A Cosmic Hunt in the Berber sky: A phylogenetic reconstruction of Palaeolithic mythology.

Abstract

As species, myths are evolving entities. Here, we built phylogenetics trees of a mythological family. The results clearly support low horizontal transmissions (borrowings), Palaeolithic diffusions and punctuated evolution. Additionaly, a probable Palaeolithic version of the story has been reconstructed.

Full text

93
Les Cahiers de l’AARS — 16 (2013) : 93-106.
1 – Introduction
Biology and mythology have historical
depth, and we need history to understand
them. Yet how do current myths and species
account for past changes ? What is their ori-
gin ? When and how have they evolved ? Could
we reconstruct their primary state ? How
can we tell the difference between chance
and common ancestry to explain similarities
between two versions or two species ? What is
the extent of vertical inheritance and horizon-
tal borrowings ?
To achieve success on these issues in the
biological field, scientists have developped
a set of methods called phylogenetics. We
have already applied many of these methods
to mythological corpus. Indeed, evolution in
organisms and myths occurs through pro-
gressive changes in heritable traits (genes
/ mythems) over successive transmissions
(Lévi-Strauss 1971 : 603-604). Two taxa are
more closely related when they share a more
recent common ancestor, and this similarity
can be used to reconstruct evolutionary his-
tories. We have had some conclusive results
in mythology (d’Huy 2012 a, b, c, d ; d’Huy
2013a, b).
For instance, we have studied the mytholo-
gical motif of the Cosmic Hunt linked to the
Big Dipper. This motif is peculiar to Northern
and Central Eurasia and to the Americas but
it seems to be lacking nowhere in the planet.
Ordinarily, three stars of the handle of the Big
Dipper are hunters and the dipper itself is an
animal(a deer or a bear) ; Alcor is a dog or a
cooking pot (Berezkin 2005).
Julien d’Huy*
A Cosmic Hunt in the Berber sky :
a phylogenetic reconstruction of a Palaeolithic mythology
Les mythes, comme les espèce, évoluent par descendan-
ce modiée. Nous avons ici construit différents arbres
phylogénétiques pour une même famille de mythes.
Les résultats montrent que les mythes se transmettent
essentiellement de façon verticale,ce qui permet de
remonter à une diffusion paléolithique et reconstruire
une proto-version. De plus, il semblerait que les mythes
évoluent par ponctuations.
As species, myths are evolving entities. Here, we bu-
ilt phylogenetics trees of a mythological family. The
results clearly support low horizontal transmissions
(borrowings), Palaeolithic diffusions and punctuated
evolution. Additionaly, a probable Palaeolithic ver-
sion of the story has been reconstructed.
Fig. 1. Reconstruc-
ted phylogeny of
the Cosmic Hunt’s
versions linked to
the Big Dipper. We
have analysed 19
versions of these
tales using bio
Neighbor Joining
(10.000 boots-
trap replications,
d’Huy 2012d).
* Doctorant au CEMAf (Centre d’Étude des Mondes africains), UMR 8171 CNRS. dhuy.julien@yahoo.fr
Received Oct. 15th 2012.

94
Julien d’Huy
We have analysed 19 versions of these tales
using computational methods from evolutionary
biology (d’Huy 2012d). An analysis with Mes-
quite 2.75 (an open-source software for evolu-
tionary biology) produced a consistency index
(CI) of 0.59 and a retention index (RI) of 0.71
that are indicative of low horizontal transmission
and vertical signal in the data. The CI and the
RI can take values from zero to one. High values
indicate a low degree of homoplasy (acquisition
of the same biological trait in unrelated lineages
that is the result of convergent evolution ) for the
CI and a high degree of synapomorphy (traits
that is shared by two or more taxa and their most
recent common ancestor) for the RI. Using both
a bio neighbor joining and a parsimony analysis,
we obtained a clear organisation of the versions
and a progression : Greece § Central Asia and
Siberia  British Columbia  Northeastern
America. (Fig. 1 for the bio neighbor analysis
with SplitsTree4, a popular program for inferring
phylogenetic trees or phylogenetic networks).
This signal showed that the similarities sys-
tematically decrease among the versions as the
geographical distance increases (for a similar
process in language, see: Holman and al. 2007).
It was consistent with our knowledge of the first
human migrations. The Big Dipper interpreted as
a Cosmic Hunt must be even older than 15.000
years, coming from Asia, when North America
was populated by migrations across the Bering
Strait. So, after two versions diverge from a com-
mon ancestor, they also become less similar to
each other with the passage of time. Note that the
use of mythemes (irreducible and unchanging ele-
ment of the myth that can be similar or different
between two versions) allow us to go back very
far in time, beyond the horizon of current phyloge-
netic method in linguistics. It may be because all
mythological versions share a bigger pool of com-
mon structures and units than language.
Phylogenetic reconstructions using parsi-
mony and maximum likelihood can attempt
to reconstruct the ancestral states of each
mytheme and of the myth itself ; we also
reconstructed the following Palaeolithic story :
a man pursues a deer, and the animal is alive
when it turns into the whole Dipper.
The Bayesian approach is considered to be
the best current method of phylogenetic recons-
truction. Yet the trouble with this method is
that human populations and mythologies do
not really take the form of a tree, but rather the
form of a bush, with exchange of both genes
and mythemes. Mythologies are not detached
from each other, and they continue to interact
after their differentiation. This interaction is
facilitated by linguistic proximity that allows
communication and by geographic proximity
that allows talks, borrowings and learnings
between speakers of different languages.
2 – Material
For this paper, we developed our database,
including mythological versions of Cosmic Hunt
centred on Big Dipper but also other versions
centred on Orion, Pleiades, etc. We added Tuareg
(Duveyrier 1864 : 424 ; Basset 1910 : 16-17 ; Ben-
hazera, 1908, 61), Wagogo (Nilsson 1920 : 119),
Khoikhoi (Nilsson 1920 : 120), Karanga, Tswana
(Koekemoer 2008: 75) Rutul (Bulatova 2003 :
222, apud Berezkin 2009 : 41), Basque (Cerquand
1875: 19-20; Vinson 1883 : 8-9), Saami (Billson
1918: 180; Berezkin 2005: 87 ; Pentikänen 2011 :
107), Evenky (Anisimov 1958 : 12-13; Jacobson
1993), Chukchi (Bogoras 1924 : 243), Ojibwe
(Speck 1915 : 63-64; Williams 1956 : 27-30),
Copper (Rasmussen 1932 : 23), Netsilik (Ras-
mussen 1931 : 211, 385), Polar Eskimo (Holtved
1951 : 50-55), Baffin land (Boas 1888 : 636-637),
Igloolik (Kroeber 1899 : 173) Coeur d’Alène (Teit
and Boas 1930 : 178-179), Tsilhqot’in (also called
the Chilcotin; Farrand 1900 : 31), Mocovi (Rivera
de Bianchi 1973 : 704), Kalina (Magana 1983),
Akawaio (Brett 1880 : 191-200; Roth 1915 : 265-
266), Chemehuevi (Fowler 1995: 147-148), Mari-
copa (Spier 1933: 146-147), Kiliwa (Meigs 1939 :
69-78) and Mojave (Fowler 1995: 147) versions.
For the others references, see d’Huy 2012d.
What about the Saharan versions of Cosmic Hunt ?
Tuareg know Ursa Minor and Ursa Major as a young camel and its mo-
ther (Tuareg 1). The North Star is sometimes a black woman that holds
the reins of the young camel as its mother is milked (Duveyrier 1864 :
424) or the post to which the young camel is tied (Bernus & Sidiyene
1989 : 155) or the head of the mother (Benhazera 1908 : 61 ; Stefanini
1926 : 127). When the North Star is a black woman, she believes that
the stars ψ, λ, μ, ν, ξ want to kill her, so she stays still (Duveyrier 1864 :
424 ; Pottier 1946 : 244-245). It’s easy to recognize here the Cosmic
Hunt motif with a move from the animal hunted to the owner hunted.
The story also seems to be inuenced by the Arab culture. Indeed, the
Arabs saw a cofn and mourners in the Big Dipper. The four stars of
the dipper’s cup are the cofn ; the handle are the three mourners, sons,
daughters or brothers of the deceased. They are following the North Star
who killed their father and seeking vengeance.
In a second Tuareg version (Tuareg 2), Ursa Major is a female camel
that belonged to Noah ; the animal was killed by seven noble people,
including one Tuareg ; the Tuareg was transformed into a desert monitor
(ăɣāta, Varanus griseus) and other people into jackal, chameleon... Sin-
ce then, the Tuareg do not eat the desert monitor whom they considered
as their maternal uncle (Basset 1910 : 16-17 ; Benhazera, 1908 : 61). This
version is clearly inuenced by the Arab culture (cf. Noah) and might be
the origin of the Greek versions, where a god places an animal amongst
the stars. Note that an inuence of Berber culture on Greek culture has
been previously showed for the Pygmalion motif (d’Huy 2011).
These tales are not cosmic hunt sensu stricto but rather deformed ver-
sions of the primitive story.

A Cosmic Hunt in the Berber Sky : A Phylogenetic Reconstruction
95
3 – Method and results
The questionnaire previously used in d’Huy
2012d was developed on the basis of the known
typological variations in this family of myths
(see the table at the end of the paper). The
absence or presence of a mytheme was coded
for each version by 0 or 1, respectively, to pro-
duce a binary matrix of 93 mythemes in 47
versions. The database was still incomplete but
has been build so that it could flexibly grow as
new versions emerge. This will allow the abi-
lity to test our model.
We took into account adaptation to the
local environment, because the choice of the
same mytheme (e.g. tapir, fisher) by neighbo-
ring cultures probably showed vertical trans-
mission from mother to daugther mythologies.
To reconstruct the first state for the myth,
we used Mesquite 2.75 (Maddison & Maddison
2011). We calculated the parsimony treelength
of the tree and matrix. Character matrices
were supplied from data files and the tree was
rearranged by subtree pruning and regrafting.
We built a majority rules consensus tree where
only those clades present in more than 50 per
cent of the trees were present (Fig. 2 ; 200
trees ; treelength : 213). The retention index
for matrix (0,6966) indicated that most of the
mythemes were shared through common des-
cent. Indeed, high CI and RI values (for ins-
tance, greater than 0.60) usually show a low
horizontal transmission (Nunn and al. 2010).
Arguments in favour of localization of pro-
typical Cosmic Hunt in Asia seem persuasive
(Berezkin 2005). We successively rooted the
tree with each eurasian version, one by one.
Phylogenetic reconstructions using maximum
likelihood allowed to reconstruct ancestral
states of mythemes and of the protomyth itself
for each root. We then identified the recons-
tructed mythemes for each eurasian root
(Fig. 3) and selected the version that holds the
majority of the wide shared mythemes (>50%)
as the better root. Consequently, the tree has
been rooted with Evenki 3. The reconstructed
Palaeolithic version was almost identical to the
previous one (see below).
The second mythological tree was
constructed with MrBayes 3.2.1 (Huelsenbeck
& Ronquist 2001; Ronquist & Huelsenbeck
2003). We set the evolutionary model to the
generalised time reversible substitution model
with gamma-distributed rate variation across
sites and a proportion of invariable sites. We
ran an ordinary Markov Chain Monte Carlo
(MCMC) analysis for 40.000.000 generations
with a sample on the screen every 10.000 gene-
rations. The summaries were based on a total
of 8.002 trees from two runs. Each run pro-
duced 4.001 trees of which 3.001 samples were
included. The fact that stationary has been
reached (burn-in) was controled with Tra cer
1.5.0 (Drummond & Rambaut 2007). As a pre-
caution, the 2.500.000 first steps (250 samples)
were discarded as “burn-in”. According to the
previous analysis, the tree was rooted with
Evenky people 3 version.
We simulated evolution of DNA sequences
on the bayesian tree with Mesquite 2.75 and
chose the Jukes-Cantor model as a user-defi-
ned model of nucleotide evolution. The four
submodels of this one were i/ the states at the
root of the tree, ii/ the equilibrium frequencies
of states on other branches, iii/ the relative
rates of characters, and iv/ the relative rates of
Fig. 2. Mesquite tree
including mytho-
logical versions
of Cosmic Hunt
centred on Big Dip-
per but also other
versions centred on
Orion, Pleiades, etc.
(red : Asia ; gray :
Greece ; yellow :
Basque ; orange :
Africa ; l ight blue :
Arctic ; magenta :
american Coast-
Plateau / British
Columbia ; pink :
northeastern Ame-
rica ; wood : ame-
rican Great Basin
/ Great Southwest ;
green : Guia na)

96
Julien d’Huy
change from state to state. We simulated 1000
characters and obtained a consistency index of
0,7358 and a retention index of 0,8548. These
results were greater than the results for the pre-
vious tree.
Using MrBayes, we also run an MCMC
analysis with the standard non-clock model
for 15.000.000 generations. Then we repeated
the procedure under the strict clock model. We
used the stepping-stone method and we estima-
ted the model likelihood by sampling a series
of distributions that represent different mixtures
of the posterior distribution and the prior dis-
tribution (Xie and al. 2011). The sampling was
based on 50 steps with 294.000 generations
(588 samples) within each step. The marginal
likelihood was bigger for the strict-clock model
than for the non-clock model, the log values
being -1057,04 and -1092,66, respectively (Fig.
4). There was good agreement between the two
independent analyses for both models, indica-
ting that we had accurate estimates of the mar-
ginal model likelihoods. The strict-clock model
Fig. 3. Recons-
tructed mythemes
for each eurasian
root by maxi-
mum likelihood
method. Claude
Lévi-Strauss calls
zoemes “animals
given seman-
tic functions.”
Zoemes “allow
mythic thought to
keep operations
within the same
framework”
(Levi-Strauss
1985 : 130).
A : Khanty
B : Saami1
C : Evenki 1
D : Evenki 2
E : Evenki 3
F : Ob-Ugric
G : Saami2
H : Chukchi
I : Rutul
J : Basque 1
K : Basque 2
L : Pausanias
M : Pseudo-
Apollodorus
N : Hesiod
O : Ovid
Last column : % of
shared mythemes.

A Cosmic Hunt in the Berber Sky : A Phylogenetic Reconstruction
97
was 35,62 log likelihood units better than the
non-clock model, a log difference above 5 being
considered very strong evidence (Kass & Raf-
tery, 1995). Thus, the marginal likelihood mean
estimator indicated that we had strong evidence
in favor of strict-clock model. In fact, myths
change over time in a more complex way.
Punctuated equilibrium is a theory in
evolutionary biology which proposes that,
instead of a slow, continuous and gradual
movement, evolution tends to take the form
of extended periods of time where species
will exhibit little net evolutionary change
(stasis), «punctuated» by episodes of very
fast events of branching speciation and
development of new forms. Punctuated and
gradual evolution result in different rela-
tionships : there is a positive correlation for
a punctuational process between path length
(the sum of length from each terminal node
to the root) and the number of nodes on the
path. In contrast, for gradual process, the
path lengths are independent to the number
of divergence events along that path (Webs-
ter and al. 2003). Using the Bayesian tree,
we computed the Pearson product-moment
correlation coefficient r (or Pearson’s r)
between the number of nodes on each path
and each root-to-tips path length. The tree
with branches proportional to lenghts (Fig.
5) showed a strong correlation (0,9146, 1
being a perfect positive correlation) between
the number of net-speciation events (nodes)
and the total path length from the root of the
tree to its tips (Fig. 6).
The node-density effect is an artifact
of phylogeny reconstruction that can cause
shorter summed lengths in areas of the tree
where fewer species (here, collected myths)
have been sampled. The artifact produces a
positive relationship between path lengths
and numbers of nodes and could impact the
results of a punctuational effect. However,
the artifact contains a distinctive signature
in the form of a curvilinear relationship
between path length and number of nodes
(Venditti and al. 2006). This curvature can
be reliably detected by means of the coeffi-
cient of determination (R2). The R2 s a num-
ber between 0 and 1 used to describe how
well a regression line fits a set of data. An R2
near 1 indicates that a regression line fits the
data well, while an R2 closer to 0 indicates a
regression line that does not fit the data very
Marginal
likehood
Standard non-
clock model
Strict-clock
model
1 -1092,88 -1057,04
2 -1092,47 -1057,03
Total -1092,66 -1057,04
Fig. 4. Margi-
nal likelihood
for standart non
clock-model and
strict-clock model.
Fig. 5. Consensus
tree based on ave-
rage branch lengths.
Labels indicate
the probability
of each node.

98
Julien d’Huy
well. We used OpenOffice 3.4.1 to calculate
the R2 for the figure 6. The R2 with a linear
regression (0,84) was bigger than a logarith-
mic regression (0,79) or than an exponential
regression (0,77). The tree also didn’t seem
to show the curvilinear trend associated
with the node-density artifact. R2 forlinear
regression could represent the percent of
variation explained by punctuational pro-
cesses(Dediu & Levinson 2012).
Then we used another technique for
constructing phylogenetic networks, Neighbor-
Net, implemented in SplitsTree 4.12 (Bryant
& Moulton 2004 ; Bryant and al. 2005), to
visualize the relationships implied by the data
(Fig. 7). We used uncorrected P-distance. This
algorithm takes a distance matrix as input, and
works by agglomerating clusters. NeighborNet
constructs networks rather than hierarchical
trees and can make a useful contribution to
representing multiple phylogenetic hypotheses
simultaneously and showing complex evolu-
tionary processes (e.g. hybridization, lateral
transfer, recombination).
We also used SplitsTree to quantify how
each version fits into the network with the
so-called delta score (Holland and al. 2002).
Briefly described, the method scored indivi-
dual versions from 0 to 1; a relatively high
delta score indicates a strong conflicting
signal in the data. The average delta-score
was 0,3393. This corroborated that the data
was tree-like. Wichmann and al. (2011) cit e
delta scores across the world’s language fami-
lies. The average which can be obtained from
their data is 0,3113. Thus some myths are
at least as tree-like in their behavior as lan-
guages, if not more.
The initial coding procedure might intro-
duces some bias in the tree. One can argue that
it is very difficult to compare the American
zoemes with others from Eurasia or Africa
because they simply do not live in these area.
To assess the robustness of the network, we
removed the zoeme names (mythemes 1-19)
and used only the mythemes 2 and 17: “the
zoeme is an herbivore” and “the zoeme is a car-
Fig. 6. Root-to-tip
path length against
number of nodes
along each path.
Fig .7. A split graph
showing the results
of NeighborNet
analyses of the Cos-
mic Hunt versions.

A Cosmic Hunt in the Berber Sky : A Phylogenetic Reconstruction
99
nivore.” Additionaly, the logically dependent
mythemes 91-92-93 were eliminated. The pro-
blem is that a “yes” on item 57-78 implies a
“yes” on items 91 for instance. This could have
a big effect on the results of our analysis by
exaggerating the appearance of clustering. The
NeighborNet network created with SplitsTree’s
analysis is shown in fig.8 and the tree created
with Mesquite is shown in fig.9.
4 – Discussion
The MrBayes’ consensus tree is shown in
fig.5. The root and the first nodes of the second
branches are eurasian. A branch groups the
Greek and Ojibwa versions. That can be explai-
ned by an ancient genetic link : haplogroup X2
appears to be essentially restricted to northern
Amerindian groups and to the Near East, the
Caucasus and Mediterranean Europe, making
it likely that some Native American founders
have European ancestors (Brown and al. 1998).
There also are mythological links (d’Huy 2012
c, e) between Europe / Western Asia and North
America. Many similar and very ancient tales
are largely restricted to the haplogroup X2 area,
e.g. i/ a person gets into a monster’s homestead,
can be killed, but escapes sticking to the hair
of an animal who is going out or under an ani-
mal’s skin (d’Huy 2012c, 2013b), or ii/ animals
ask riddles that the hero must answer on pain
of death (dHuy 2012e). All of that reinforces
the idea that the peoples with the haplogroup
X were part of the original founders of Native
American populations. Yet the cluster Greek /
Ojibwa may be weakly supported (see below).
Two Salish versions (Snohomish people /
Twana people) are incorrectly placed before the
Basque versions and belong to the same cluster
than Evenki. Note that the larger and contro-
versial family “Dene-Caucasian” includes the
Basque and maybe the Salish language fami-
lies (Shevoroshkin 2003, 2008). Moreover, the
Yeniseian languages that belong to the Dene-
Caucasian family share many contact-induced
similarities with the South Siberian Turkic
languages, Samoyedic languages and Evenki
(Anderson 2003) ; these contacts could explain
the proximity between Evenki, Basque and
Salish versions. If we exclude this hypothesis,
it can be assumed that during the last glacial
maximum, one migration spreads from Asia to
the Pyrenees and the Northwest Coast.
By observing the figure 5, several conclu-
sions can be drawn. North and South Ameri-
can versions may come with many migrations
from Siberia rather than just a single wave of
migrants, one by ancestors of present Eski-
mos and some Northeastern Amerindians,
one by ancestor of Algonquian (who sha-
red haplogroup X with European), another
Fig. 8. Splitgraph
without some
mythemes.

100
Julien d’Huy
by people whose descendants are confined
to the Southwest of North America and the
South America, and another one by ancestors
of many Salishian people. These migrations
probably occurred at least 15.000 years ago,
before the rising sea water submerged Berin-
gia. The fact that African and Great Basin
versions belong to the same cluster can be
explained by a common eurasian origin. The
African versions probably indicate a reverse
migration from Asia to Africa.
Our results also suggest that, as for spe-
cies and language (Atkinson and al. 2008;
Dediu and al. 2012), newly formed versions
diverge at a rapid pace around myths splits,
and then adopt a slower rate of change, where
they evolve at a rate relatively constant over
time (see also: d’Huy 2013b). The punctuatio-
nal bursts may account for about 84% of the
total divergence among the studied versions. It
could be due to a cultural need to foster group
identity and social cohesion (we think that ver-
sus they think that) : for instance, a story of
the origin of fire was told by an amerindian to
offset another story by an amerindian of ano-
ther tribe (Goddard 1904 : 197). The second
hypothesis is a mythological founder effect.
In a similar effect to its genetic counterpart,
a mythological founder mutation appears in a
myth known by individuals who are founders
of a distinct population; then this mutation can
get passed down to other generations. Here,
the myth is probably characterized by serial
founder effects as people settled Eurasia and
North America by rapid and long distances fol-
lowed by periods of settlement. Consequently,
mythological differentiation tends to increase
with geographic distance. Future works will
hopefully give a better measure of this punc-
tuational effect and of the role of distance in
mythological differentiation.
The SplitsTree’s network shows a non star-
like aspect according to a deeply stable signal
(Fig. 7). This network highlights higher level
cluster in the data with an African, an Eura-
sian and an Amerindian groups. Additionally,
it correctly organises some of the versions
into well-known linguistic groups, with Ura-
lic (Ob-Ugric, Khanty people, Saami), Saha-
ran (Tuareg 1, Tuareg 2), Greek (Hesiod /
Ovid / Pausanias / Pseudo-Apollodorus),
Tungusic (Evenki 1 and 3), Eskimo (Copper,
Baffin land, Polar Eskimo, Netsilik, Igloo-
lik), Salishan (Snohomish / Twana ; Lillooet
/ Coeur d’Alene / Nlaka-Pamux), Iroquoian
(Mohawk / Cherokee / Seneca, with three
geographically close Algonquian versions :
Mi’kmaq, Lenape and Meskwaki), Algon-
quian (Ojibwa 1 / Ojibwa 2), Yuman (Mari-
copa / Kiliwa / Mojave ; Chemehuevi is Uto-
Aztecan) and Carib (Kalina / Akawaio) area
being the most distinct. The low number of
boxes showing conflicting signals and the
coherent groupings of mythological versions
indicate that the structural features of these
versions were deeply stable through time (for
the same analysis with language: Greenhill
and al. 2010). There is a clear link between
languages and mythological versions. Myths
belong probably to the category of cultural
features that are most similar to languages in
their distribution and the both seem change
relatively little over their history.
According to the haplogroup X’s hypothesis,
the Greek cluster is not placed into the eurasian
cluster. The Khoikhoi version is also incorrectly
Fig. 9. Tree drawn
with an altered
version of the
database, without
some mythemes.

A Cosmic Hunt in the Berber Sky : A Phylogenetic Reconstruction
101
placed. Its place might indicate a quick migra-
tion from Asia to Africa and America. Note that
a specific version can only appear around the
perimeter (and not in the midst) of a network of
reticulations. If borrowings are identified in a
version, NeighborNet cannot place such a ver-
sion in an intermediate position between many
others around it, and that can biase its position
(Heggarty and al. 2010). A cluster groups the
Tuareg, Basque, Saami, Evenki and Uralic ver-
sions, which could be interpreted as indicating
an ancient common ancestry. This grouping
may confirms that the Franco-Cantabrian
refuge area was the source of expansions of
hunter-gatherers that colonized northern Africa
and recolonized much of northern and central
Europe after the last glacial maximum. It also
reveals a common mythological and palaeo-
lithical bond that unites the European hunter-
gatherer populations and the Berbers.
To assess the solidity of the network,
we repeated the NeighborNet analysis with
Splitstree after removing some doubtful
mythemes (see Methods and results). The new
network (Fig. 8; delta-score: 0,3514) is very
similar to the previous one, with approxi-
matively the same linguistic clusters, except
for Saharan. The Eurasian group is divided
into two groups that remain at the root of the
other branches. The Eskimo cluster is splited
between Netsilik / Igloolik / Polar Eskimo and
Copper / Baffin land groups, yet remains at the
base of the Northeastern amerindian branch
(as it actually does with more mythemes; see
fig.7). Yuman and African versions continue
to be closely associated, as Basque / Snoho-
mish / Twana and Greek / Ojibwa versions
do. The Khoikhoi version is always incorrectly
placed. The Wasco version changes position.
With a Mesquite analysis of the modified
database (Fig. 9; consensus tree of 200 trees ;
treelength: 172 ; retention index: 0,6566), the
relationship within each of the main clusters
remains approximatively the same and the
linguistic and cultural groups are relatively
similar. The tree groups the Ojibwa versions
with the Iroquoian tales, which weakens the
haplogroup X’s hypothesis. It also groups one
Tuareg version with Greek versions. These
new clusters could be explained by borrowings
between neighboring people (see the Box for
the Tuareg version). Additionaly, the Iroquoian
family does not form a monophyletic group
and Mojave and Netsilik versions are mis-
placed. To conclude, in agreement with the
two modified data analyses, a deeply change
of mythemes may cause only small changes
in the topology of the tree and network, and
misplacements do not affect their general
meaning. Yet we should note that the results
obtained with the modified data could be less
reliable because there may be not a sufficient
amount of data analysed.
According to the reconstructed Palaeolithic
version (see Methods and results), there is an
animal that is a horned herbivore, especially an
elk. One human pursues this ungulate. The hunt
locates or get to the sky. The animal is alive when
it is transformed into a constellation. It forms the
Big Dipper. With the altered tree (Fig. 9), the only
change is that the zoeme is an herbivore, without
any additional information (according with infor-
mation partially given by the modified data).
The transition of the Ursa Major designa-
tion from elk to bear in Indo-European and
Mediterranean area (not earlier than 2000
B.C.) is corroborated by various cultural and
linguistic sources (Lushnikova 2003). The
reconstructed prehistoric story also explains
why there is no bear representations similar
to the Big Dipper in Upper Palaeolithic (Hay-
den and Villeneuve 2011), the cosmic animal
being a horned herbivore. The Cosmic Hunt
myth is probably reflected in the rock art of
Karelia, Siberia, the Far East and Northern
Mongolia (Ernits 2010) and may be repre-
sented in the famous Lascaux shaft ‘scene’
showing a bird-headed ithyphallic man with
an apparently disembowelled bison (d’Huy
2012d). Yet the contents of the myth and the
specifics of rock art make it difficult to prove
definitively the relevant connection. Additio-
nally, the phylogenetic approach suggests that
the Cosmic myth would have existed in many
area, including the prehistoric Sahara : we
now need to find where and how this myth
has been represented in this area.
5 – Conclusion
Our phylogenetic approach of myths allows
us to : 1/ confirm the Palaeolithic datation of the
Cosmic Hunt linked to a horned herbivore, 2/
reconstruct the first version of the myth, 3/ cor-
roborate the hypothesis of a European post-gla-
cial human recolonization of Europe and a part
of North Africa from the Franco-Cantabrian
refuge, 4/ prove the existence of a mythological
root common to Berbers and European hunter-
gatherers, 6/ document at least four migrations
from Eurasia into America and suggest that
there was reverse migration from Eurasia to
Africa, 5/ find support for the punctuated evo-
lution of myths, 6/ offer some suggestions to
interpret prehistoric rock art images that should
be discovered. Thus, the study of human mytho-
logy directly opens up new ways for genetics,
archaeology and rock art researches.
Acknowledegments :
The author thanks
Yuri Berezkin,
Brian Hayden, Jean-
Loïc Le Quellec,
Steven Pinker,
Yves Vadé and
Søren Wichmann
for their helpful
suggestions and
encouragements.

102
Julien d’Huy
Ch
ero
ke
e
Nl
ak
a'p
am
ux
Tw
an
a
Sn
oh
om
ish
Me
sk
wa
ki
Mi
km
ac
Mo
hw
ak
Kh
ant
y
Ev
en
ki
1
Ev
en
ki
2
Ob
ugr
ic
Ba
sq
ue
1
Sa
am
i1
Tu
are
g 1
Tu
are
g 2
He
sio
d
Ps
eu
do-
Ap
oll
od
oru
s
Ov
id
Pa
us
ani
as
Lill
oo
et
W
as
co
Oji
bw
a
Se
ne
ca
Ch
uk
chi
Co
pp
er
Ne
tsili
k
Pol
ar
Es
ki
mo
Ba
ffin
La
nd
Co
eur
d'A
lèn
e
Tsi
lhq
ot'i
n
Oji
bw
a2
Ba
sq
ue
2
Kal
ina
Ak
aw
aio
Le
na
pe
Mo
jav
e
Igl
ool
ik
Ch
em
eh
ue
vi
Ma
ric
op
a
Kili
wa
Ev
en
ki
3
Kh
oik
hoi
Ka
ran
ga
Ru
tul
Sa
am
i2
Ts
wa
na
Wa
go
go
1
The zoeme is a big mammal.
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
0
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
2
The zoeme is a herbivore.
0
0
1
1
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
1
1
1
0
1
0
1
1
1
1
1
1
1
1
1
1
3
The zoeme is a horned animal.
0
0
1
1
0
0
0
1
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
1
0
0
0
1
0
1
1
1
1
0
0
1
1
0
0
4
The zoeme is an ungulate.
0
0
1
1
0
0
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
1
1
1
0
1
0
1
1
1
1
1
0
1
1
1
1
5
The zoeme is an elk.
0
0
1
1
0
0
0
1
1
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
6
The zoeme is a reindeer.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
7
The zoeme is a deer.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
8
The zoeme is a moose.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
9 The zoeme is a camel. 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
10
The zoeme is an ass.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
11 The zoeme is a mountain sheep. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 1 1 0 0 0 0 0 0 0
12
The zoeme is an antilope.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
13
The zoeme is a zebra.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
14
The zoeme is a pig
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
1
15
The zoeme is an ox.
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
16
The zoeme is a tapir.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
17
The zoeme is a carnivore.
1
1
0
0
1
1
1
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
0
1
1
1
1
1
0
1
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
18
The zoeme is a bear.
1
1
0
0
1
1
1
0
0
0
0
0
0
0
0
1
1
1
1
1
1
0
1
0
0
1
1
1
1
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
19
The zoeme is a fisher.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
20
It's a domestic animal.
0
0
0
0
0
0
0
0
0
0
0
1
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
21
It’s a six legged-animal.
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
22
The zoeme captures the sun
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
23
There is one animal.
1
1
0
0
1
1
1
1
1
0
1
0
1
0
1
1
1
1
1
1
0
0
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
0
0
0
1
0
0
0
1
0
0
24
There is two animals.
0
0
0
0
0
0
0
0
0
1
0
1
0
1
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
1
25
There is three animals.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
0
1
1
0
0
1
0
26
There is four animals.
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
27
There is seven animals.
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
28
An animal is punished for his pride.
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
29
There is one pursuer.
0
0
1
0
0
0
0
1
1
0
1
0
0
0
0
0
1
1
1
0
0
0
0
1
0
0
0
0
0
0
1
0
0
1
0
1
0
0
1
1
1
1
1
0
1
0
1
30
There are two pursuers.
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
31 There are three – or at least three -
pursuers.
1 1 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 0 1 0 0 0 0 1 1 0 1 1 0 0 1 0 0 1 0 0 1 0 0 0 0 0 1 0 1 0
32
There are four pursuers.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
33
There are five pursuers.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
34
There are seven pursuers.
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
35 Pursuers are
members of the same
family.
0 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
36
Pursuers are dogs.
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
1
0
1
0
37
An animal pursues an animal who
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0

A Cosmic Hunt in the Berber Sky : A Phylogenetic Reconstruction
103
pursues an animal
38
Pursuers located or get to the sky.
1
1
1
1
1
0
1
0
0
1
1
1
1
1
0
0
0
0
0
0
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
39
A man goes down alone from the sky to
earth and destroys the way to access to
the sky.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
40
A man turns his brothers into stars.
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
41
A woman breaks a taboo.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
42
Hunters become stars because of a
relative.
0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0
43
A god transforms a nymph into a bear.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
44 A man is about to kill his own mot
her
transformed into a bear.
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
45
A god transforms animal into a
constellation.
0
1
0
0
0
0
0
0
?
0
0
1
0
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
46
A divine person stops a hunter.
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
47
The animal is alive when transformed
into constellation or in the sky.
0
1
1
1
0
0
1
1
1
1
1
1
1
1
0
1
0
1
0
1
1
0
0
1
?
?
1
1
1
0
0
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
48
The animal is dead when transformed
into constellation.
1 0 0 0 1 1 0 0 0 0 0 0 0 0 1 0 1 0 1 0 0 1 1 0 ? ? 0 0 0 1 1 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0
49
The hunt continues until the fall.
1
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
50
The grease or the blood dripping from
the animal's body falls on earth and
becomes something else.
1 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
51-
53
The dripping blood of the animal
tinges the autumn foliage.
0
0
0
0
1
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
The grease becomes honey-dew.
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
The grease becomes snow.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
54
The dog is transformed into a star.
0
1
1
1
1
0
1
0
0
0
0
1
1
0
0
0
0
0
0
1
1
0
0
0
0
1
1
1
1
1
0
1
0
0
1
0
1
0
0
0
0
0
1
1
0
1
1
55
The cosmic animal is
hit with a
dart or arrow.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
1
0
0
0
0
1
0
1
1
1
0
1
0
0
0
0
0
56
The hero is the origin of warmth
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
57
Each animal is transformed into a star
of the Big Dipper.
0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
58
One animal turns into a star of the
Dipper.
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
59
Two animals turns into two stars of the
Dipper.
0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
60
Four animals turn into four stars of the
Dipper.
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
61
Seven animals form seven star of the
Dipper.
0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
62
The animal forms four stars of the
Dipper.
1
1
0
0
1
1
1
0
1
1
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
63
The animal forms the Dipper.
0
0
0
0
0
0
0
1
1
0
0
0
1
1
1
1
1
1
1
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
64
Members of the same family turn into
Ursus Major and Ursus Minor.
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
65
Cut off limbs are stars seen in winter.
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
66
Three stars are the shadow of the
animal.
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

104
Julien d’Huy
67
Hunters are the handle of the Dipper.
1
1
0
1
1
1
1
0
0
1
0
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
68
Hunters form four stars of the Dipper.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
69
Hunters form five stars of the Dipper.
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
70
Hunters form seven stars of the Dipper.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
71
Animals are associated to their owner.
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
72
Alcor is a dog.
0
1
1
1
1
0
1
0
0
0
0
1
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
73
Alcor is an object.
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
74
Alcor is a cooking pot.
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
75
Alcor is an arrow.
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
76
Alcor is a knife.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
77 Alcor is a pack 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
78
The Big Dipper is a drawing.
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
79 Orion is a pursuer. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0
80
The three stars of Orion’s Belt are
interpreted as three animals
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 1 1 1 0 1 0
81 Th
e three stars of Orion’s Belt are
interpreted as three pursuers.
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0
82
Three stars of Orion’s Belt are
interpreted as one single animal.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
1
0
0
0
83
One of the neighboring stars of Orion is
a sword shot by the hunter to hit the
animals.
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 1 0 1 1 1 0 1 0 0 0 0 0
84
The sword of Orion is a pursuer.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
85
Canopus is a hunter
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
86
The Hyades are a game.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
87
The Pleiades are hunters.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
1
0
0
1
1
0
0
1
0
0
0
0
0
0
0
0
0
0
88
The Pleiades are an animal.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
89
Cassiope is a game.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
90
Betelgeuse is a game.
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
91
One of the main constellation of the
story is the Big Dipper
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
0
0
0
1
0
1
1
0
0
1
0
0
0
0
0
1
0
0
0
0
0
0
92 One
of the main constellation of the
story is Orion.
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 1 0 1 0 0 0 1 0 1 0 1 1 1 0 1 1 1 1 1 1
93
One of the main constellation of the
story is the Pleiades
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
1
1
0
0
1
0
0
0
0
0
0
0
0
0
0

A Cosmic Hunt in the Berber Sky : A Phylogenetic Reconstruction
105
Allen Richard Hinckley 1899. Star-Names and their
Meanings. New York & Leipzig : G. E. Stechert, 563 p.
Anderson Gregory 2003. Yeniseic languages in Siberian
areal perspective. Sprachtypologie und Universa-
lienforschung 56 (1/2) : 12-39.
Anisimov A.F. 1958. Religiya evenkov v istoriko-gene-
ticheskom izuchenii i problemy proiskhozhdeniya
pervobyt- nykh verovanii. Moscow-Leningra : Aka-
demiya nauk.
Atkinson Quentin D., Andrew Meade, Chris Venditti,
Simon J. Greenhill & Mark Pagel 2008. Languages
evolve in punctional bursts. Science 319: 588.
Basset René 1910. Recherches sur la Religion des Ber-
bères. Paris : Ernest Leroux, 52 p.
Bernus Edmond & Ag Sidiyene Ehya (1989). Étoiles
et constellations chez les nomades. Awal, Cahier
d’Études Berbères 5 : 141-153.
Benhazera Maurice 1908. Six mois chez les Touaregs du
Ahaggar. Alger : A. Jourdan, 233p.
Berezkin Yuri 2005. The cosmic hunt: variants of a Sibe-
rian – North-American myth. Folklore 31: 79-100.
——— 2012. Seven Brothers and the Cosmic Hunt:
European sky in the past. Paar sammukest XXVI,
Eesti Kirjandusmuuseumi aastaraamat 2009,
Tartu: Eesti kirjandusmuuseum : 31-69.
Billson Charles James 1918. Some mythical tales of the
Lapps. Folklore 29 (3): 178–192.
Boas Franz 1888. The Central Eskimo. 6th Annual Report
of the Bureau of Ethnology to the Secretary of the
Smithsonian Instituition (1884-1885) : 39 9 -6 6 9.
Bogoras Waldemar 1924. New Problems of Ethnogra-
phic Research in Polar Countries. In : Proceedings
of the 21stInternational Congress of Americanists,
First Part. Held at The Hague, August 12-16, 1924,
Leiden, E.J. Brill : 226-246.
Brett William Henry 1880.Legends and Myths of the
Aboriginal Indians of British Guiana. London :
William Wells Gardner. 206p.
Brown M.D., S.H. Hosseini, A. Torroni, H.J. Bandelt,
J.C. Allen, T.G. Schurr, R. Scozzari, F. Cruciani & D.
Wallace 1998. mtDNA haplogroup X : an ancient link
between Europe/ Western Asia and North America ? The
American Journal of Human Genetics 6 3 : 18 52 –1861.
Bryant David & Vincent Moulton 2004. Neighbor-Net :
an agglomerative method for the construction of
phylogenetic networks. Molecular Biology and Evo-
lution 21 : 255 –265.
Bryant David, Flavia Filimon & Russell D. Gray 2005.
Untangling our past : languages, trees, splits and
networks. In : R. Mace, C. J. Holden & S. Shennan
(ed.), The evolution of cultural diversity : phylogene-
tic approaches, Londres, UCL Press : 67–84.
Bulatova Angara Gamidovna 2003. Rutultsy in XIX –
nachale XX vekov (istoriko-etnograficheskoe iss-
ledovanie). Moscow : Institute of Ethnology and
Anthropology Press.
Cerquand Jean-François 1875. Légendes et récits popu-
laire du pays basque – 1. Pau: Léon Ribaut, 74 p.
Clark Ella Elizabeth 1953. Indian Legends of the Paci-
fic Northwest. Berkeley / Los Angeles : University of
California Press, 225 p.
Dediu Dan & Stephen C. Levinson 2012. Abstract pro-
files of structural stability point to universal tenden-
cies, family-specific factors, and ancient connec-
tions between languages. PlosOne 7 (9) e45198.
Duhard Jean-Pierre 1996. Réalisme de l’image mascu-
line paléolithique. Grenoble : Jérôme Million, 245 p.
Duveyrier Henri 1864. Les Touaregs du Nord. Paris :
Challamel Ainé, 499 p.
Ernits Enn 2010. On the Cosmic Hunt in North Eurasian
Rock Art. Folklore (Tartu) 44 : 61-76.
Farrand Livingston 1900. Traditions of the Chilco-
tin Indians. New York : Memoires of the American
Museum of Natural History, vol, 4. Publications of the
Jesup North Pacific Expedition, vol.2, part 1. 54p.
Fowler Catherine S. 1995. Mountain sheep in the sky :
Orion’s Belt in Great Basin mythology. Journal of
California and Great Basin Anthropology 17(2) :
146 -152.
Goddard Pliny Earle 1904. Hupa Texts. Berkeley : The
University Press, 278 p.
Greenhill Simon, Quentin Atkinson & Russell Gray
2010. The shape and tempo of language evolution.
Proceedings of the Royal Society B Biological
Sciences, 10.1098/rspb.2010.0051.
Hayden Brian & Suzanne Villeneuve 2011. Astronomy
in the Upper Palaeolithic ? Cambridge Archaeologi-
cal Journal 21(3) : 331-355.
Heggarty Paul, Warren Maguire & April McMahon
2010. Splits or waves ? Trees or webs ? How diver-
gence measures and network analysis can unravel
language histories. Philosophical Transactions of
the Royal Society B 365 : 3829-3843.
Holland B. R., K. T. Huber, A. Dress A. & V. Moulton
2002. δ plots : a tool for analyzing phylogenetic dis-
tance data. Molecular Biology and Evolution 19 :
2051-2059.
Holtved Erik 1951.The Polar Eskimos. Language and
Folklore. II. Myths and Tales translated by Erik
Holtved. København : C. A. Reitzels Forlag. Med-
delelser om Grønland. Udgivne af Kommissionen
for Videnskabelige Undersøglser i Grønland 152 (2),
153 p.
d’Huy Julien 2011. Le mythe ovidien de Pygmalion trou-
verait l’une de ses origines dans la Berbérie préhis-
torique. Les Cahiers de l’AARS 15: 19-25.
——— 2012a. Le motif de Pygmalion : origine afrasienne
et diffusion en Afrique. Sahara 23 : 49-58.
——— 2012b. Mythes, langues et génétique. Mythologie
française 247 : 25-26.
——— 2012c. Le conte-type de Polyphème : essai de
reconstitution phylogénétique. Mythologie française
248 : 47-59.
——— 2012d. Un ours dans les étoiles : recherche phylo-
génétique sur un mythe préhistorique. Préhistoire du
Sud-Ouest, 20 (1) : 91-106.
Références

106
d’Huy Julien 2012e. L’Aquitaine sur la route d’Oedipe.
La Sphinge comme motif préhistorique. Bulletin de
la SERPE 61: 15 -21.
——— 2013a. A phylogenetic approach of mythology
and its archaeological consequences. Rock Art
Research 30(1), to be published.
——— 2013b. Polyphemus (Aa. Th. 1137): a phylo-
genetic reconstruction of a prehistoric tale. Nou-
velle Mythologie Comparée / New Compara-
tiveMythology1,Online.
Drummond Alexei J. & Andrew Rambaut 2007. BEAST:
Bayesian evolutionary analysis by sampling trees.
BMC Evolutionary Biology 7, 214.
Holman Eric W., Christian Schulze, Dietrich Stauffer
& Søren Wichmann 2007. On the relation between
structural diversity and geographical distance
among languages: observations and computer simu-
lations. Linguistic Typology 11(2) : 395-423.
Huelsenbeck John P. & Fredrik Ronquist 2001.
MRBAYES: Bayesian inference of phylogeny.
Bioinformatics 17 : 754-755.
Huson Daniel H. & Bryant David 2006. Application
of Phylogenetic Networks in Evolutionary Studies.
Molecular Biology and Evolution 23(2) : 254-267.
Jacobson Esther 1993. The Deer Goddess of ancient
Siberia : a study in the ecology of belief. Leiden :
E.J. Brill, 291 p.
Kass Robert E. & Adrian E. Raftery 1995. Bayes factors.
Journal of the AmericanStatistical Association 90
(43 0) : 773 -79 5.
Koekemoer Gerhardus P. 2007. Lighting design in an
African context. Submitted in partial fulfilment of
the requirements for the degree magister techno-
logiae : performing arts technology in the Depart-
ment of entertainment technology faculty of the arts,
Tshwane university of technology, 202 p.
Kroeber Alfred Luis 1899. Tales of the Smith Sound
Eskimo. Journal of American Folklore 12(46) : 166-
182.
Lévi-Strauss Claude 1971. Mythologique 4. L’Homme
nu. Paris, Plon, 681p.
——— 19 85. La Potière jalouse. Paris: Plon, 314 p.
Lushnikova A.V. 2003. Ursa Major : ot losya do med-
vedya (Ursa Major: from elk to bear). Istoriko-Astro-
nomicheskie Issledovaniya 28 : 189-216, 342- 343.
Maddison W. P. & D.R. Maddison. 2011. Mesquite: a
modular system for evolutionary analysis. Version
2.75 http://mesquiteproject.org
Magaña Edmundo 1983. Star myths of the Kaliña
(Carib) Indians of Surinam. Latin American Indian
Literatures 7(1) : 20-37.
Meigs Peveril 1939. The Kiliwa Indians of Lower Cali-
fornia. Berkeley : Ibero-Americana, 88 p.
Nilsson Martin P. 1920. Primitive Time-reckoning : A
study in the origins and first development of the
art of counting time among the primitive and early
culture peoples. Paris : Champion (Acta societatis
humanorum litterarum lundensis 1) 384 p.
Nunn Charles L., Christian Arnold, Luke Matthews &
Monique Borgerhoff 2010. Simulating trait evolution
for cross-cultural comparison. Philosophical Tran-
sactions of the Royal Society 365 (1559) : 3807-3819.
Pentikäinen Juha 2011. Mythologie des Lapons. Pa ris :
Imago, 350p.
Pottier Jeanne-René 1946. Légendes touareg. Paris :
Nouvelles Editions Latines, 253 p.
Rasmussen Knud 1931. The Netsilik Eskimos. Social
Life and Spiritual Culture. Copenhagen : Report of
the Fifth Thule Expedition 1921-24, 8 (1-2), 542 p.
——— 19 32 . Intellectual Culture of the Copper Eski-
mos. Copenhagen : Report of the Fifth Thule Expe-
dition 1921-24, vol.9. 350 p.
Rivera de Bianchi Mabel 1973.Mitología de los pueblos
del Chaco, según visión de los autores de los siglos
XVII y XVIII. América Indígena 23(3) : 695-733.
Ronquist Fredrik & John P. Huelsenbeck 2003.
MRBAYES 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19 : 1572-1574.
Roth Walter E. 1915. An Inquiry into the Animism and
Folklore of the Guiana Indians. 30th Annual Report
of the Bureau of Ethnology to the Secretary of the
Smithsonian Institution (1908-1909) : 26 -745.
Speck Frank Gouldsmith 1915.Myths and Folklore of
the Timiskaming Algonquin and Timagami Ojibwa.
Ottawa : Canada Department of Mines (Geological
Survey Memoir 71. Anthropological Series 9) 87 p.
Spier Leslie 1978. Yuman Tribes of the Gila River.
Chicago : University of Chicago Press, 1433 p.
Stefanini Jean 1926. Au Pays d’Antinéa. Paris : Plon,
180 p.
Shevoroshkin Vitaliy 2003. Salishan and North Cauca-
sian. Mother Tongue 8: 39–64.
——— 2008. On The Origin of Salish, Wakashnan, and
North Caucasian Languages. International Journal
of Modern Anthropology 1 (1) : 84-121.
Teit James A. & Franz Boas 1930. The Salishan tribes
of the Western Plateau. 45th Annual Report of the
Bureau of Ethnology to the Secretary of Smithso-
nian Institution (1927-1928) : 23 -396.
Venditti Chris, Andrew Meade & Mark Pagel 2006.
Detecting the node-density artifact in phylogeny
reconstruction. Systematic biology 55 (4) : 637-643.
Vinson Julien 1883. Le Folk-lore du Pays Basque. Pa ris :
Maisonneuve, 396 p.
Webster Andrea J., Robert J. H. Payne & Mark Pagel
2003. Molecular phylogenies link rates of evolution
and speciation. Science 301 : 478.
Wichmann Søren, Eric W. Holman, Taraka Rama &
Robert S. Walker 2011. Correlates of reticulation
in linguistic phylogenies. Language Dynamics and
Change 1: 205-240.
Williams Mentor L. 1956. Schoolcraft’s Indian Legends.
Michigan : Michigan State University Press, 322p.
Xie Wangang, Paul O. Lewis, Yu Fan, Lynn Kuo &
Ming-Hui Chen 2011. Improving marginal likeli-
hood estimation for Bayesian phylogenetic model
selection. Systematic Biology 6 0(2) : 150 –160.