By Yuval Goren
Department of Archaeology and Ancient Near Eastern Cultures
In September 2001, an anonymous private collector brought a black stone tablet with an allegedly ancient Hebrew inscription to the Geological Survey of Israel (GSI) in order to verify its authenticity. The inscription includes a dedication text by Jehoash, son of Ahaziahu King of Judah, to commemorate his repairs of the Jerusalem temple. Three experts from the GSI studied the stone tablet and the patina coating its surface by using various chemical and mineralogical methods. In addition, tiny carbonized particles within the patina were dated using the radiocarbon AMS technique in the Beta Analytic Radiocarbon Dating Laboratory in Miami, Florida (Ilani, Rosenfeld, and Dvoracheck 2003:109).
The results led the GSI team to conclude that the Jehoash inscription was authentic and that it should be considered as belonging to the assemblage of Hebrew inscriptions of the first millennium BCE (Ilani, Rosenfeld, and Dvoracheck 2003:115-116). However, as we shall see below, although the analytical results are doubtlessly precise, the conclusion reached by the GIS scientists seems to be unduly biased. The mineralogical and chemical data indicate that the patina and other materials that appear on the tablet reflect acute anomalies that raise serious doubts about the authenticity of this artifact, or at least make it highly questionable. It is obvious that had they been more familiar with common forgery techniques and perhaps less enthusiastic to publish "an intriguing story," the GSI team may have come to entirely different conclusions. Other easily available analytical methods known from the vast literature on the topic of rock surface dating, which could better fit this particular case and supply decisive results, were overlooked. In this note, we shall survey in brief the main results of the GSI analysis and review them against the background of known forgery techniques, as well as the processes of patination and dating techniques of engraved rock surfaces.
METHOD, RESULTS, AND INTERPRETATION OF THE PREVIOUS STUDY
The detailed survey of the analytical methods that were applied by the GSI team is presented by Ilani, Rosenfeld, and Dvoracheck (2003) and will be only summarized here together with their main conclusions. I shall not review here the details of the engraving method, the script, and the contents of the text that are discussed to some detail by the above authors (ibid.) as these data should be reviewed by epigraphers, philologists, and Bible historians. The analytical method and results are as follows:
Nine samples of the patina coating the stone in various locations and three samples of the rock were taken by peeling with a scalpel, scratching with a diamond pencil and drilling with a fine hand drill. Surface examination as well as chemical and mineralogical analyses were performed using stereoscopic microscopy, Scanning Electron Microscopy (SEM) equipped with an Energy Dispersive Spectrometer (EDS), X-ray diffraction (XRD), and inductively coupled plasma atomic emission spectroscopy (ICP-AES). The carbonized particles within the patina were dated by radiocarbon AMS technique in the Beta Analytic Radiocarbon Dating Laboratory at Miami, Florida.
The rock is black, fine-textured, well-bedded arkosic sandstone composed mainly of silt-sized minerals with the addition of some quartz grains up to 0.5 mm in size. It was identified by its petrography and chemistry as originating from the Cambrian Shehoret Formation of southern Israel and Sinai, or its equivalent beds east of the Dead Sea and the eastern Arabah Valley.
The patina samples are composed of Si, O, Ca, Al, Mg, K and Fe. They contain angular carbon particles (presumably carbonized wood) and an anomalous number of globules of pure gold, each about one micron in size. The patina is enriched relatively to the mother rock by about 12% Fe2O3 and 4% CaO and diluted in K and Al. The carbon particles from within the patina were submitted for AMS radiocarbon age determination. The conventional radiocarbon age was 2250±40 BP (2-Sigma calibrated result: 390-200 BCE).
Ilani, Rosenfeld, and Dvoracheck (2002:115) conclude that the tablet and the inscription on it are genuine. They declare that "no indications of any adhesive materials or any other artificial substances were found. No evidence was found that could indicate the addition, adhesion or dispersion of artificial patina (or natural patina from a different stone) to the plate or the letters" (ibid. 114). The siliceous composition of the patina is similar to that of the arkosic sandstone and results only from it. The radiocarbon dates that postdate the destruction of the Temple by the Babylonians (namely 586 BCE) by about 200-400 years are connected in their opinion with the fire that destroyed Jerusalem and the temple, with the addition of younger carbon from later strata. The gold globules within the patina are interpreted as stemming from gold from the destruction of the gilded walls of the temple by fire (ibid. 115-116).
REVIEW OF THE PREVIOUS INTERPRETATIONS
Prior to the interpretation of the GSI data, we need to define some terminology. First, there are three principal ways to produce a fake artifact - forgery, imitation, and copying (e.g., Craddock and Barker 1990:11-27; Whittaker and Stafford 1999). Copies are artifacts that are tied to an information source, namely the imagery created to constitute a "brand world" to which these artifacts belong. Imitations ("soft fake") are artifacts that generate a false attribution to a context - a style, country, period, or producer. However, in the case of forgery ("hard fake"), a false attribution is created to the work or style of an individual, usually a major figure in his genre. The attribution is generated either within the work, for instance, through a false signature, or outside of it through claims by persons who are a source of authority in the field. In either case, a relationship is claimed between form and historical context, which never existed. If the Jehoash inscription proves to be a fake, it should undoubtedly be categorized as a forgery.
Depending upon a number of factors, the clues to forgery range from the simple, crude, and easy to identify saw marks and applied modern residues to more subtle characteristics such as weathering and the polish-like patina imparted by immersion in water, electrolysis, or burial for some years in an extreme environment (like under sheep enclosures). Faking antiquities is an ever-expanding field of recreational and professional interest, and in many cases, clever forgers can bestow the "correct" look to thoroughly modern fakes. Alongside the development of advanced laboratory techniques for the examination of archaeomaterials, forgers seemingly read the scientific literature and challenge it with ever-more sophisticated and creative performances. With tools and chemicals ranging from acids, potassium permanganate, rock tumblers, microwaves, ultrasonic baths, and electrolysis equipment, an experienced artifact faker can sometimes fool the best experts (Craddock and Barker 1990:11-27; Siegel, Saukko, and Knupfer 2000). Indeed, there were several cases in the history of archaeological research when fakes "survived" years of laboratory tests until they were finally exposed. One of the best-known examples is "the Piltdown Man," declared as genuine by the best natural scientists in Britain and revealed as a hoax only after nearly half a century. More relevant to our case is the ongoing scientific debate around "The Amana Mansoor collection" that has lasted for decades longer than the existance of the city of Amarna itself, involving the debate over the nature of rock patination in statues (Mansoor 1995).
Coming back to the Jehoash inscription, the analytical data supports the following interpretations:
Patination and weathering
Patina is the natural coating that is created over the rock surface due to the absorption or loss of various elements. It is commonly thought that the process of patination is slow; thus, genuine patina may be used as indication for the antiquity of an item. A vast amount of literature has accumulated over the past four decades concerning the processing, weathering, and dating possibilities of patina on archaeological petroglyphs (rock art), with special regard to sandstones (Bednarik 1979, 1994, 1996, with further references). From this data, it is evident that rock patination by itself is an unreliable indicator of antiquity since patina-like coating or even true patina can be created in the laboratory by various methods (see below). While one can readily accept that genuine patina formed over an engraved groove is younger than the design, there are several difficulties in evaluating its age. Climatic factors and pH have considerable influence and can accelerate, delay, or completely inhibit patina formation (Bednarik 1979:29-30). Moreover, the application of various dating techniques to the patina is dependent upon many factors (Bednarik 1996, with references) that will be surveyed below only with relevance to the Jehoash inscription. Therefore, the presence or absence of patina over a rock artifact cannot be used a priori as an indicator for its authenticity. In fact, it has been suggested that repatination of a groove is not so much a function of age but of groove depth and thickness of crust (Bednarik 1994:70-71).
In the creation of patina, two factors play a crucial role. The first is the composition of the substrate over which the patina is processed, its capillary, and absorbed water content. The second is the environment, namely the nature of the sediment, pH, temperature, and humidity that surround it as well as the physical conditions (such as aeolian erosion, granular mass-exfoliation, etc.). Obviously, an artifact having been subjected to a subterranean environment would develop different patination from an atmospheric one. Since the composition of patina is the result of reaction between the base material and the surroundings, it is expected to reflect in its composition some combination of the elements from the rock and its depositional environment.
The declaration that "no evidence was found that could indicate the dispersion of natural patina from a different stone to the plate or the letters" should be noted (Ilani, Rosenfeld, and Dvoracheck 2003:114). In Jerusalem, there are no geological formations containing any type of sandstone. As the Jehoash inscription was supposedly found in Jerusalem (a hypothesis adopted without any questioning by the GSI experts and used for augmenting their conclusions about the abnormal presence of gold in the patina), the analytical results make the option of burial completely impossible because the patina has only autochthonous minerals. In other words, it has absorbed nothing significant from its depositional environment that (in the case of Jerusalem) is very different from the stone. Since, in this case, one should expect the patina to be markedly different in composition from the stone, not similar to it, it raises the possibility that the patina was created out of another fragment of the same stone by methods that will be suggested below. The lithology around Jerusalem includes a set of Cenomanian - Turonian limestone, chalk, and dolomite series, typically capped by terra rosa or brown rendzina soils. This would result in significant enrichment of calcite in the patina as compared with the original stone. Moreover, several elements, especially those with high ionic charge or ionic radius, are more sensitive to post-depositional processes that may occur in buried artifacts due to their solubility in groundwater. Other elements (P2O5, SO3, and K2O) are enriched in archaeological sediments due to anthropogenic activities, especially in ashy layers such as those suggested by the GSI researchers to be the source of the carbon in the patina. Table 1 presents the calculated composition of the rock and the patina (obtained by two different methods).
Table 1: calculated composition of the Jehoash inscription rock and patina (after Ilani, Rosenfeld, and Dvoracheck 2003:114).
As can be clearly seen in Table 1, while the ICP-AES results show that the rock contains 11.1% CaCO3 (weight percent), SEM-EDS indicates that the patina contains only 15%, which means that it is enriched only by 3%-4%. In fact, this enrichment is lower due to the dilution factor of the Al2O3. Obviously, there is no addition of elements that are expected to come from an archaeological deposit, such as P2O5 or K2O. Given the fact that the patina is enriched with Fe2O3 (iron oxides) by 12% for reasons that will be explained later, there is actually no enrichment of calcite at all in the patina. Indeed, the GSI researchers use this as evidence for the authenticity of the patina. However, the interpretation should be completely the opposite, signifying that the stone has never been exposed to calcareous soil bearing anthropogenic features for a significant sequence of time. This means that the patina is unlikely to be created in the Jerusalem environment or, in fact, in any archaeological site in the Central Hill Country and beyond.
The examination of the patina cannot rely only on tiny, powdered or drilled samples. Natural patina would show, for example, microlaminar structure, and, therefore, it would be possible to distinguish it from a fake one if examined as an undisturbed cross-section (Wainwright and Taylor 1978). This method is capable of providing convincing minimum dates, but the severe restrictions on sample sizes prevent dating with the apparatus currently available. However, working in secrecy and not being allowed to take a minuscule section of the stone to analyze it is not the way to solve the question of authenticity of the artifact. In a research project that was dedicated to the provenance of clay cuneiform tablets, we managed to examine in thin sections minute samples, almost invisible to the naked eye (Goren, Finkelstein and Na’aman forthcoming). The same method could be applied also here.
The presence of gold in the patina
One of the arguments for the authenticity of the patina that covers the inscription is the anomalous number of pure gold globules, sizing one micron, that are embedded within the patina. Ilani, Rosenfeld, and Dvoracheck (2003:115-116) declare that such globules can be formed only by burning of pure gold at high temperatures. They further interpret this data based on Kings I: 6 where King Solomon is said to have gilded the inner section of the temple with fine gold. In their opinion, the gold globules were the result of the fire set by Nebuzaradan, the captain of the guard of Nebuchadrezzar, the King of Babylonia in 586 BCE. With tongue in cheek, one may say that if the gold globules should stem from the burned Salomonic temple, then the Inscription is much better to disprove the Bible than to prove it because according to Kings II: 24-25, the temple was thoroughly plundered by the Babylonians before burning.
This far-fetched explanation, given to such an anomaly that should have immediately given rise to suspicion, reflects the biased interpretation of the GSI researchers. If the gold microglobules were indeed the result of the strong fire that melted the gold-plated walls above or near the inscription, then we would naturally expect to find drops of gold and not one micron-sized globule. Moreover, for the obvious lack of any evidence for burial in soil, explained above, it is impossible that the patina would take only the gold microglobules and the charcoal from the sediment. Hence, one far less romantic, though stronger interpretation, could be that the anomalous number of pure gold globules under the patina is the result of a very shallow gold plating of the inscription that was made for sustaining its conductivity in order to apply the "patina" over its surface by electrolysis. This could result in the significant enrichment of iron in the patina, as indeed the analysis indicates. However, it is hard (though not impossible) to see how this could technically be done. Considering the reference to the temple, these globules could also be the result of gold treatment using a gas burner above the object in order to generate this linkage. This was most likely done prior to the coating of the inscription by the "patina," as we shall see below.
The numerous ambiguities that occur when carbon particles from petroglyphs are dated by AMS radiocarbon analysis are presented in detail by Beck et al. (1998) and supported by Dalton (1998). Their main conclusion is that "if a sample submitted for radiocarbon dating is found to contain two types of carbonaceous materials, each with a different radiocarbon age, then the analysis of the bulk mixture will not yield a reliable radiocarbon age. An apparent age can be determined, but this apparent age has no true age significance. Clearly, in these cases the bulk radiocarbon ages are ambiguous and do not represent the true ages of the samples" (ibid.: 2135). Just as serious is the plethora of general environmental variables that can significantly affect carbon isotopes. Among them is the hard water effect (the deposition of calcium and magnesium salts from aqueous solution in ground water), exchange with the atmosphere, humic acid, and especially the effect of the introduction of old carbon from a variety of sources (Bednarik 1996: 2-4). The greatest difficulty, however, is misinterpretation of results by archaeologists. In rock art research, it has been assessed that a piece of charcoal within a petroglyph has clearly no connection with its dating. When a motif yielded two different dates, it was seen by various authors as proving that repainting had occurred. In fact, Bednarik (ibid.) suggests several possible alternatives that are of relevance for our case:
a. The true age lies outside stated tolerances of one or both samples;
b. Charcoal fragments of different ages were used at the same time;
c. One or both samples are contaminated;
d. One or both samples provided erroneous results;
e. Or there could be any combination of some of the above factors.
In fact, the average age of the carbon samples that were dated by AMS radiocarbon dating is offset by half a millennium. Radiocarbon dating measures the date when an organism (a tree in this case) died, not the time when it was burned. The temple was burned at 586 BCE, but it was built at the 10th century BCE and perhaps restored at the 9th, so an average date of 300 BCE is much too late. However, this detail is insignificant since radiocarbon dating of charcoal from petroglyphs (or rock inscriptions) is too ambiguous anyway.
Given the suspicious composition of the patina and the presence of gold, one may suggest that charcoal from an Iron Age II stratum at some Israeli site has been mixed with the patina when it was created in order to make it "datable." This possibility could be easily tested by other laboratory methods (as we shall see below), but in its present state, it remains hypothetical. In any case, the radiocarbon dating from the charcoal that was embedded in the patina cannot serve by its own as evidence for the authenticity of the inscription since it could easily be "planted" to fool the scientists.
OTHER POTENTIAL METHODS
The ambiguity of the above results could be minimized and perhaps even settled by other, well documented, and easily available examination techniques. We shall survey below some of them in short:
The term "microerosion" refers exclusively to solution processes affecting the rock surface which can be seen only at the microscopic level. Microerosion analysis includes two related practices: the measurement of microscopic erosion on fractured crystals (Bednarik 1992) and the selective retreat in certain rock types of components that weather at vastly different rates (Pope 2000). While microerosion analysis is not thought to provide great accuracy, it is probably more reliable than most alternative methods of dating petroglyphs, and it is certainly cheaper and simpler than most. It can certainly distinguish between a newly polished surface (even if it was "eroded" by airbrush blown sand) and an old one.
Pure gold has a melting point of 10630C. When alloyed with other base metals, the melting temperature of the resulting alloy is somewhat reduced. It may be assumed that if indeed harsh fire caused the melting of gold over the stone, it must have been exposed to considerable heat. In the laboratory, we experimented with the resistance of similar sandstone (of the Shehoret formation) to heating in an electric kiln at 10500C for four hours. Apart from some color change, the rock revealed no cracks or signs of crumbling. However, the question as to whether the rock has been exposed to heat or not, which would examine the "temple burning scenario," could be easily tested by thermoluminescence (TL) dating. Amusingly enough, the GSI houses one of the most modern TL dating laboratories of its kind.
TL dating made its debut in archaeology primarily to help in estimating the ages of pottery remains. The use of its principles to determine when sand grains had last been exposed to sunlight is a more recent development (Smith et al. 1990). In the case of fired sandstone, this method can be readily applied.
Uranium/Thorium dating: An additional method that could be potentially used for age examination of the patina is the isotopic analysis of its uranium and thorium content (Thompson 1973). The uranium-thorium to helium-lead method has been previously used to date a small range of minerals, particularly magnetite (hence relevant for the case in question). The method involves various chemical purification processes and controlled re-precipitation. After final purging, the U and Th are plated onto a metal disc so that the activities of the various isotopes (230Th, 232Th, 234U and 238U) may be measured by alpha spectrometry. The recorded 230Th/234U activity ratio is then employed to determine the absolute age.
EXPERIMENTAL: REPLICATION OF THE FORGERY PROCESS
This section presents the results of an experimental simulation that was performed at the Laboratory For Comparative Microarchaeology, Tel-Aviv University, in order to demonstrate how, by the use of relatively simple equipment, a forgery that could yield the same analytical results can be produced.1
Monumental, royal stelae of the first millennium BCE (such as the Mesha Stele, the Tel Dan inscription and the Aramaic dedication inscriptions), were regularely engraved in basalt when a dark, majestic stone was desired. In this context, we may explain the abnormal choice of iron rich sandstone for the production of the forgery. Creating patina on a clay and iron rich, hard sandstone is by far easier. First, the inscription was easily carved on a pre-polished surface of the rock using iron tools that leave no traces of nickel, chromium, etc. (Fig. 1). At this stage the inscription looks "fresh" (Fig. 2), but it can be "aged" easily by blowing fine quartz (or carborundum, which may leave traces) on the surface using an airbrush system. This creates a surface that looks "weathered" and "old" even under the binocular stereomicroscope (Fig. 3). The patina is then created by crushing another fragment of the same rock in an agate mortar (to prevent contamination) and in an ultrasonic bath (to disaggregate the stone), then producing a watery solution of the powder. If gold traces are desired to be revealed by the anticipated chemical analysis which would be tied immediately with the reference to the temple gold in the text, it can be sprayed over the tablet using a gas burner. This would create micron-sized globules that will not be seen by the naked eye. Another possible way to fool the scientists would be to "plant" in the cracks and the grooves Iron Age II charcoal from an archaeological site, which can be easily obtained at any Near Eastern department of archaeology.
After the "hints" about time (datable charcoal) and space ("temple" gold) are embedded on the surface, the "time capsule" needs to be sealed (Fig. 4). The pasty solution of the ground rock is used for carefully coating the entire surface, including the inscription, and then let dry (Figs. 5,6). Due to the clay contents, it can be hardened to some extent by gradual heating in an electric kiln to about 3000C-4000C, which will not affect the charcoal. The "patina" may be "aged" even more by exposing it to microwaves which would result in the appearance of minute cracks and grooves within it.
It must be emphasized that, had we brought the little replica that we produced to the GSI and subjected it for the same set of analyses, it would produce exactly the same analytical results. The decision whether these results are enough to form an opinion about the authenticity of the object remains completely within the interpretational level. It has been shown above that different analyses (by thin sections and TL dating) could materialize more conclusive results.
A FINAL WORD OF CAUTION
It is evident that by bringing the Jehoash inscription to the GSI, the antique dealer or collector who owned it intended to obtain a "certificate of authenticity" in order to offer it to a museum for sale. Apart from the obvious ethical question as to why a governmental institute would agree to promote the financial interests of individuals who illegally possess looted antiquities, there is also the question of the responsibility of the scientist. Giving an objective, serious, well-considered, and well-substantiated "verdict" about the authenticity of an object is not only a matter of scientific integrity, but it also becomes a matter of legal responsibility when this opinion is being used by an antiquities dealer in order to convince a museum to buy the item for millions of dollars (as in this case). It goes without saying that in such cases scientists should better resist the temptation of sensational publicity and restrict their scientific work to items that are owned by formal institutions, if not to items that were discovered only in legal, well-recorded excavations.
The results of this experiment are published as a presentation on the World Wide Web at http://www.bibleinterp.arizona.edu/presentations/index.
I wish to thank Paul Craddock and Ian Freestone from the Department of Scientific Research of the British Museum, Chris Doherty from the Laboratory for Conservation and the History of Art, Oxford University, and Neil A. Silberman for their useful comments on various aspects of this paper.
Beck, W., Donahue, D. J., Jull, A. J. T., Burr, G., Broecker, W. S., Bonani, G., Hajdas, I. and Malotki, E. 1998. "Ambiguities in direct dating of rock surfaces using radiocarbon measurements." Science 280: 2132-9.
Bednarik, R. G. 1979. "The potential of rock patination analysis in Australian archaeology," Part 1. The Artefact 4: 14-38.
Bednarik, R. G. 1992. "A new method to date petroglyphs." Archaeometry 34: 279-91.
Bednarik, R. G. 1994. "A taphonomy of palaeoart." Antiquity 68: 68-74.
Bednarik, R. G. 1996. "Only time will tell: a review of the methodology of direct rock art dating." Archaeometry 38: 1-13.
Craddock, P. and Barker, N. 1990. Fake?: The Art of Deception. Berkeley: University of California Press.
Dalton, R. 1998. "Dating in doubt as researcher is probed." Nature 392: 218-9.
Goren, Y., Finkelstein, I. and Na’aman, N. Forthcoming. Inscribed in Clay I: Provenance Study of the Amarna Tablests and Other Related Near Eastern Texts. Tel Aviv.
Ilani, S., Rosenfeld, A. and Dvoracheck, M. 2003 (in press). "A stone tablet with an ancient Hebrew inscription attributed to Yehoash, King of Judea – archaeometry and epigraphy." GSI Current Research 13:109-116.
Mansoor, C. 1995. The Scandal of the Century: The Mansoor Amarna Expose. New York City.
Pope, G. A. 2000. "Weathering of petroglyphs: direct assessment and implications for dating methods." Antiquity 74: 833-43.
Siegel, J.A., Saukko, P.J. and Knupfer, G.C. 2000. Encyclopedia of Forensic Sciences (3 vols), "Document Analysis" in Vol. II, 556-597, San Diego: Academic Press.
Smith, B. W., E. J. Rhodes, S. Stokes, N. A. Spooner and M. J. Aitken 1990. "Optical dating of sediments: initial quartz results from Oxford." Archaeometry 32: 19-31.
Thompson, P. 1973. "Procedures for extraction and isotopic analysis of uranium and thorium in speleothems." McMaster University Geology Department Technical Memo 73/9:21.
Wainwright, I.N.M. and Taylor, J. M. 1978. "On the occurrence of a parallel pigment layer phenomenon in the cross-sectional structure of samples from two rock painting sites in Canada." In C. Pearson (ed.), Conservation of rock art, 29-31. Institute for the Conservation of Cultural Material, Ottawa.
Whittaker, J.C. and Stafford, M. 1999. "Replicas, Fakes, and Art: the Twentieth Century Stone Age and its Effects on Archaeology." American Antiquity 64: 203-214.
The above article refers only to the geoarchaeological analysis (and publication), and not to any previous epigraphic or philological work that concerns the Johoash inscription.