Content uploaded by Robert G. Bednarik
Author content
All content in this area was uploaded by Robert G. Bednarik on Jan 25, 2022
Content may be subject to copyright.
Content uploaded by Robert G. Bednarik
Author content
All content in this area was uploaded by Robert G. Bednarik on Jan 25, 2022
Content may be subject to copyright.
AURA Newsletter, Volume 14, Number 2, September 1997, pp. 9-12.
Direct dating results from rock art: a global review
ROBERT G. BEDNARIK
Direct dating of rock art
‘Direct dating’ of rock art was introduced as an alternative
to archaeological or stylistic dating, the only methods used until
1980. In recognising the subjective or inductive nature of these
traditional approaches of determining the age of rock art,
scientists began developing a new methodology in the 1970s
and 1980s. ‘Direct dating’ is predicated on two crucial precon-
ditions. First, the physical relationship of the art and the dating
criterion must be direct and indisputable. Second, the proposi-
tions concerning the chronological relationship of the rock art
and the dating criterion must be falsifiable.
A series of methods have been developed that can comply
with these strict demands, and are likely to offer scientific (i.e.
refutable) information relevant to the age of rock art. None of
them provide real and absolute dates, and their precision and
reliability are of a considerable range. Nevertheless, provided
that their epistemic limitations are appreciated and appropriately
accommodated in interpretative pronouncements, the results of
these methods provide scientifically sound information.
Problems have arisen where they have been interpreted without
recourse to the relevant qualifications (Bednarik 1994a, 1994b;
Ward 1994).
Since 1990, direct dating of rock art has been widely used in
many parts of the world, and it is of interest that 40 percent of
the published dating projects so far (late 1995) were conducted
in Australia, or by Australians abroad, since the methodology
was first introduced in 1980, in Australia (Bednarik 1993).
Australia has been very much in the forefront of this
development. This methodology has led to a revolution in our
understanding of various palaeoart systems, and in many cases
to the challenge or invalidation of previous archaeological
models. Its value as part of an analytical rock art science being
developed has been amply demonstrated already, but it must be
cautioned that it remains severely limited. It is an experimental
methodology and it is being developed quite randomly, through
the preferences of individuals and with very little overall stra-
tegy. It is therefore useful to review the achievements of this
approach, as well as the dynamics determining progress and
priorities, in order to more clearly see the shortcomings,
strengths and direction in this field. Policies and strategies of
researchers, institutions and funding agencies should be devel-
oped only on the basis of such analytical data. Without them,
this field will continue to develop haphazardly and opportunis-
tically, with individual researchers seeking to maximise the
effects of their work. In this paper the work so far produced is
reviewed.
Overview of data
I have summarised in Table 1 the direct dating results so far
obtained which I regard as offering a reasonable level of
credibility. This list was compiled in late 1995 and it does not
include cation-ratio or cosmogenic nuclide results because they
are not considered sufficiently reliable. Additions to this list,
which is to be maintained as a cumulative record, are invited
from readers.
Each entry includes the year when the date was either
obtained or published, the name of the site or region, the
method used, the number of dates secured, approximate
results, and the name and country of the researcher(s) who
produced the results.
1980 - Malangine Cave, South Australia - Radiocarbon dating of reprecipitated carbonate over and under petroglyphs - Two -
Minimum age of c. 5550 BP for some petroglyphs - R. G. Bednarik, Australia.
1982 - Malangine Cave, South Australia - Uranium/thorium dating of reprecipitated carbonate over and under petroglyphs - Two -
Minimum age of c. 28 000 BP for some petroglyphs - R. G. Bednarik, H. H. Veeh, Australia.
1987 - Boontjieskloof, South Africa - AMS radiocarbon dating of charcoal pigment - One - 500 BP - N. J. van der Merwe, J. Sealy,
R. Yates, South Africa.
1987 - Huashan, China - Radiocarbon dating of reprecipitated carbonate over and under painting - Two for one motif - 2370 BP
max. and 2115 BP min. - Qin Shengmin, China.
1990 - Kakadu National Park, Australia - Radiocarbon dating of oxalate associated with paintings - Two - Up to c. 8880 BP - A.
Watchman, Australia.
1990 - Lake Onega, Russia - Microerosion dating of petroglyph - One - E4000 BP - R. G. Bednarik, Australia.
1990 - Cougnac Cave, France - AMS dating of charcoal pigment - One - 14 300 BP - M. Lorblanchet, France.
1990 - Gnatalia Creek, Australia - AMS dating of charcoal pigment - Two - 6000 BP and 29 800 BP from one motif - J. McDonald,
K. Officer, T. Jull, D. Donahue, H. Head, B. Ford, Australia.
1990 - Waterfall Cave, Australia - AMS dating of charcoal pigment - One - 600 BP - J. McDonald, K. Officer, T. Jull, D. Donahue,
H. Head, B. Ford, Australia.
1990 - Judds Cavern, Tasmania - AMS dating of blood residue in pigment - Two - 10 000 BP - T. H. Loy, R. Jones, D. E. Nelson, B.
Meehan, J. Vogel, J. Southon, R. Cosgrove, Australia-Canada.
1990 - Laurie Creek, Australia - AMS dating of blood residue in pigment - One - C. 20 000 BP - T. H. Loy, R. Jones, D. E. Nelson,
B. Meehan, J. Vogel, J. Southon, R. Cosgrove, Australia-Canada. (This result is now doubtful.)
1990 - Seminole Canyon, Texas - AMS dating of organic carbon in pigment - One - 3900 BP - J. Russ, M. Hyman, H. J. Shafer, M.
W. Rowe, USA.
1990 - Cangyuan, China - Radiocarbon dating of reprecipitated carbonate under and over painting - Several for one motif - 3100 BP
max. and 2960 BP min. - Woo Sheh Ming, China.
1992 - Cosquer Cave, France - AMS dating of charcoal pigment - Seven - Ranging from c. 18 000 BP to 27 900 BP - J. Clottes, J.
Courtin, H. Valladas, France.
1992 - Niaux, France - AMS dating of charcoal pigment - Three - Ranging from c. 12 900 BP to 13 850 BP - H. Valladas, H.
Cachier, M. Arnold, F. Bernaldo de Quiros, J. Clottes, P. Uzquiano, France-Spain.
1992 - Altamira, Spain - AMS dating of charcoal pigment - Three - Ranging from 13 600 BP to 14 300 BP - H. Valladas, H. Cachier,
M. Arnold, F. Bernaldo de Quiros, J. Clottes, P. Uzquiano, France-Spain.
1992 - Castillo, Spain - AMS dating of charcoal pigment - Two - C. 13 000 BP - H. Valladas, H. Cachier, M. Arnold, F. Bernaldo de
Quiros, J. Clottes, P. Uzquiano, France-Spain.
1992 - Laura South, Australia - AMS dating of oxalate containing paint residue - One - 24 600 BP - A. Watchman.
1992 - Lower Pecos River region, Texas - AMS dating of organic carbon in pigment - Two - Both c. 3000 BP - J. Russ, M. Hyman,
M. W. Rowe, USA.
1993 - Olary Province, Australia - AMS dating from organics on petroglyphs - Nineteen - Ranging from c. 1500 BP to 43 140 BP -
M. Nobbs, R. Dorn, Australia-USA.
1993 - Cougnac, France - AMS dating of charcoal pigment - Five - Ranging from c. 13 800 BP to 25 100 BP - M. Lorblanchet,
France.
1993 - Yam Camp, Laura, Australia - AMS dating of plant-fibre binders in rock paintings - Two - Both c. 700 BP - A. Watchman, N.
Cole, Australia.
1993 - Lower Pecos River region, Texas - AMS dating of organic carbon in pigment - Two - 3000 and 4200 BP - S. D. Chaffee, M.
Hyman, M. W. Rowe, USA.
1993 - Arnhem Land, Australia - AMS dating of beeswax figures - Several - Up to 4000 BP - D. E. Nelson, C. Chippindale, G.
Chaloupka, P. Taçon, J. Southon, Canada-Australia-U.K.-USA.
1994 - Pryor Mountains, Montana - AMS dating of organic carbon in pigment - One - 840 BP - S. D. Chaffee, L. L. Loendorf, M.
Hyman, M. W. Rowe, USA.
1994 - Canyonlands National Park, Utah - AMS dating of charcoal pigment and organic carbon in pigment - Two - 600 and 750 BP -
S. D. Chaffee, M. Hyman, M. W. Rowe, N. J. Coulam, A. Schroedl, K. Hogue, USA.
1994 - All American Man painting, Utah, U.S.A. - AMS dating of paint residue - Two - 575 BP and 753 BP - S. D. Chaffee, M.
Hyman, M. W. Rowe, N. J. Coulam, A. Schroedl, K. Hogue, U.S.A.
1994 - El Raton, Baja California Sur, Mexico - AMS dating of paints - Four - 300 BP to 5000 BP (probably contaminated) - J. M.
Fullola, V. del Castillo, M. A. Petit, A. Rubio, E. Sarrià, R. Viñas, Spain.
1994 - Auditorium Cave, India - Microerosion dating of petroglyphs - One - Beyond range of method - R. G. Bednarik, Australia.
1995 - Le Portel, France - AMS dating of charcoal pigment - Two - 11 600 and 12 200 BP - W. A. Ilger, M. Dauvois, M. Hyman, M.
Menu, M. W. Rowe, J. Vezian, P. Walter, USA-France.
1995 - Grotte Chauvet, France - AMS dating of charcoal pigment and torch soot - Nine - Ranging from c. 26 100 to 32 400 BP - J.
Clottes, J.-M. Chauvet, E. Brunel-Deschamps, C. Hillaire, J.-P. Daugas, M. Arnold, H. Cachier, J. Evin, P. Fortin, C. Oberlin, N.
Tisnerat, H. Valladas, France.
1995 - Pech Merle, France - AMS dating of charcoal pigment - One - C. 24 600 BP - M. Lorblanchet, H. Cachier, H. Valladas,
France.
1995 - Drakensburg, Natal, South Africa - AMS dating from plant fibre binders - Two - Of c. 300 and 400 BP - A. Mazel, A.
Watchman, South Africa-Canada.
1995 - Walkunder Arch, Chillagoe, Australia - AMS dating of oxalates associated with paintings - Four - from c. 6800 - 28 100 BP -
A. Watchman, J. Campbell, C. Tuniz, Canada-Australia.
1995 - Côa site complex, Portugal - AMS dating from organics on petroglyphs and adjacent surfaces - Eight - Primary dates ranging
from c. 2000 to 7000 BP, corrected to 100 BP to 1700 BP - A. Watchman, Canada.
1995 - Côa site complex, Portugal - AMS dating from organics on petroglyphs - Ten - Primary dates from c. 2000 to 5100 BP - R.
Dorn, USA.
1995 - Côa site complex, Portugal - Microerosion analyses of petroglyphs - Numerous - Oldest figure c. E6500 BP, youngest a few
centuries - R. G. Bednarik, Australia.
1995 - Grosio, Italy - Microerosion dating of petroglyph - One - E4900 BP - R. G. Bednarik, Australia.
1995 - Lewis Canyon, Texas - AMS dating of organic carbon in pigment - One - 1100 BP - W. A. Ilger, M. Hyman, J. Southon, M.
W. Rowe, USA.
1995 - Opeleva Cave, Angola - AMS dating of organic charcoal in pigment - Two - Both c. 1900 BP - W. A. Ilger, M. Hyman, J.
Southon, M. W. Rowe, USA.
1995 - Seminole Canyon, Texas - AMS dating of organic carbon in pigment - Four replicates - From 2950 to 3600 BP - W. A. Ilger,
M. Hyman, J. Southon, M. W. Rowe, USA.
Table 1. Direct rock art dating projects.
It follows from this information that soon after direct dating of rock art was introduced, a strong bias for just one method
developed. I divide here the direct dates so far reported into three groups: those derived by radiocarbon from organic matter; by
radiocarbon from mineral matter; and by other methods (Table 2).
1980
1982
1987
1990
1992
1993
1994
1995
Totals
14C, organics
1
8
17
32
9
39
106
14C, mineral
2
4
3
1
4
14
Other methods
2
1
1
2
6
Totals
2
2
5
12
18
32
10
44
126
Table 2. Direct rock art dates obtained in various years, comparing the respective numbers secured by different means. The table
shows a distinct bias in favour of radiocarbon dates derived from organic matter.
This shows a massive preference for the former method,
which accounts for 84 percent of all such dates since I began
direct dating in 1980. The systematic neglect of all methods
other than radiocarbon determination of organic matter is amply
evident and illustrates also how initially alternative methods
were tried and are still being applied occasionally, but how the
analysis of organic matter came to dominate the field entirely by
1992 or 1993.
Discussion
The conclusion one is likely to draw from this pattern is that
the apparently preferred method was found to be the best or
most reliable, and was therefore widely adopted around the
world. But this is not the case at all. There are several alternative
methods available, and there are numerous techniques possible
or potentially applicable in the field, yet their use has not been
attempted so far. One of the reasons for this is that the
developing monopoly of just one method actively discourages
the use of alternative methods.
But there are still other drawbacks involved. By permitting a
whole methodology to be entirely dominated by a single method
we needlessly create a dependence, and if the one method we
use almost exclusively turned out to have severe flaws, years of
effort may go to waste. It seems to be a very imprudent strategy
to favour one method before we have even explored the whole
range of methods available to us, and before we have had any
opportunity to compare the one method with all the others. But
perhaps most importantly, it is already known that the favoured
method does have severe shortcomings. I have demonstrated
that not one of the many 14C dates from charcoal in European
caves represents a real age of the picture in question (Bednarik
1994a, 1994b, 1996), they are all maximum ages, and perhaps
very conservative maximum ages. Similarly, 14C dates from
organic material in mineral accretions, while no doubt related in
some complex fashion to the age of the art they cover, do not
actually date it, and may in fact be so different from its age that
they can only be misleading.
At this stage in the development of the new methodology,
confidence can only be acquired through a pluralist approach, in
which a variety of techniques are harnessed to guard against
errors resulting from hasty pronouncements and misinterpreta-
tions. In nearly all cases this is not happening, the most notable
exception being the Côa valley sites in Portugal (where a series
of blind tests was conducted; Bednarik 1995; Watchman 1995,
1996). In Australia, a multi-pronged approach was initially
introduced, which differed significantly from the strategy
adopted in European Palaeolithic art dating. While this and the
rather individualistic approach were the main strengths of the
discipline in Australia and contrasted with European initiatives,
current developments in this field would soon extinguish that
spirit and replace it with a bland research ‘industry’ centred
around the AMS dating of organic matter. That would be most
unfortunate, because it would soon erode the credibility of the
discipline. Radiocarbon occurs naturally in most rock substrates
(Bednarik 1979) and in most paint residues. Some of it occurs in
accretions covering petroglyphs, some of it in charcoal or other
paint residues. Some of it relates to the age of the rock art, some
of it does not. We cannot distinguish between the two sources,
hence we cannot date rock art by this single-method approach.
This inherent limitation questions the validity of radiocarbon
analyses even before we consider the general methodological
constraints of the radiocarbon method (Ward 1994; Ward and
Wilson 1978; Wilson and Ward 1981; Bednarik 1994a).
In summary, direct dating of rock art remains in an expe-
rimental stage and we need to keep our options open and exer-
cise methodological pluralism. At present we seem to adopt a
Figure 1. Graphic depiction of the number of direct rock art
‘dates’ per year, comparing 14C dates derived from organic
matter (n = 106), 14C dates from mineral matter (n = 14),
and dates resulting from other methods (n = 6). CR dates
are not considered.
policy of placing all our bets on one option that is known to
provide only questionable dating results (Bednarik 1996). While
it is true that Australians have pioneered the specialist field of
direct dating of rock art, the institutional follow-up to these
individual endeavours has generally been inadequate and
discouraging. It should be adequately clear from the data pre-
sented here that Australian efforts in scientific rock art dating
have been most impressive by any standards. To continue this
strong research tradition I believe that it is imperative to adopt a
significantly more diversified approach.
Acknowledgments
The help of Dr A. Watchman and Professor M. W. Rowe in com-
piling Table 1 is acknowledged.
REFERENCES
BEDNARIK, R. G. 1979. The potential of rock patination analysis in
Australian archaeology - Part 1. The Artefact 4: 14-38.
BEDNARIK, R. G. 1993. The direct dating of rock art. Rock Art
Research 10: 48-51.
BEDNARIK, R. G. 1994a. Conceptual pitfalls in Palaeolithic rock art
dating. Préhistoire Anthropologie Méditerranéennes 3: 95-102.
BEDNARIK, R. G. 1994b. About rock art dating. International
Newsletter on Rock Art 7: 16-18.
BEDNARIK, R. G. 1995. The age of the Côa valley petroglyphs in
Portugal. Rock Art Research 12: 86-103.
BEDNARIK, R. G. 1996. Only time will tell: a review of the method-
ology of direct rock art dating. Archaeometry 38: 1-13.
WARD, G. K. 1994. On the use of radiometric determinations to ‘date’
archaeological events. Australian Aboriginal Studies 1994/2: 106-9.
WARD, G. K. and S. R. WILSON 1978. Procedures for comparing and
combining radiocarbon age determinations: a critique. Archae-
ometry 20: 19-31.
WATCHMAN, A. 1995. Recent petroglyphs, Foz Côa, Portugal. Rock
Art Research 12: 104-8.
WATCHMAN, A. 1996. A review of the theory and assumptions in the
AMS dating of the Foz Côa petroglyphs, Portugal. Rock Art
Research 13: 21-30.
WILSON, S. R. and G. K. WARD 1981. Evaluation and clustering of
radiocarbon age estimates: procedures and paradigms. Archaeome-
try 23: 19-39.
