D 2036 Dating Methods (Absolute and Relative) in Archaeology of Art is of a phenomenon related to but not the actual Dating Methods (Absolute time of manufacture of the art. If these indirect ages and Relative) in Archaeology of Art are in a stratigraphic relation to rock art (older or younger), then they produce minimum/maximum Juan Francisco Ruiz1 and Marvin W. Rowe2 direct dates for related imagery (Bednarik 2007; 1Facultad de Ciencias de la Educacio´n Morwood et al. 2010;Ruizetal.2012). y Humanidades, Universidad de Castilla La Rock art research has been treated for years as Mancha, Cuenca, Spain a minor aspect of archaeology. Lack of reliable 2Texas A&M University at Qatar, Museum of methods to date ancient imagery, both New Mexico, Santa Fe, NM, USA pictographs and petroglyphs on open-air sites or inside of deep caves, kept it outside of main- stream archaeology. This began to change with Introduction the introduction of scientific dating approaches, and there are reasons to feel optimistic about Chronology of rock art, ranging from Paleolithic dating rock art at this time. Several dating groups to present times, is a key aspect of the archaeol- are currently working on this around the world, ogy of art and one of the most controversial. It and it is now possible to hope for interlaboratory was based for decades in nonscientific methods comparison tests to help evaluate the reliability that used stylistic analysis of imagery to establish and accuracy of the techniques. one-way evolutionary schemes. Application of scientific methods, also called absolute dating, started to be used in the 1980s and since then Key Issues/Current Debates/Future has increased more and more its significance, as Directions/Examples judged by the large number of papers published in the last two decades on this subject (Rowe 2012). Archaeological studies of rock art demand a temporal framework in which a particular imagery was produced, as it is the only way to Definition relate decontextualized imagery to archaeological cultures. The earlier traditional methods to estab- Absolute and relative dating methods have been lish chronologies of rock art sites and imagery used to establish tentative chronologies for rock were based on assumptions made on iconography, art. Relative dating refers to non-chronometric style, and comparison with excavation evidence methodologies that produce seriation based on sty- and technical analysis. For example, Paleolithic listic comparison and stratigraphic assumptions. mobiliary art from excavated sites in Europe was On the other hand, absolute dating methods are used as a base for stylistic comparison with cave based on scientific techniques that yield imagery. These evidences, supported by superim- a chronometric age for a phenomenon in direct or position analyses, produced the great stylistic indirect physical relation to rock art (same age, schemes for Paleolithic art in Western Europe, older, or younger). Dating of some binders in pic- which defended a one-way evolution from simple tographs or the alterations of surfaces by petro- to complex figures, expanding from Aurignacian glyphs are examples of direct ages related to rock to Magdalenian times (Pettitt & Pike 2007). These art production. However, it is controversial to con- systems lack enough resolution to produce an sider these dates as “absolute” as they merely accurate temporal frame for rock art, above all reflect experimental propositions, which often for styles without consensus on their mobiliary lack independent verification (Bednarik 2007; parallels. Pettitt & Pike 2007). Most scientific dating Weaknesses of these stylistic paradigms were methods are indirect because they produce pointed out (see (Bednarik 2007; Pettitt & Pike constraining ages for imagery, and the age obtained 2007) for recent reviews of them), but it is Dating Methods (Absolute and Relative) in Archaeology of Art 2037 D important to recognize that they are still useful Dating of charcoal pictographs has been for rock art chronology because it is obviously broadly used in French and Spanish Paleolithic impossible to date every figure in a site and every caves (Alcolea & Balbı´n 2007; Steelman & Rowe site all over the world. A well-defined proxy with 2012), but also in North America and Australia, stylistic, technical, and chemical composition and other regions of the world. The largest part of data would be very helpful as a complement to the charcoal pigment dates is considered reliable scientific dating. A date archaeologically but, for example, in Chauvet Cave, dates decontextualized is of little value, so it must be are controversial because they are unusually stressed that any dating should be included in old and conflict with stylistic paradigm. Several D archaeological hypotheses. authors claimed for a likely contamination of The first radiocarbon dating on rock paintings datings of Chauvet, as the dates from different was carried out on a charcoal pictograph in South samples of one single figure in Pen˜a de Candamo Africa in the late 1980s quickly followed by (Asturias, Spain) showed that results by others in 1990 in Australia, the USA, and Europe, Geochron Lab (USA) were 15,000 years younger which added to pioneer research on engravings than those produced by LSCE (France), respon- dating. A few years later, a broader conscience sible of all Chauvet dates (Pettitt & Pike 2007). about these new possibilities dictated that This situation reflects pitfalls of the method that scientific dating of the passage of time became could be accompanied by contamination of an alternative to stylistic paradigms (Lorblanchet unknown origin, possible repainting for younger & Bahn 1993). However, debates on very old dates, mistakes in laboratory treatment of sam- AMS 14C dates from Chauvet Cave (France) ples, and the presence of carbon of different ori- and very young ones on open-air engravings gin, for example, incomplete dissolution of dated by several methods in Foz Coˆa (Portugal) calcium oxalates. An improved specific protocol showed that style and scientific dates were still to remove contamination produced by calcium necessary for archaeology of art. oxalates from charcoal paintings has recently The most common technique for dating rock been developed (Bonneau et al. 2011). paintings worldwide is the radiocarbon dating of It is indispensable to follow a strict protocol to the charcoal pigments often used to construct the collect samples during fieldwork. The protocol drawings. A large number of publications have described in literature tries to avoid contamina- been collected in the bibliography composed by tions using sterile latex gloves and surgical Rowe (2012). Dating charcoal has been well masks. Samples are removed from walls with honed by the radiocarbon community, and the a sterile surgical blade, which is changed and results can be considered to be generally reliable. discarded after each sample. They are put inside The main disadvantage to radiocarbon dating of a sheet of folded sterile aluminum foil and charcoal pigments is that the date measured is placed inside of a labeled plastic bag. The exact NOT that of the time of execution of the painting. position where samples were removed should be Rather it dates the pigment and there are two recorded with photographs. Extreme care caveats that accompany any date of charcoal: ought to be observed on the selection of the old wood and old charcoal (Steelman & Rowe sampling points to avoid major visual impact or 2012). Old wood phenomena are situations, usu- harm to the pictographs, for example, selecting ally encountered in desert areas where wood flakes that appear likely to spall from the walls decays slowly, in which the wood burned to naturally. The size of samples required is uncer- make charcoal may be up to centuries old. Old tain but around 2 cm2 is generally used for charcoal may occur when freshly hewn wood is pictographs with inorganic pigments and much burned, but not used to construct a painting until less for charcoal-pigmented paintings, as for much later. Both these caveats should be kept in AMS 14C dating only 50–100 mg of carbon is mind at all times when interpreting charcoal pig- needed for an accurate date. For pictographs ment radiocarbon dates. with inorganic pigments, e.g., iron ochre or D 2038 Dating Methods (Absolute and Relative) in Archaeology of Art manganese oxides, it is essential to take an It has been shown to be useful to get minimum unpainted rock sample as near to the sample ages for petroglyphs and minimum/maximum taken as is feasible. That background rock sample ages for pictographs. On certain locations, should be processed identically to the pictograph researchers has bracketed dates for rock paintings sample. A more detailed report on sampling between two oxalate skins, producing a temporal protocol, reporting of radiocarbon results and frame for pictographs in agreement with archaeo- laboratory pretreatment of samples, has been logical expectations (Ruiz et al. 2012). just published (Steelman & Rowe 2012). Sample removal procedure is similar to AMS 14C has been used to date any other kind that described for radiocarbon dating (Cole & of carbon-bearing substances related to picto- Watchman 2005). Sample sizes range from graphs or petroglyphs (Aubert 2012). The 25 mm2 to1cm2, depending on oxalate content. presence of binders has been used to produce Oxalate dating demands microstratigraphic direct radiocarbon dates of beeswax paintings in analysis and micro-excavation techniques to Australia (Morwood et al. 2010). Vegetal resins avoid contamination between upper and lower and wax are binders of these paints. It is consid- layers of calcium oxalate. Mechanical proce- ered that wax would have been fresh when duresandlaserablationhavebeenusedsofar applied on walls to construct the drawings, so it for this purpose (Watchman 2000). There are should be ambient source of carbon to date rock two main drawbacks for oxalate dates: (1) radio- art. In the Australian Kimberley area, a range of carbon age of any calcium oxalate crust is dates from 3,780 Æ 60 BP to present times were a weighted “average” of oxalate deposited for obtained (Morwood et al.