that the
following equations existed it probably was:
• *kami- MM; FFZ
• *papi- FM; MFZ
• *ngaji- MF; FMB
• (the root for FF remains less clear) ?mayi-li ?pula
CONCLUSIONS
• A number of kinship terms can be reconstructed to high-level
subgroupings in Australia including to proto-Pama-Nyungan
• This provides additional support for the existence of Pama-Nyungan
• The terms often vary in meaning across the continent so reconstruction
of meaning can be an exacting task
• However the systems one can plausibly reconstruct are highly
constrained by what we know of possible systems world-wide and in
Australia
• Further, change in meaning is also constrained by the principle that
generally, change from meaning A to B proceeds via a stage of
polysemy A/B and the limited number of possible polysemies that exist
in kinship systems eg FM=MM or FM=MF are common but M=F rare
• In some cases, change in the meaning of a kinship term may be part of
a bundle of shared innovations that defines a sub-group
• However because the polysemies involved naturally occur
independently, this in not always the case