Determination of genuineness using the thermolumines-
cence method ("TL" ) on cast cores of African bronze
objects 1:

Not an Illusion !
by Ralf Kotalla "Laboratory Kotalla - Haigerloch, founded 1979

It is possible based on practical experience and work on approx. 800 African
"bronze" objects to allocate old casting work to various ages and determine
work performed in modern times.

1. Basis

The determination of genuineness fulfils two necessary imperatives of
scientific dating methods (absolute dating):

1. A process with constant speed must take place, the reaction speed
should not depend on external conditions or environmental influence.

2. The reaction must take place at the time of manufacture of the object
being dated.

The "TL" method requires for this purpose the cast core of the bronze object -
a clay which comprises quartz and feldspar which was heated by metal flowing
into the casting mould, the geological TL is however, eliminated - zero position.
The elimination of the geological TL is the required reaction, the process of a
constant speed is the result of the annual new inflow ="annual dosage" of
energy to be stored until the time of measurement - test day - of the energy
accumulated since manufacture = zero position! The TL method determines
therefore the final time of heating which is usually the time of manufacture.

Demands for exact scientific dating not only of cast cores but also of customary
terra cotta from collections, trade and museums are extremely difficult to fulfil
in practical work as the site of discovery and its specific environmental factors
and their measurement parameters can no longer be reconstructed to calculate
the age exactly!
The "authenticity test with limited dating" whose result refers mainly to
material-internal effects of inner radioactivity and thus requires a wider range
of interpretation, suffices fully to recognise modern age work and to allocate
work to varying ages.
Due to the widely used "TL" in the meantime, we renounce on giving an account
of the physical basis of the "TL". The present level of knowledge can be found
in Aitken (1985) in detail or in Goedicke/Henschel (1994) - Possibilities and limits
of dating methods on bronze objects.

Copper-tin alloys are defined as "bronze". African metal objects contain mainly
zinc so that "brass" would really be the correct term. Lead and tin are to be
found in African alloys which is the reason for keeping to the term "bronze" as
ethnological literature mainly uses this term and thus avoids confusion.

2. Conditions

The age of a "bronze" object can be determined indirectly when a cast core of
fired clay (-with Benin max. 80% partly blackened quartz+feldspar due to wax)
is found for a TL analysis. The direct relation of the age to "bronze" is possible
when this cast core is connected with the manufacture = casting procedure =
reaction.

If the cast core originated directly from the "bronze object", the determined
age is also the age of the object as far as the last heating as manufacturing
process is concerned, irrespective of the subsequent external changes whether
natural or unnatural. (Cleaning, touching up, restoration etc.).

The determination of whether the cast core belongs to the object, is the
responsibility of the sampler whose experience and thoroughness in taking
samples is fundamental. Manipulation by attaching old cast cores or pulverised
clay must be determined and proved. (Details in Chapter "Practical experience").

3. Experience in applied practical work

a. Tested objects - sample sites - sampling - sampled amount -

Over the last 21 years, I have been presented with approx. 1100 African
"bronze objects" for performing a TL analysis. (Frequency ratio: 1.Head plates,
2. Pendant figures 3. Receptacles)
In approx. 300 objects no cast core material whatsoever or too low amounts
of usable cast cores were found. Of the approx. 800 examined objects,
20 % proved to be geologically contaminated and therefore could not be evaluated.

According to my practical experience, the claim that in open plates, pendants
and commemorative heads no core remains are to be found, must be
contradicted. (This also refutes the holdings of the Berlin Ethnology Museum
which at the present time has approx. 372 objects of this type of which over
300 objects have adequate cast cores. Also in the Ethnology Museum in
Dresden and similar museums there are sufficient examples to refute this claim).

The sampling sites are in some cases not easily accessible. The cast core
material must be removed near to the wall. With larger cores, it should be
noted that due to their poor thermal conductivity characteristics despite a
cast temperature of approx. 1000 °C , there is a risk that the geological TL
was not eliminated thus making classification impossible.

The sampling amounts which are removed for performing the authenticity
test, are around min. 40-100 mg so that a careful processing of samples
and at least one measuring sequence with 10 measurements per sample
can be realised. This relatively small sampling amount can be carefully
removed from the smallest pores of the wall, e.g. corners, edges, folds.
Contaminated samples give an unnatural curve diagram when measured
with an age which is probably too high or geological.

2O. Langevin "Determining the age of metals": An illusion? Tribal art 2000 II.
In the case of pulverised clay used to fake terracotta objects
(of corresponding age!), experience made by the laboratory in question, shows
that this can be detected without doubt using the natural TL curve diagram.

Old cast core of the same age used for faking!? This does not have the same
resistance and incrustation to the metal surface as a natural cast core.
If this resistance is produced by adhesive/bonding agents, this can be reliably
proved by the FT-IR Fourier-Transform infrared analysis. The scanning electron
microscope and the EDX (Energy dispersive x-ray analysis) also provide proof.

b Does the TL analysis require stylistic specifications ?

N O ! ! ! See point 1. and 2 on the demands on scientific absolute
dating/determination of genuineness !

On the contrary: The stylistics of African bronze sculpture require the
specifications of the TL analysis, if one can believe Willet 1983 who determined
that "the history of Nigerian sculpture still has an episodic character" Duchateau,
1989, Eszra 1992; Eisenhofer 1998 confirm this estimation by extensive work
on Benin "bronze casting".

Knowledge gained from experience (Langevin - Tribal Art 2000/ II), that the
large part of "bronze objects" offered on the African art market are not old,
must be "examined individually" and not generalised based on stylistic
specifications which are not reliable.
Absolute neutral analysis results must be taken into consideration for stylistic
determinations.

c. Accuracy - Tolerance

Accuracy based on practical work on approx. 800 African objects and approx.
400 Asian and numerous European bronze objects, can be determined as being
approx. 100% "New or old"
Although the age tolerance of approx. +/-25-30% as regards the determined
overall age is acceptable due to the usually unknown circumstances of the
site of discovery and the missing measurement data on site and their
replacement by interpretation variants, so-called mean values. This allows
stylistic considerations a natural scope. (300 years +/- 75 = 1625-1700-1775 a.d.)

The risk that the smallest geological remaining amounts see a recent signal
with a "fictitious age" in the range of 100-1100 years, is according to my
experience in the range of 2-4 per mil. This possibility only applies to so-called
extremely small sampling amounts around and under 40 mg.

d. Independent further scientific examination methods of "bronze" objects

Quoting references, so far, such as "Craddock" or "Werner" and their
examination methods as a possibility for determining genuineness and dating,
represent an outdated standard which is recognised as a standard examination
but which is supplemented by contemporary, more highly sensitive analysis
equipment and their use and which are far more convincing.

The latest Patina examinations - since 1997 - with e.g. species determination,
correlating, highly precise element analysis, data base-supported molecule
spectroscopy and atom spectrometry provide adequate scientific statements
on proving the genuineness of "bronze" objects.

Riederer 1994 - genuine and false - gives detailed insight into the multitude
of analysis possibilities and provides impartial interested parties with
extensive information.


e.: Final summary - personal concern

Thanks to TL analyses as the only possible absolute way to determine
genuineness/dating methods on cast cores and the newest metal and patina
analyses, it is possible, to provide references for developing reliable stylistics
and dating of the ancient cultures of Africa - e.g. Nigeria.

Traditions in cultures south of the Sahara were only provided orally and not
in writing (documents from that age which could be used for stylistics and
reasoning of same are virtually all missing). Here is the possibility of placing
methodical standpoints of ethnological circles, knowledge and experience
gained from trade and scientific methods on a moderate "ethnological,
practice-related scientific" foundation. I am aware that it is impossible to unite
all competitors when one-sided interests and ignorance block the readiness
to solve future tasks with new methods.

As far as my work with thermoluminescence is concerned, I would be grateful
if open discussions could take place on this difficult topic and cooperation
surmount existing barriers. This would strengthen each individual position
and would not lead to individual spheres or areas of responsibility becoming
superfluous.

It goes without saying, of course, when in individual cases, analysis results
are doubted, that the entire examination result is questioned. The obvious
reaction in such cases is to speak to the relevant expert before the validity
and accuracy of proven analytic methods are questioned and it is alleged
that these are no good. As an example, I would like to present the argumentation
on the use of old pulverised clays for manufacturing entire objects which today
can be recognised as such or the influence of sacrificing metal objects to
the patina growth process which is also recognisable. Those who ask experts
about this problem will find them willing to pursue this matter and answer
relevant questions.

The basis for a scientific/technical application of an analysis method remains
the neutral approach to the object without giving consideration to the mercantile
environment.

Literatur:

Aitken M.J.
1985 Thermolumineszenz Dating , London

Craddock, Paul T.
1985 Medieval Copper Alloy Production and West African Bronze
Analysis

Duchâteau
1989 Benin , Kunst einer Königskultur , Paris

Eisenhofer 1998

Ezra , K.
1992 Royal Art of Benin , The Metropolitan Museum of art , New York

Goedicke/Henschel
1994 Zur Chronologie der Berliner Benin-Bronzen , Möglichkeiten und
Grenzen naturwissenschaftlicher Datierungsmethoden Baessler Archiv
bd 41 Heft 2 Berlin 1994

Riederer , J.
1994 Echt und Falsch , Schätze der Vergangenheit im Museumslabor
Berlin, Heidelberg , New York

Werner , O.
1970 "Metallurgische Untersuchungen der Benin-Bronzen des
Museums für Völkerkunde Berlin "

Willet F. und Eyo. E.
1983 Kunstschätze aus Alt-Nigeria , Mainz