Change in kinship and marriage systems and its reflection in languages and genes

1
Change in kinship and marriage systems and its
reflection in languages and genes

• Patrick McConvell
• AIATSIS/ANU

2
AUSTKIN PROJECT http//austkin.pacific-credo.fr
Australian Research Council 2008-10. Harold
Koch, Ian Keen(ANU) Laurent Dousset (CREDO/CNRS)
et al incl. McConvell (ANU)
3
AUSTKIN ON-LINE DATABASE
4
DYNAMICS OF HUNTER-GATHERER LANGUAGE CHANGE
PROJECTNATIONAL SCIENCE FOUNDATION, USA 2008-11
Claire Bowern (Yale) Jane Hill (Arizona) Pattie
Epps (Texas Austin) Keith Hunley (New Mexico) et
al. incl. McConvell (ANU) website not yet
established

• How the languages of hunter-gatherer groups have
  changed and spread
• Language is a basic element in human social
  identity and change and spread of languages is a
  key aspect of human and social dynamics.
• The spread of farming and languages associated
  with it has claimed a lot of recent attention but
  very little corresponding work on the languages
  of hunter-gatherers.
• What drives change and spread among
  hunter-gatherers? Are such processes are
  fundamentally the same or different from what
  occurred after the farming revolution a
  relatively recent event in the history of humans.
• Comparison of hunter-gatherers in North and South
  America and Australia
• The project is interdisciplinary in using the
  results of biological anthropology and genetics
  in conjunction with those of linguistics, to
  clarify what the relative contributions of
  migration and language shift were to language
  spreads.
• Among the linguistic data to be collected are
  vocabulary in the fields of plants and animals,
  and kinship and social organization. These
  provide evidence of changes in human ecology and
  social patterns respectively and relate to the
  disciplines of archeology and socio-cultural
  anthropology, also represented in the project
  team, as well as to biological anthropology since
  both nutrition and lifestyle, and marriage
  patterns can be reflected in genetic
  trajectories.

5
HUNTER-GATHERER LANGUAGE CHANGE CASE STUDY AREAS
NORTH AND SOUTH AMERICA
6
HUNTER-GATHERER LANGUAGE CHANGE CASE STUDY AREA
AUSTRALIA
7
Resurgence of interest in kinship and kinship
change The last decade has seen a resurgence of
interest in long-term patterns of change in
kinship systems, with reworking of structuralist
ideas of transformations which do not always
lead to actual historical hypotheses (Godelier,
Trautmann and Tjon Sie Fat eds.1998). There are
also actual hypotheses about prehistorical
change emerging (or reemerging), both regional
sequences and general evolutionary patterns and
constraints (eg Dziebel 2007 James et al eds
2008). Linguistic evidence of kinship change A
prominent theme in some of this work is the
importance of linguistic evidence (emphasised in
many papers by the late Per Hage and colleagues).
Reconstruction of kinship terminologies (systems)
at various proto-language levels enables us to
view prehistoric systems and the changes they
have undergone.
8
KINSHIP AND MARRIAGE

• Relationships between marriage types and kinship
  systems have been proposed throughout the history
  of anthropology from Morgan on. These have often
  been framed in terms of correlations between
  marriage patterns or rules and kinship systems
  synchronically.

LINKED CHANGE IN KINSHIP AND MARRIAGE
There are also studies which attempt to show how
diachronic developments in marriage might have
led to change in kinship systems. Dole (1969) for
instance argues that a change to generational
(Hawaiian) terminology from Dravidian/Iroquois
cross-cousin terminology among the Kuikuru of
Brazil represents an adaptation to language group
endogamy due to a long-distance migration.
9
Dravidian endogamous circulation
Dravidian systems would generally tend to keep
marriage alliances in relatively tight closed
loops.
10
SYSTEMS THAT DISPERSE ALLIANCES

• The other kinds of marriage system which are
  known to develop from Dravidian bilateral
  cross-cousin marriage have terminological
  structures distinctively different from
  Dravidian. Some also tend to disperse marriage
  alliances more widely in various ways. These
  include
• Asymmetric cross-cousin marriage eg matrilateral,
  where a man marries a (classificatory) MBD, not
  an FZD
• Skewing systems eg Omaha where an MBD is
  classified as a mother and therefore usually
  unmarriageable
• Second cousin (Aranda) systems in Australia
• (possibly) systems involving marriage classes
  like sections and subsections
• In Australia, the development of Omaha skewing
  can be shown by linguistic evidence to be
  involved in the transition to matriliateral
  marriage. Keen also hypothesises that
  matrilateral systems are correlated with high
  polygyny, and old men having young wives.

11
LOSS OF CROSS-PARALLEL DISTINCTIONS
As well as the loss of symmetry in marriage
direction, a common feature in both Australia and
North America is the loss of an original system
which distinguishes between cross-cousins and
parallel cousins (the latter equated with
siblings) in favour of one which suppresses or
downgrades this distinction. This seems to
correlate with expansion of language groups into
tougher environments on both continents. Lower
levels of polygyny might be expected in this
situation. The loss of symmetry seems to
correlate with what I have called downstream
spread and loss of cross-parallel distinctions
with upstream spread. Details of why these
correlations are present still need to be worked
out. Cross-parallel distinctions in grandparent
terminology are also lost under apparently
similar circumstances yielding
grandfather/grandmother systems from systems
which distinguished FF and MF and FM from MM, eg
in the Chiracahua variety of Apachean and inland
Northern Athapaskan and in the Luritja system
of the Australian Western Desert.
12
MIGRATION AND LANGUAGE SHIFT
The individual migration of spouses to
post-marital residence localities is different
from the migration of whole groups. It may be
better to use movement of spouses for the
former. Migration of groups can happen with or
without significant language shift. Groups may
just live side by side without language shift or
one group may displace another physically without
language shift. Language shift if it occurs can
be from or to the migrating language. It may
occur in conjunction with high rates of
intermarriage or not. A likely hypothesis is that
this conjunction is common. If intermarriage and
language shift cooccur, language shift will
accompany high gene flow and this may be
sex-asymmetric.
13
GENETIC PREDICTIONS ABOUT MIGRATION AND LANGUAGE
SHIFT
Language spread can occur through large-scale
migrations or through language shift. In the
simplest migration scenario, marital exchange
occurs between local groups subsequent to the
migration. The exchange will produce a
correlation between the genetic and geographic
distances between groups. The migration scenario
can be tested by collecting genetic data from the
different groups and measuring the correlation
between genetic and geographic distances. The
hypothesis of language shift may be tested by
collecting genetic data from many groups in a
region that speak both similar and different
languages. Groups that experienced language shift
will be genetically closer to groups with which
they share a more recently common biological
ancestry than to groups with which they share a
language. The hypothesis has previously been
formally tested in South America using
multidimensional scaling (Cabana et al. 2006).
The predictions and tests of the migration and
shift processes are fairly clear cut if groups
have persisted in the same region for
considerable periods, if they have not moved much
within the region, if they are evenly distributed
across the landscape, and if genetic exchange is
limited to geographic neighbors. These conditions
are never met for long in humans, but the tests
can be adjusted to take into account aspects on
population history estimated from independent
sources (e.g., archaeological data) and by taking
into account geographic features such as
waterways.
14
DOWNSTREAM Greater carrying capacity
Initial upstream spread
100 migration
UPSTREAM Lesser carrying capacity
15
DOWNSTREAM Greater carrying capacity
CORRIDOR
Range expansion
Migration
UPSTREAM Lesser carrying capacity
16
DOWNSTREAM Greater carrying capacity
High mobility Dense networks
CORRIDOR
Language focussing Low contact influence
UPSTREAM Lesser carrying capacity
17
DOWNSTREAM Greater carrying capacity
Downstream spread
CORRIDOR
UPSTREAM Lesser carrying capacity
18
DOWNSTREAM Greater carrying capacity
RAIDS
INTERMARRIAGE
RITUAL TIES
Downstream spread
CORRIDOR
SUBSTRATUM
UPSTREAM Lesser carrying capacity
19

• Genetics and marriage
• Patrilocality and dispersal of mtDNA
• There have been a number of studies of
  mitochondrial and Y-chromosome DNA (female and
  male-linked respectively) which have linked wide
  distributions to preferred post-marital residence
  patterns and resultant migration of spouses. A
  common pattern is the wide dispersal of mtDNA,
  linked to patrilocal residence, and perhaps to
  certain marriage patterns.
• In other cases, the results reflect known
  historical events. For example, colonizations
  consisting primarily of men have resulted in the
  introgression of European Y-chromosomesbut not
  mitochondriainto native populations
• patrilocal groups show more geographic structure
  in their Y-chromosomes, while matrilocal groups
  have more geographically structured mitochondria.
  (Wilkins Marlow 2006290)
• the ethnographic dataindicate that the transition
  to agriculture is associated with an
• increase in patrilocality.We propose a model in
  which male and female migration are similar over
  most of human history, and female-biased
  migration is a recent phenomenon.
• (Wilkins Marlow 2006291)

20
CLAIMS DISPUTED This follows on claims by Marlow
(2004) that hunter-gatherers tend not to have
bilateral descent as well as no strong bias in
post marital residence in contrast to
agriculturalists. These hypotheses are doubtful.
Wilkins Marlow assume that there has been a
broad change towards more patrilocality over the
course of (pre-) history due to the change to
agriculture. However when we look at
hunter-gatherer groups we can see not only that
there is variation in these patterns between them
but also that they go through processes of change
which affect marriage dispersal patterns and
therefore distibution of genetic markers.
21
The connections of kinship systems with genetic
variation are also worth pursuing, particularly
because different kinship systems may have
different consequences for patrilineally and
matrilineally transmitted genes the
interaction between kinship as a social
institution and population processes like
migration and diffusion may be a particularly
rewarding topic for future investigation. For
example, prehistorians commonly argue that demic
expansions are driven by innovations in
subsistence, especially domestication. But which
groups spread and both when and how they did is
sometimes a function not just of material
technology but of social structure. Instead of
kinship systems being passively carried along by
population expansions and diffusion like neutral
genetic poymorphisms,they may play an active role
in these processes,which may in turn feed back to
influence kinship. This article has argued that
demic expansions have been associatedwith the
spread of particular social systems future
research may demonstrate that these social
systems haveplayed some role in causing these
expansions. If so, then cultural anthropologys
long-standing interest in kinship systems and
their structural consequences may contributeto
explaining some of the major events in
prehistory. DOUG JONES (2003) Kinship and Deep
History Exploring Connections between Culture
Areas, Genes, and Languages AMERICAN
ANTHROPOLOGIST 105(3)501514.
22
Jones contd. For Na Dene speakers in North
America, there is substantial agreement on
linguistic relationships and ancestral kinship
systems, For New Guinea and Australia, the
role of demic expansions and genetic and cultural
diffusion in the origin of population clusters,
language families, and culture areasis less
certain, although a provisional case can be made
for parallel transmission. Australia is
particularly interesting Although not widely
known, there is quite suggestive
genetic,linguistic, and archeological evidence
for a fairly recent (from about six k.y.a.) demic
expansion. This obviously may bear on the origin
of the continents distinctive social
organization. cites McConvell Evans etc. on
Pama-Nyungan expansion hypothesis
23
KINSHIP IN NORTH AMERICA
Athapaskan Numic Cross-parallel lost in outer
areas, then lineal (Crow) in origin
area Algonquian Cross-parallel lost in outer
areas, then lineal (Omaha) in further southern
area
24
Suggests that genetics supports a migration
scenario for Numic expansion in the main but
Cabana et al 2008 advocate a more rigorous method
which may affect that finding
25
Proto-Numic Kariera
26
Gene flow across linguistic boundaries in Native
North American populations Keith Hunley and
Jeffrey C. Long PNAS February 1, 2005 vol.
102 no. 5
27
ATHAPASKAN (Hunley Long contd)
a history of pervasive genetic exchange across
linguistic boundaries. The distribution of mtDNA
haplogroups in the Apache and Navajo presents the
clearest example. As shown in Fig. 4, the
distribution of the canonical Native American
mtDNA haplogroups differs markedly between the
far North and the Southwest. Notably,mtDNA
sequences belonging to haplogroup B are not
observed in the northern Na-Dene-attributed
populations, and members of haplogroup C occur
rarely (Fig. 4). By contrast, mtDNA sequences in
Southwestern non-Athabascan speakers are
characterized by the predominance of members of
haplogroups B and C and the absence of members of
haplogroup A. The haplogroup configuration for
non-Athabascan speakers in the Southwest is
exemplified in the present study by the Pima
mtDNA sequences (Fig. 4) The Navajo and Apache
possess many haplogroup A sequences typical of
Northwestern populations with languages
attributed to the Na Dene language family.
However, DNA sequences belonging to haplogroups B
and C are also common in the Navajo and Apache,
and these are most likely due to immigrants from
the local non-Athabascan speaking populations.
the pattern of genetic exchange is not
reciprocal.A-group haplotypes would have appeared
in the Pima sample if they had absorbed a
substantial number of Athabascan-speaking
migrants. The pattern of asymmetrical genetic
exchanges is allthe more interesting given
current mate exchange practices.Today, marriage
practice in both the Western Apache and Navajo is
strongly matrilineal . On this basis, we would
not expect to see the inclusion of female
lineages introduced from the surrounding
non-Athabascan-speaking populations. However,the
practice of matrilineality in these populations
is likely to have begun after the Navajos and
Apaches arrived in the Southwest(31). This
practice makes it likely that the haplogroup B
and CmtDNA sequences carried in the Navajo and
Apache today wereintroduced early in their
experience in the Southwest, and before the
current cultural practices were initiated.
28

• THIS IS PUZZLING
• matrilineal refers to a type of descent not a
  type of marriage
• presumably this refers to a requirement that
  Apacheans need to have a mother from the same
  group to be a member of a matriclan, in effect
  prohibiting men marrying outside the group
• not everyone is convinced that Apachean
  matrilineality is a product of late contact with
  Pueblos Dyen Aberle argue it is old in
  Athapaskan
• matrilocality has been seen to be a response
  unifying a group after migration (Divale 1974)
• BUT IT IS AN INTERESTING TYPE OF MODEL
• it proposes a past cultural change in marriage
  which affects the pattern of gene flow between
  groups
• the present-day gene distribution mainly reflects
  a previous period when gene flow was higher, and
  putatively, a different marriage regime existed
• incidentally - this is not in the article -
  parallel types of argument can be drawn from
  historical linguistics - kinship terminologies
  that reflect old states of affairs that help us
  understand change in social practices
• generally however geneticists do not engage with
  this type of anthropological linguistics

29
Jack Ives and Sally Rice (2006) Correpondences in
Archaeological, Genetic and Linguistic Evidence
for Apachean History UCSB/MPI-EVA Language
Genes Workshop
Human biological data provide unambiguous
evidence for this hypothesis that Apachean
ancestors came from the Subarctic, had a small
founding population, and followed a route
southward that did not take them through the
Great Basin (e.g., Li et al. 2002 Malhi et al.
2003 Smith et al. 2000 . Both mtDNA haplotype A
and ALNaskapi incidences confirm a northern
origin for all Athapaskan populations. Lower
sequence variation for mtDNA haplotype A among
Apachean peoples, as well as the characteristics
of Athabascan Severe Combined Immunodeficiency
Disorder and Athabascan Brainstem Dysgenesis
Syndrome, imply genetic bottlenecking in the
Apachean past. ALNaskapi is absent among Numic
speakers, but mtDNA haplotypes B and C among
Apachean speakers show that gene flow did take
place with their immediate, recent neighbours
(Puebloan peoples for Navajo, Piman peoples for
Apaches).
30
(No Transcript)
31
(No Transcript)
32
SOME POSSIBLE TRANSITIONS IN PAMA-NYUNGAN KINSHIP
SYSTEMS
?original Kariera/Dravidian
Cape York Peninsula
Yolngu
gtasymmetric (matrilateral)
Karajarri
Kariyarra
Arrernte
gtAranda (MMBDD/FFZDD wife)
Western Desert
gtLuritja, weakening of crossness
33
NGUMPIN-YAPA
Ngumpin downstream spread
ARANDIC
THE ARANDA SCARP Tawny-hair
WESTERN DESERT DIALECTS
34
DRAVIDIAN TO MATRILATERAL(Kariera to
Karadjeri)
35
KINSHIP AND MARRIAGE SYSTEMS IN CAPE YORK
PENINSULA AND N.E.ARNHEM LAND
There is a strong genetic connection between
western CYP and Yolngu of NE Arnhem Land (White
1997)
Ayapathu
36
THOMSON (1972)
Ayapathu (Rigsby)
Kariera
Prescriptive equations
Omaha skewing
37
(No Transcript)
38
The Indigenous Australian Marriage
ParadoxSmall-World Dynamics on a Continental
Scale

• Douglas R. White drwhite_at_uci.edu
• Woodrow W. Denham wwdenham_at_gmail.com

39
Data - Problematic but Generally Accepted
White Denham contd

• Ethnographers estimate that the populations of
  Indigenous Australian language groups were
  consistently small, averaging perhaps 500 people
  each.
• Classical models of Indigenous Australian kinship
  systems consistently embody endogamous marriage
  as both a norm and a logical requirement.

40
The Australian Paradox
White Denham contd

• Paleodemographers argue that small reproductively
  closed human populations are doomed due to
  stochastic variations in birth rates and sex
  ratios.
• If both the population estimates and the models
  are right, how did these small closed societies
  avoid extinction and indeed persist in Australia
  for 40,000 years and more?

41
A Counter-Intuitive Approach
White Denham contd

• We weaken the axiom for endogamy simply to a
  preference, one that might vary through time.
• We argue that widespread restrictions on
  marriages, especially when mates are scarce, may
  reduce choices locally, but facilitate
  integration of populations globally by forcing
  people to marry outside their own language
  groups.
• Simply put, local restrictions encourage the
  dispersion of marriages.
• REPRODUCTIVE STRESS theory of changes in
  marriage/kinship systems being developed by White
  Denham

42
CONCLUSIONS

• Different kinship systems are correlated with
  different marriage preferences and prescriptions
• Changes in kinship and marriage can be
  investigated using linguistic reconstruction of
  terminologies and changes of meaning of terms
• Different marriage systems are correlated with
  different distributions of genes
• Some systems like Dravidian bilateral
  cross-cousin marriage tend to limit distribution
  of spouses (and genes) while others disperse them
• The tendency for dispersal may be related to
  phases of spread of peoples and languages
  upstream with endogamy and fission downstream
  with exogamy, and language shift
• The proportion of language shift vs. pure
  migration, and its direction. can be investigated
  using genetics