Transcript Roflf’s linguistic map (1970) Aquitanian substrat in toponyms

Linguistic and genetic
singularities
in the Western Pyrenees:
a review of the latest
proposals about Basque
B. Oyharçabal (CNRS, UMR 5478)
HIPVAL (OHLL)
Aussois, 22/09/2005
1
Linguistics
Anthropology
Presence of a non IE
language spoken in
the Aquitanian-Iberian
region since a remote
period
Basques may be the
most direct descendants
of upper Paleolithics
(Michelena 1964,
Trask 1997)
(Bernard & Ruffié 1976,
Cavalli-Sforza et al 1994)
2
Presentation’s steps
• Basic background on the past of the Basque
language and the Basque population
• Review of the Dene-Caucasian hypothesis
• Review of the Vasconic hypothesis
• Conclusion
3
I. Background
4
What do we know about the past of
the Basque language?
1. Spoken in Western Europe before IndoEuropeanisation occurred
2. What Basque words looked like 2000 years
ago
3. Where Basque was spoken at this time.
5
Basque preceded the spread of IE
languages across Western Europe
Non-IE languages present in Western Europe 2000
years ago (or before) are supposed to have
preceded the spread of Indo-European languages
(except Punic settlements) : Basque, Iberian,
Ligurian, Tartessian, ...
Only indirect evidence, because first evidence of the
presence of Basque is recent (2000 years ago).
6
First mention of Basque language:
J. Caesar’s De Bello Gallico
“Gaul is divided into three parts, one of which is
inhabited by the Belgae, another by the Aquitani,
and the third by those who in their own language
are called Celts, in ours Gauls.
All these differ from each other in language,
customs and laws. The river Garonne separates
the Gauls from the Aquitani.”
Aquitanian = Proto-Basque
(Luchaire, Lafon, Michelena, Gorrotxategi)
7
First Basque words and texts
 Aquitanian anthroponyms and theonyms (1-3 c.)
 First Basque words (11th c. on) (onomastic lexicon)
 Connected texts (16th c.)
8
What Basque words looked like
2000 years ago.
Thanks to:
• Aquitanian forms,
• Latin and Romance borrowings,
• Internal reconstruction,
 It was possible to reconstruct the phonology of
Basque (major achievement of Michelena
1960/1976)
9
The Basque speaking region
2000 years ago
• J. Caesar’s and Strabo’s indications,
• Geographical distribution of Aquitanian names,
 Basque was spoken between the Garonne and
the Ebro, as far as the Central Pyrenees.
However, it is not clear where Basque language’s
western limits were at this time.
10
Histiographical background
regarding genetic studies on Basque
population
• Genetic singularity of the Basque population was
established a long time ago using data linked to
classical polymorphisms, such as:
 Rhesus Negative Blood Group frequency (Boyd
& Boyd 1937, Mourant 1947)
 O-Allele frequency (Bernard & Ruffié 1976)
11
Frequency map of the
Rh Negative Blood Group factor
(based on Mourant et al. 1976, from Sykes & Renfrew 2000)
12
Rohlf’s linguistic map (1935/1970)
• Red line: Western limit
of compact diffusion of
toponyms with -an and
-ac endings (early
Romanisation)
• Orange points:
toponyms with the
Aquitanian -os suffix.
13
Rohlf’s map and isogenic curves
(after Bernard & Ruffié 1976)
Thick lines: isogenic
curves
O-allele frequency: 0,70
(yellow)
O-allele frequency: 0,73
(green)
14
Postglacial population expansion
Late paleolithic
population expansion
as revealed by
mtDNA analysis
(Haplogroup V)
From Torroni & al
1998
15
How linguists have linked their
research work to genetic studies
• Until a few years ago most linguists (contrary to
anthropologists) didn’t try to relate their work on
the history of Basque to population history.
• However, more recently, two proposals have
been made to establish remote relationships for
Basque.
16
More recent linguistic proposals
for remote relationships
• Dene-Caucasian hypothesis (Bengtson,
Ruhlen)
• Vasconic hypothesis (Venneman)
These linguists try to make their proposals
coherent with geogenetic data
17
Main proposals in search of
Basque’s relatives until 1990
• Iberian: Humboldt (1821), Schuchardt
(1908).
• Hamito-semitic: Mukarovsky (1981)
• Pan-Caucasian: Bouda (1951), Lafon
(1968)
18
Why the two major hypotheses
were set aside
Vasco-Iberian
hypothesis
Vasco-Caucasian
hypothesis
abandoned when it became
clear that despite some
similarities Basque was of
no help to understand
Iberian texts, which can be
read since the middle of the
20th century
(contrary to what happens
with Aquitanian names).
abandoned when it was
noticed that the proposed
comparisons had serious
methological flaws,
especially because all the
Causasian languages
(Karvelian + Eastern and
Western Northern
Caucasian languages) were
considered as forming one
family.
19
II. The Dene-Caucasian
hypothesis (DCH)
Bengtson (1991,1995, 2003),
Macro Caucasian (MC):
Basque + Burushaski + ‘Northern Caucasian’
20
Dene-Caucasian macro-family
(from http://ehl.santafe.edu/maps5.htm)
21
DCH and genetic studies:
•No genetic correlation with the DC hypothesis
(Basques are similar to other Europeans).
 Ruhlen (1994) argues that DC is so remote
(prior to Eurasiatic) that genetic relationships
between populations have been lost,
 only linguistic relationships would be
traceable.
22
Appraisal of the DCH
Despite recents efforts by Bengtson (2003) to
establish phonological correspondances with
Proto-Caucasian,
 nothing convincing remains: no specialist
shares Bengtson’s and Ruhlen’s view
regarding Basque’s connection to DC.
23
III. The Vasconic hypothesis
(VH)
T. Venneman (1994,1998, 2003)
 follows the tradition of
substrate analyses
24
Substrate analysis in North-Alpine
Europe
• Non-IE features in IE languages can be
explained by substrate features.
• There were only two language families in
Europe before IE expansion; some features
can be traced because there are direct
descendants of these proto-languages.
25
Non-IE languages in Europe
after last Ice-Age
(Venneman 2003)
Vasconic (VH)
Basque
Old
European
Semitidic
Semitic
Atlantic
(Pictish)
26
The VH and genetic studies showing
post-glacial colonisation of Europe
(Torroni et al 1998, 2001)
Vasconic expansion
would have followed
the repopulation of
Europe by Vasconic
people leaving the
Aquitanian-Iberian
refugium after the last
ice-age
(from Venneman 2003)
27
Linguistic arguments for the VH
• Visegimal counting in some Romance, Celtic and
Germanic languages.
• First syllable accent innovation in Germanic, Celtic
and Italic languages.
• River-names of North-Alpine Europe with Vasconic
roots (is-, ur-, aran-, -alde) and morphems (-a); +
agglutinative morphology.
• Etymologies of non-IE words found in IE
languages.
28
Negative appraisal of the VH
• Analysis of hydronyms criticised on several grounds:
eg. -a (suffixed det.) is lately derived from dem. har-;
• No further confirmation of phonological proposals
made in order to establish Vasconic etymologies.
eg. lat. grandis, bq. handi or haundi> vc.*grandi or *graundi
(Proto-Basque words admitted no initial consonant clusters)
• Unlikely late (medieval) borrowings:
eg. -ing. knife < fr. canif / canibet < bq. ganibet
(von Wartburg: o.frq. *knif > fr. canif / canibet).
(Old Basque nouns had no final plosives)
29
IV. CONCLUSION
30
Negative general balance
• The most recents attempts made to establish
long distance relationships for Basque have
failed,
 Despite the support of genetic studies on
population, disappointing results of
Vennemann’s substrate analysis ...
31
two reasons
- inherent limits of substrate analysis, especially
in the case of remote substitution (very little or
no information at all regarding the substituted
language);
- the alleged linguistic homogeneity of W-Europe
before IE-expansion may be wrong; cf. Trask’s
patchwork pattern, which makes substrate
analyses unavailable (because unrestricted).
32
Moreover,
During the same period, historical studies on
Basque have progressed outside multilateralist
studies (Gomez & Sainz, Hualde, Igartua,
Lakarra, Trask, etc.), which have been of no
benefit at all to Basque historical studies;
 Internal reconstruction is still the best tool to
make our knowledge of the history of Basque
advance.
33
New orientations for interdisciplinary
studies in the W-Pyrenees
As genetic studies are making progress,
the correlation between isolated
languages and population histories can be
studied in a more effective way focussing
on shorter periods of the past.
34
Some questions to be adressed
• Is the genetic profile of Basques and W-Pyreneans the
result of late neolithisation of W-Pyreneas?
• What kind of links did endogamic behaviour of populations
have with language survival? Was there any cultural
anthropological basis (family structure?) to such a behaviour
(outside language itself)?
• If there was some late Basque language spread, did it occur
into non Basque populations, or was it the result of
population expansion?
• Is there any link between tribal limits and Basque dialects?
In this case, how can we explain them?
35
Some recent studies
(Iriondo et al 2003)
36
Work in progress
(CHCB, CNRS, INSERM, Univ. Bordeaux II,
Univ. del País Vasco, Univ. Pompeu Fabra)
Population History and Linguistic Variation
in the Western Pyrenees (HIPVAL):
heterogeneity factors
among Basque populations
First results : 2007
37
Researchers
Anthropologists
B. Arveiler , F. Bauduer, P. Darlu, N. Izagirre
Geneticists
D. Comas, L. Quintana-Murci
Linguists
R. Etxepare, E. Montoya, B. Oyharçabal, J.
Salaberria
38