Horse, bison and reindeer, Chauvet Cave, France. Internet image, public domain.
From the
discovery of the painted caves of Europe (and the eventual acceptance of their
authenticity) people have been trying to reason out a meaning for these images.
Of course, one of the first theories posited to explain them was hunting magic,
the idea that creating the image of a game animal would improve one’s chances
of success in the hunt. Another way of looking at this art was to attempt
organizational analysis, what would be called ‘design’ in art history.
In ‘The
Hunters of Prehistory’, in 1989, Andre Leroi-Gourhan wrote “it is certain that the men of the Upper Paleolithic had a highly
developed system of beliefs, expressed in symbolic images related to the world
of hunting. It was long believed that the animals represented prey, whose
portrait was ‘killed’, and so, by magic, the animal itself. The detailed study
of some one hundred ornamented caves in France and Spain revealed that,
although the caves don’t follow the rules of modern composition and perspective,
the positions of the images could not be explained solely by chance.”
(Leroi-Gourhan 1989:137-138)
Horses and bison, Ekain, Spain. Image from Leroi-Gourhan, 1989, page 139. Image clarified by chatGPT.
When
interest turned to the design of the painted panels (their patterning)
relationships between various species have been posited and expressed. “The list of animals represented on walls is
shorter than for those depicted on bone artifacts. On walls horse-bison plus
ibex is the most widely distributed assemblage. Horse and bison (or auroch)
make up the basic formula, but examples of just these two species are rare. In
ond case, the figures are located face to face on two walls.”
(Leroi-Gourhan 1989:140) So here is posited a relationship between animal
imagery that leads to design decisions on placement.
Horse, auroch, lion and reindeer, Chauvet Cave, France. Internet image, public domain.
Other
species are assigned places in these assumed relationships as well. “The most dangerous animals – bear, large
felines and rhinoceros – have positions at the edge of the engraved panels (Les
Trois-Freres), in the grouping farthest away from the entrance (Lascaux), or
simply at the end of the grouping (Labastide).” (Leroi-Gourhan 1989:140) In
this he is leaning on a much shakier assumption, indeed, perhaps the bear,
large feline and rhinoceros are the important groupings, and the other animals
grouped somewhere away from them.
What seems
to have been missing in these analyses is any indication of the painting style
of individual animals. In some cases the hand of the artist is recognizable in
a group of animals, but in other instances animals in close proximity may have
been done by different hands at different times. This would make it somewhat
harder to state a certain relationship between animals. Different minds and
different artists may not see the supposed rules quite the same way. In other
words, I might paint the bison here just because I want the bison here,
irrespective of other animals in its vicinity.
Horses, aurochs, and reindeer, Lascaux Cave, France. Internet image, public domain.
A new
study, and statistical analysis, of painted caves in France, Spain, and Italy
has come out with conclusions that re-open the arguments about intentional
groupings of painted subjects.
“Taken together, the network-based
results support the existence of an underlying organizational grammar governing
the combination and distribution of themes in Magdalenian cave art. This
structure operates at multiple, interrelated levels: through recurrent thematic
associations within panels; through hierarchical differentiation, whereby
certain motifs occupy structurally central positions; and through spatial
patterning, expressed in the uneven distribution and reiteration of motifs
across compositional units. Within this framework, large
herbivores—particularly bison, horse, and ibex—emerge as key organizing themes,
while other motifs appear to fulfill more contextual or articulatory roles.” (Intxaurbe 2026)
“These findings are consistent with
long-standing structural interpretations of Upper Paleolithic parietal art and
extend them by demonstrating their statistical robustness within an expanded,
systematically documented, and partly unpublished corpus. The minimum
spanning tree analysis further refines this picture by identifying bison as the
primary structuring node of the thematic hierarchy. This configuration closely
matches patterns previously described for both Cantabrian and Pyrenean contexts
and gains particular significance in the Basque region, which occupies an
intermediate geographical and cultural position between these areas. However,
sensitivity analyses indicate that the precise hierarchical ordering may vary
depending on the inclusion or exclusion of specific categories or assemblages,
with ibex or horse occasionally occupying the most central position. This
variability does not undermine the overall structure; rather, it suggests the
existence of a stable triadic core—bison, horse, and ibex—whose internal
ordering remains flexible. The recurrence of this triadic configuration across
analytical variants supports the hypothesis of partially shared organizational
principles operating within interconnected Magdalenian symbolic networks, while
the shifting prominence of each element points to complementary rather than
strictly hierarchical thematic roles.”
(Intxaurbe 2026) This analysis has a renewed reference to Leroi-Gourhan’s
bison, horse and ibex centrality.
Horses, aurochs, lion and reindeer, Chauvet Cave, France. Internet image, public domain.
“Beyond these substantive results,
this study highlights the methodological importance of grounding statistical
analyses of Paleolithic art in empirically controlled and taphonomically
assessed datasets. By integrating geoarcheological evaluation, detailed spatial
documentation, and an open, reproducible computational workflow, the approach
aims to minimize interpretative biases previously identified in the reuse of
fragmentary or insufficiently contextualised corpora. At the same time, we
deliberately refrain from advancing speculative interpretations concerning
symbolic meaning or narrative content. While the analyses demonstrate
structured organization and relational regularities, they do not—and
cannot—provide direct access to the social meanings or underlying symbolic
intentions.” (Intxaurbe
2026) The organizational and relational regularities mentioned may be as much
an artifact of the viewer than the creator. ‘I will keep looking until I find
what I want to see.’ So Intxaurbe may have covered more aspects of the subject
than I am aware of. For instance, has he considered things like ratios of size
of images to distances apart? Or, did he look at colors in his relational
analysis?
Horses, aurochs and reindeer, Chauvet Cave, France. Internet image, public domain.
With an
admirable lack of hubris, Intxaurbe offers his work as a tool or an approach to
aid others. He is not claiming to have solved everything.
“Rather than offering definitive
interpretations, this work should be understood as a methodological
contribution and an invitation to further comparative research. More broadly,
the approach offers a formal tool for investigating the internal organization
of symbolic systems and is therefore potentially relevant to adjacent fields
such as epigraphy, paleography, or historical linguistics, where co-occurrence
and combinatorial structure play a central role. The accumulation of
comparable, well-documented datasets and the adoption of shared analytical
frameworks represent the most promising path toward refining our understanding
of Paleolithic graphic systems.” (Intxaurbe
2026) There is, indeed, much truth to this. There are great portions of the
paper by Intxaurbe that are based upon mathematics and statistical analysis,
and I confess that I cannot understand much of that. An education in
traditional art history does not include any of these concepts. I point this
out to say that he may cover certain aspects of this subject that my reading of
his study did not give me insight to.
I will not
argue with these findings because, to a certain extent, I can believe that this
may be true for certain cave artists, at certain times and in a certain place.
My objection is the old one that having found something that works in
interpreting or analyzing an example of rock art it becomes the only
explanation considered. I have also frequently referred to my skepticism about
analyses proven by statistics, especially when the data has to be massaged to
get to the end result sought for – ‘I
want to find this so I will keep massaging the data until I do.’ This is, to
me, analogous to Laurenty Beria’s answer to Joseph Stalin, “show me the person and I will find the
crime.”
Horses, aurochs, reindeer and rhinoceros, Chauvet Cave, France. Internet image, public domain.
As someone who
has studied art history for sixty years I am just unable to accept any ‘one
answer fits all’ theory. I see no reason not to assume that one artist painted
a horse on the cave wall, and fifteen years later another artist wanted to
paint a bison and found a convenient spot. In such a case it would be hard for
me to agree that they are there in a mutual relationship. Intxaurbe has,
however, produced a piece of work that deserves serious consideration. Read his
paper, and who knows, you may agree with him.
NOTE: Some images in this column
were retrieved from the internet with a search for public domain photographs.
If any of these images are not intended to be public domain, I apologize, and
will happily provide the picture credits if the owner will contact me with
them. For further information on this you should read the original report
listed below.
REFERENCES:
Intxaurbe,
Iñaki, 2026, Mapping the Symbolic Structure
of Palaeolithic Rock Art Using Co‑occurrence
Network Analysis, Journal
of Archaeological Method and Theory 33:4. doi:10.1007/s10816-026-09796-y.
Accessed online 12 August 2026.
Leroi-Gourhan, Andre, 1989, The Hunters of Prehistory, translated byClaire Jacobson, Atheneum,
New York.
Posts
