The Oxford Research Encyclopedia of African History will be available via subscription on April 26. Visit About to learn more, meet the editorial board, learn about subscriber services.

Dismiss
Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, AFRICAN HISTORY (oxfordre.com/africanhistory). (c) Oxford University Press USA, 2019. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 26 April 2019

Scientific Dating Methods in African Archaeology

Summary and Keywords

The African archaeological record is particularly remarkable in that it covers timescales relevant to all human history and prehistory. Different dating techniques are therefore fundamental to constructing reliable chronologies for the continent. The principal factors that determine the usefulness of a dating technique are (1) applicability to the material in question, (2) the expected precision of the technique, and (3) the age range over which it is expected to be useful. Radiocarbon is applicable to the past fifty thousand years of human history, encompassing the Later Stone Age, Iron Age, and historical periods, and is a highly-refined method applicable to organic materials such as bones, plant matter, charcoal, teeth, and sometimes eggshell. However, African archaeological contexts often present challenges to the preservation of material, and it is important to establish the context of the material under investigation. Materials of preference for radiocarbon dating, such as plant cellulose, are thought to be resistant to alteration during burial (diagenesis). The age ranges of luminescence and uranium-series dating stretch well into the African Middle Stone Age. Luminescence dating is applied to sediments and burnt objects, and uranium-series (U-series) dating is applied to geological materials such as carbonates and stalagmites. In some special cases, U-series dating can also be applied to fossil bones, teeth, and eggshell. For all dating methods the importance of context cannot be overstated. Other techniques, such as archaeomagnetic dating and rehydroxylation (RHX) dating, should be applicable over the historical period, but these new methods are under development. Dating methods are an active area of interdisciplinary research, continuously refined and developed, and collaboration between African archaeologists, geologists, and dating specialists is important to establish accurate regional chronologies.

Keywords: scientific dating, radiocarbon dating, African archaeology, Quaternary period, U-series dating, luminescence dating, archaeomagnetism, tephrochronology

The Importance of Dating Methods in African Archaeology

Dating methods are central to African archaeology. The continent has the longest continuous record of human occupation, and its rich archaeological heritage stretches across timescales that encompass the historical period, over the past few hundred years, to the precolonial period, to the emergence of farming over the past several thousand years, and ultimately the evolution of humans over hundreds of thousands of years. There is therefore abundant need for a range of analytical dating methods, applicable to a wide variety of archaeological contexts and materials (see figure 1). By far the most important and widely used of these techniques is radiocarbon dating. Applicable to a broad class of organic (previously living) materials, and relevant over much of modern human history and prehistory, radiocarbon dating is a powerful technique that has driven several “radiocarbon revolutions” in archaeology since the original development of the method in the 1950s.1 In fact, one of the oldest high-precision radiocarbon laboratories in the world was established in 1969 in Pretoria, South Africa; it was instrumental in making considerable improvements to the development of radiocarbon dating as a technique, and helped establish a chronological framework for much of African archaeology during the 1970s and 1980s, especially in southern Africa. Radiocarbon dating has been much refined since the mid-20th century, and it is a highly reliable and much utilized tool. That said, since the turn of the 21st century, the method has been so easily applied to archaeological contexts, frequently with only minimal oversight by radiocarbon or chronology specialists, that there is abundant scope for misapplication. However, numerous excellent published guides for archaeologists and ecologists can help maximize the utility of radiocarbon dating methods.2

Scientific Dating Methods in African ArchaeologyClick to view larger

Figure 1. Age range of dating techniques in African archaeology. The asterisk (*) indicates that the method is regarded as developmental. Adapted from Martin Jim Aitken, Science-Based Dating in Archaeology (New York: Routledge, 2014).

Dating methods are themselves an active area of research, constantly under development and refinement. The field is often a highly technical collaboration between applied physicists and chemists who develop and refine methods of pretreatment and measurement, and archaeologists and historians who supply the materials.

Here, a brief overview of the principles of the various dating methods applicable to African archaeology is presented, and the present status of each technique is reviewed. The importance of the context of dated materials cannot be overstated, and establishing the context is vital to ensure accurate dating. This is as important for radiocarbon, which is applied to the Later Stone Age, Iron Age, and historical periods, as it is for other techniques that are more regularly applied to older Quaternary settings, often in the context of human evolution, such as optically stimulated luminescence dating (OSL) and uranium series dating (U-series).

Methods

Radiocarbon Dating

Fundamental Principles of Radiocarbon Dating

Radiocarbon, or 14C, is a radioactive isotope of carbon, present in all living things on earth, albeit in tiny amounts (only about 1 in 1012 carbon atoms is 14C). It is predominantly produced in the upper atmosphere, when atmospheric nitrogen interacts with incoming high-energy cosmic rays, and is rapidly distributed throughout the atmosphere and incorporated into the global carbon cycle alongside stable isotopes 12C and 13C, largely via plant photosynthesis. The natural production of radiocarbon by this mechanism is fairly constant, but there are small fluctuations in production, attributable to modulation by both the solar wind and the geomagnetic field, which can affect the global amount of atmospheric radiocarbon available for uptake in a given year. Luckily, we can reconstruct these variations by measuring the concentration of 14C in tree rings, corals, and other annually laminated materials of independent known age. The procedure for accounting for these fluctuations in known as calibration (described below), which converts a radiocarbon measurement into a calendar age.

Radiocarbon dating is based on the decay of the 14C atom. An excess two neutrons (eight) over the number of protons (six) in the nucleus causes the radiocarbon atom to be unstable. This instability induces the β‎ decay of one excess neutron, via the production of a negatively charged electron, and converts radiocarbon to nitrogen (146C → 147N + e). Crucially, the rate of radiocarbon decay is constant, and so upon death, when newly produced radiocarbon is no longer incorporated into a living organism, the steady tick of radioactive decay diminishes the supply of radiocarbon in the material. During the early development of the radiocarbon dating method, the radiocarbon half-life (the time for radiocarbon to decay to half of the initial amount) was measured as 5568 years.3 The half-life, together with the precision of measurement, account for the practical upper limit of the radiocarbon method. Older materials contain fewer 14C atoms, and consequently precise radiocarbon dating becomes impossible at timescales more than 50,000 years before the present day (roughly 8–9 half-lives).

The very first radiocarbon dating techniques measured the rate of β‎ emissions within a sample. These techniques can produce very precise age estimates, but they require large sample mass to produce sufficient numbers of β‎ decay events. Consequently, very small or ancient samples require unfeasibly long counting times (several months) to produce precise age estimates. Using modern accelerator mass spectrometry (AMS) methods, samples can be much smaller than β‎-counting techniques, and analysis time has been greatly reduced. AMS methods have several important implications for archaeologists. Small or precious samples, which previously would have been dated only by association with a stylistic or stratigraphic sequence, can now be directly dated using AMS. Previously, small amounts of excavated organic material, such as charcoal fragments, were often aggregated to form larger samples, potentially mixing material from different organisms or even across stratigraphic units. AMS methods now allow individual tiny fragments to be dated. However, a well-defined relationship is needed between the organic sample and the archaeological event, and care should be taken during excavation to rule out the potential movement of small remains that may have occurred during burial.

Calibration of Radiocarbon Dates

An essential step in translating a radiocarbon measurement into an age estimate is the process of radiocarbon calibration. A radiocarbon measurement is converted into a calendar “date” (typically reported as bp, “before present,” where present is 1950 ce), with an associated measurement error, usually presented as ± one standard deviation (1σ‎). It is important to recognize the statistical nature of a radiocarbon measurement, and the uncertainty is integral to the radiocarbon date. Calibration curves, the most recent of which are IntCal13 for the Northern Hemisphere and SHCal13 for the Southern (to account for a very slight offset between the two hemispheres), rely on numerous lines of independently dated evidence to estimate past 14C levels, including tree ring sequences, annual varves from lakes, and deep-sea sediment cores.4 Although the computational methods of filtering the normal distribution of a radiocarbon measurement through the nonlinear calibration curves can be complex, there are fortunately several open-access software programs designed for radiocarbon calibration (OxCal, Calib, CalPal). Such programs present a time range within which the date estimate is consistent, and which should be reported in full as the calibrated range (often on the cal bp scale or a journal-specific nomenclature). Due to “wiggles” and “plateaus” in the calibration curves, a single date may equally correspond with separate periods, which can be cumbersome to report, but which reflect the calendar age most accurately. Plateaus in the calibration curve can sometimes limit the precision of radiocarbon dates in certain periods. One method of tackling such limitations requires the measurement of multiple samples, whereby the overlapping age ranges, in combination with stratigraphic and independent age data, can be statistically modeled in specialist open-access software (including OxCal, BChron, BCal, Bacon) using Bayesian principles to narrow the likely age range (see figure 2).5

Scientific Dating Methods in African ArchaeologyClick to view larger

Figure 2. Two different calibrations for a single radiocarbon date for charcoal from the Iron Age site of Thulamela in South Africa. Vogel (2000) reports the radiocarbon determination as follows: Pta-7243, 520 ± 40 (1σ‎) 14C years bp. On the left, the radiocarbon determination is calibrated with the atmospheric curve IntCal13 (blue curve), which leads to an incorrect date. On the right, the same determination is calibrated with SHCal13 (green curve, Hogg et al., 2013), which is the most appropriate calibration curve for the Southern Hemisphere. Both dates were calibrated using the OxCal software, version 4.3.2. The inappropriate calibration curve (IntCal13, left) results in subtle differences in the calibrated age ce, including a second possible region of the 95.4 percent probability range at 1310–1360 ce. Illustration by the authors.

In the past there has been a trend in southern African archaeology to leave 14C dates uncalibrated, a practice that is now out-of-date considering recent advances in the accuracy of the Southern Hemisphere calibration curve and/or corrections for marine reservoir offset. Omission of proper calibration will lead to inaccuracies in chronology, sometimes by as much as several hundred years, and should no longer be considered an appropriate approach to African prehistory. For any radiocarbon date, four quantities should always be reported: (1) the raw, uncalibrated 14C date, (2) the uncertainty ±1σ‎, (3) the calibrated 14C date (in cal years bp or bce) and (4) the confidence level (either 68.2 or 95.4 percent). The archaeologist should also make it clear which calibration curve has been used (e.g., SHCal13, IntCal13, or an appropriate Marine curve with reservoir offset).

Reservoir Effects and Their Importance for Radiocarbon Dating of Marine Materials

Although newly produced radiocarbon mixes very rapidly through the atmosphere, it mixes more slowly through the global oceans. Therefore, radiocarbon dates from marine contexts typically appear several hundred years too old, with the exact offset depending on regional oceanographic dynamics such as upwelling systems.6 This marine reservoir effect can be a considerable problem for dating archaeological remains in coastal settings, and materials such as marine shells or unidentified bone fragments (if potentially from seals, fish, etc.) should be avoided. The radiocarbon age of terrestrial animals or humans is also affected by the reservoir effect if marine-derived foods are consumed. The effect is proportional to the amount of marine food in the diet, which can be difficult to establish with any accuracy.7 Other radiocarbon reservoir effects exist in diverse environmental settings, especially where there is a considerable contribution from geological carbonates that are radiocarbon “dead,” such as in lakes or limestone cave systems, or due to behaviors of organisms such as snails. Ultimately, the source of carbon needs to be carefully considered in any dated material.

Anthropogenic anomalies have arisen with the widespread burning of 14C-depleted fossil fuels since the Industrial Revolution (the Suess effect) and testing of nuclear technologies in the mid-20th century have increased 14C production in the atmosphere (the bomb spike). These atmospheric effects will thus affect material formed after their inception. However, both of these effects have been well understood for most of radiocarbon dating’s development and are typically corrected for upon measurement.

Contamination and Material Requirements for Radiocarbon Dating

Incorporation of either younger or older radiocarbon is a central concern with radiocarbon dating. The effect is known as contamination, and it can arise during excavation, subsequent handling, or alteration in the burial environment (diagenesis). Radiocarbon dating of archaeological bones is entirely dependent on the adequate preservation of collagen8, for which the conditions are comparatively poor across most of Africa. Certainly, beyond a few thousand years, collagen is typically preserved only in caves and rock shelters or other highly stable depositional environments. Dating of bone apatite has long been considered unreliable and requiring rigorous pretreatment, given apatite’s susceptibility to post-depositional diagenesis. Some dates on apatite are published in older literature, and can indicate unfeasibly ancient ages—for example, dates of livestock older than 5000 bp in East Africa have subsequently been disregarded.9 Charcoal, more resistant to degradation, is thus one of the most commonly dated materials in African archaeological sites. The issue of in-built age needs to be considered with charcoal samples, where the age of the wood that formed the charcoal may be considerably older than the archaeological event. This is more of a concern in old-growth forests or hyper-arid deserts, where long-dead wood can preserve on the open landscape, than in grasslands or woodlands. Moreover, the small sample requirements of AMS methods allow for the selection of recognizably twiggy fragments, which are likely to have been burned very soon after growth. Another common material for radiocarbon dating in African contexts is ostrich eggshell, as the species is widely distributed and eggshell preserves well. However, ostrich hens consume relatively large amounts of geological carbonate during the laying season to meet their calcium requirements—this has been suggested to make the age of a modern egg appear a couple of hundred years too old, but the effect is relatively poorly constrained by experimental studies.10 In older archaeological contexts, the impact of this small reservoir is outweighed by the advantages of an otherwise largely reliable dating material, but it will be of greater concern for more recent periods.11 Marine shells also preserve well and are frequently the most abundant organic material in coastal sites, but dating of marine shell is problematic for several reasons, including the uncertainty related to the local marine reservoir effect, and how it changed through time. A further concern is the susceptibility of carbonates to diagenetic alteration, whereby the original carbonate dissolves and recrystallizes, potentially incorporating carbon atoms from the depositional environment in an unpredictable way. Such diagenesis can be difficult to detect visually and should be evaluated using analytical chemistry approaches to eliminate unsuitable samples prior to dating.12 Another problem, possibly exacerbated by limited funding and expertise in some regions, pertains to the curation and storage of samples, which may be susceptible to microbial attack after excavation. Samples suspected to be affected require careful examination and aggressive pretreatment strategies.

To tackle contamination, where extraneous carbon is incorporated into a sample after death, radiocarbon laboratories employ several cleaning and chemical pretreatment methods that are crucial to acquiring a reliable date and in extending the limit of the radiocarbon method. Glues, varnishes, or inks applied for conservation purposes must firstly be removed using chemical and physical cleaning methods. Samples are commonly abraded or sandblasted to remove the outer layer likely to be most affected by any contamination or diagenetic processes. ABA (acid-base-acid) cleaning methods are still widely applied to a broad range of materials including bone, charcoal, wood and sediments. An ABA treatment consists of an initial acid wash to dissolve secondary carbonates, a base wash to attack secondary organic components such as humic acids, and a final acid wash to neutralize dissolved CO2 introduced by the base wash.13 Variants on this method include ABOx-SC (acid-base-oxidation-stepped combustion) for old charcoal samples and ultrafiltration for bone collagen, each of which have been shown to produce older dates than other pretreatment methods, suggesting that they remove more contaminating young carbon.14 Advanced pretreatment methods may employ aggressive chemical strategies that cause much of the original sample to be lost, but given the tiny amount of carbon ultimately required for measurement by AMS instruments, the larger samples necessary to attempt these pretreatment methods are still acceptable in many circumstances. The development of more targeted pretreatment strategies, whereby single organic compounds known to be present in the original sample (e.g., the amino acid hydroxyproline in bone collagen) but unlikely to be found in any depositional contaminant, continues apace with considerable success.

Uranium Series Dating and Its Applicability to African Contexts

Radiocarbon dating is arguably the most widely employed dating method in Africa, but it should be noted that there are other important radiometric methods that are relevant to African archaeology. In particular, uranium-series dating (also “U-series,” or “uranium-series disequilibrium dating”) is one such family of chronological techniques, particularly useful for materials older than the range of radiocarbon dating (50,000 years). This family of techniques is based on two radioactive decay chains of the naturally occurring isotopes of uranium, 238U and 235U. The parent isotope 238U ultimately decays via a chain of daughter radioisotopes, including thorium-230 (230Th) to the stable isotope of lead 206Pb, with a half-life of around 4.47 billion years. The other parent isotope, 235U, ultimately decays to a different stable isotope of lead, 207Pb with a half-life of 704 million years. There are two main types of U-series dating: U-Th dating, and U-Pb dating. U-Th dating is based on the decay series 238U→234U→230Th, and is used to date material from a few hundred years old to more than 500 million years. However, instead of accumulating indefinitely, the daughter isotope 230Th itself decays, so the U-Th dating technique is ultimately limited by the half-life of 230Th (75,584 years), which means that after a certain number of years 230Th decays as fast as it is produced. When this happens, the system is said to be in a state of secular equilibrium. The other type of U-series dating is U-Pb. This technique is based on measurement of both 238U and 225U decay series, and their ultimate stable daughter products 206Pb and 207Pb. The U-Pb dating technique is mostly applied to materials older than 1 million years.15

Uranium is naturally incorporated into most organisms and geological materials. In these contexts, the uranium series is said to be either an open system or a closed system. An open system is where all the products of a uranium decay series remain in the system, or when uranium itself has migrated to or from the sample during burial. A closed system is where there is has been no migration and all the decay products of uranium remain in material. A key assumption of U-series dating is that the system is closed, or the open system behavior can be accounted for. This requires detailed analysis of the geological history, taphonomy, and stratigraphy of a particular site, as well as complementary isotope measurements on the material under examination. Closed system behavior is essential for U-Th dating of speleothems, tufa, and calcrete.16 It is sometimes established in coral and eggshell, but fossil teeth, bone, and invertebrate shell are more often than not open systems, and these materials are therefore more difficult to date by U-Th dating. If possible, 14C should rather be used for shell, teeth, bones, etc. over the past 50 thousand years.

U-series dating requires specialist clean facilities, to avoid contamination with lead and other metals. The technique is often highly accurate, achieving 1 percent precision in many cases during the last interglacial period. U-Th dating is especially accurate over this period, particularly in speleothem contexts. The introduction of multiple-collector inductively coupled plasma mass spectrometer (MC-ICPMS) techniques, which can simultaneously measure several isotopes in the uranium series, and require small sample mass between milligrams and grams, have greatly improved the precision and practicality of the technique. By convention, U-series ages are always reported with 2σ‎ uncertainties and as the number of years (in units a or ka) before the time at which the measurement was made.

Luminescence Dating and Its Methodology

Luminescence dating encompasses a suite of related techniques that rely on similar principles to calculate the time since common minerals such as quartz and feldspar were last exposed to heat or light. These minerals absorb radiation emitted from natural environmental sources at a quantifiable rate. The absorbed energy is stored in minuscule defects within the crystal lattice of the mineral, which act as electron traps, and continues to accumulate as long as the mineral is shielded from sunlight or high temperatures, or until the traps are completely filled or “saturated.” When the minerals are re-exposed to light or heat, or “bleached,” the trapped electrons are released as luminescence, a light emission far too faint to see by eye, but detectable in a laboratory.

The technique is broadly split into two categories: optical dating and thermoluminescence (TL) dating.17 Optical dating, where stored energy is released by sunlight, actually comprises several methods, including optically stimulated luminescence (OSL), thermally transferred OSL (TT-OSL), infrared stimulated luminescence (IRSL) and post-infrared IRSL (pIRSL). Thermoluminescence methods, where the stored energy is released at high temperatures, are widely applied to fired ceramics and other heat-treated material, such as burnt flints and iron-smelting furnaces.18 Luminescence dating methods have much older age limits than does radiocarbon dating, potentially extending beyond 1 Ma under the right conditions, although typically back to c. 200 ka. Moreover, poor preservation conditions for organic materials in humid tropical regions make luminescence a valuable alternative to radiocarbon dating, even in comparatively recent contexts.

An OSL age is composed of two measurements—the equivalent dose (De, equivalent to the “palaeodose”) represents the stored energy that is released from the sediments in the laboratory, and this is divided by the rate at which the energy accumulates, the environmental dose rate (DR), an estimate of the annual radiation experienced by the mineral grains that is unique to each depositional context. The equivalent dose, presented in units of grays (Gy), is measured by stimulating ultraviolet radiation (using blue/green light for quartz, or infrared for feldspars), which is captured using highly sensitive photomultiplier tubes in specialist darkroom facilities. Various laboratory procedures, such as repeated heating steps in conjunction with exposure to variable light intensities, can be employed to facilitate the release of stored energy.

The environmental dose rate DR comprises all sources of radiation, including alpha, beta, and gamma radiation and cosmic rays, which have affected the sample since the last emptying of the electron traps. This radiation largely derives from radioisotopes of uranium, thorium, and potassium in soils and underlying geology, and it can be very challenging to accurately estimate the various radiation sources at a sufficiently resolved spatial scale. These sources are measured using different techniques, for example, laboratory measurement of a second sediment sample, or field measurements using portable radiation detectors or buried dosimeters, which are placed in the original sample location. Various radiation sources affect parts of a mineral grain differently, and environmental fluctuations may have altered the dose rate through time (e.g., by the leaching of uranium out of sediments). Changes to the water content of the sediments through time also need to be estimated, and all of these calculations require meaningful error estimates that can be propagated through to the uncertainty in the luminescence age. A working rule is that the uncertainty provided by OSL techniques is usually about 10 percent of the age (and is often higher for TL measurements), but uncertainties in dose rate can mean that greater uncertainties are more realistic.

OSL Methodologies for Quartz and Feldspar Grains

Quartz grains, the most commonly targeted mineral for OSL dating, are susceptible to low saturation thresholds and low sensitivity. This means that they are typically useful as a dating tool back to only about 200 ka.19 In contrast, feldspar grains, especially those with high potassium content, have the potential to record much older ages, but feldspar is subject to a phenomenon known as “anomalous fading” whereby electrons leak out from the traps, and therefore provide ages that are too young. Methods have been developed to correct for this problem only back to c. 50 ka, and while there is active research into fading-resistant traps, these traps are less easily bleached by sunlight and so may yield inaccurate ages if not properly reset by exposure to light. Recent development of infrared radiofluorescence (IR-RF) techniques applied to potassium-rich feldspar minerals is particularly promising for contexts older than c. 200 ka, as K-feldspars are not subject to the anomalous fading constraints of other feldspar minerals.20

Early OSL techniques measured the bulk luminescence signal across multiple subsamples, or aliquots, altogether comprising many thousands of grains, and combined these measurements to derive an equivalent dose. However, this approach assumes that each grain is equally saturated upon last exposure. If the sample is mixed, and grains were not equally saturated upon last exposure, then the OSL age will be an amalgamation of ages over which the grains were last exposed to light. A major breakthrough was the development of the single aliquot regenerative dose (SAR) protocol for quartz and feldspar, which incorporates sensitivity tests that can correct for these effects.21 The SAR protocol can be extended to single mineral grains, whereby the equivalent dose is measured independently from dozens of grains, each a separate aliquot to be incorporated in or excluded from a combined age estimate. This procedure allows for the dating of sediments with more complex formation histories that may include grains from different sources.22 Single-grain OSL dating is often undertaken in combination with detailed sedimentological assessments, and has become an especially important method in archaeological sites, which are susceptible to post-depositional disturbances. While several hundred grains are typically measured, only a relatively small number are particularly suitable for OSL dating, because of the unique properties of each grain, and a large proportion of grains will be excluded from analysis. Statistical models are sometimes deployed to distinguish and combine individual single-grain measurements, but correct implementation of such approaches requires some understanding of the site formation and depositional processes.23

Archaeomagnetic Dating, Rehydroxylation (RHX) Dating, and Tephrochronology as Developmental or Complementary Techniques

Radiocarbon, U-series, and luminescence are the most widely used dating methods in African archaeology, but there are a number of analytical dating methods that are less often used. This may change in the future as techniques develop and their practicality and accuracy are improved. In some cases (e.g., archaeomagnetism, rehydroxylation) these techniques are still an active research area. In others (e.g., tephrochronology) the technique is well established, but the dated material is relatively rare in archaeological contexts.

The Potential for Archaeomagnetic Dating in Africa

Materials made of clay such as pottery, kiln structures, and burnt hearths often acquire a magnetization at the time they are fired. This magnetization is usually carried by magnetic minerals within the fired clay matrix, such as magnetite or hematite. When these minerals are heated to above their respective Curie points (around 570–680oC), and allowed to cool back to ambient temperatures, they acquire a remanent magnetization that is proportional to both the direction and strength of the local magnetic field at the time of firing. The Earth’s magnetic field changes over timescales ranging from tens to thousands to hundreds of thousands of years (full reversals). Changes in direction and intensity of the geomagnetic field on timescales from tens to thousands of years are called secular variation. Archaeomagnetic dating generally relies on secular variation in the direction of the Earth’s magnetic field, but sometimes intensity changes are also used. The directional signature of the geomagnetic field at a particular point is measured as declination and inclination. Declination is the angle between geographical north and the geomagnetic field. In Africa in 2018 declination ranges from about 4o to −28o, with strongest variation in southern Africa. Inclination, or dip, is the angle made with the horizontal by geomagnetic field lines. It ranges from −90o at the geomagnetic South pole to 90o at the geomagnetic North pole, with 0o near the equator (where magnetic field lines are parallel to the Earth’s surface).

To perform archaeomagnetic dating one first needs to have a good understanding of past secular changes in the geomagnetic field. A regional secular variation curve is required to date archaeological objects from a particular area, because these changes will be different in each region. To build up the required temporal resolution for a regional secular variation curve, it is first necessary to perform archaeomagnetic analysis on many features, independently dated by 14C or other stratigraphic constraints. Only once the secular variation curve is constructed can the archaeologist use the declination and inclination to date a burnt feature of unknown age. Another requirement for this dating method is that the burnt feature is preserved in situ without significant post-depositional disturbance that can change the direction of the preserved magnetic signal, leading to incorrect ages.

Archaeomagnetic dating is a highly interdisciplinary field, with archaeologists and geophysicists teaming up to ensure that each burnt feature is properly preserved, sampled, and measured. To reconstruct ancient magnetic directions preserved by burnt materials, both extensive field measurements and laboratory experiments are required. In the field, the samples are orientated in situ, and a record of this orientation is preserved on each fragment. This orientation allows the geophysicist to properly prepare several small (1–4 cm3) orientated samples for subsequent measurement in a laboratory. Furthermore, to improve the accuracy of the technique it is common to collect several large fragments, from which multiple subsamples are measured. The results are then averaged into single values for declination and inclination, using Fisher statistics.24 The declination and inclination are then matched with the known secular variation curve, from which an age is inferred for the material.

In the laboratory, directions are measured by progressive heating until the Curie point. The remanent magnetization of these materials is relatively weak, and the directions are sometimes difficult to measure. It is therefore common to use sensitive cryogenic magnetometers, located in special magnetically shielded rooms. To ensure that the magnetic remanence is properly preserved, tests (including low-field susceptibility versus temperature and magnetic hysteresis experiments) will often be performed to identify the mineral carrier. Despite the large, costly, and time-consuming requirements, archaeomagnetic dating can yield very precise chronologies (e.g., down to a decade or two) if measurements are conducted carefully, and accurate regional secular variation curves are obtainable. Such curves have been refined in Europe for several decades, Western Europe in particular.25

Scientific Dating Methods in African ArchaeologyClick to view larger

Figure 3. “Stereonet” plot of secular variation in the magnetic field inferred at the location of Mapungubwe, South Africa, between c. 350 ce and the present day (shown by the star). The curve is constructed from archaeomagnetic analysis of ten independently dated Iron Age burnt structures (Hare et al., 2018). Direction of green arrows shows the evolution of the field. Ninety-five percent confidence ellipses for each site direction. Illustration by the authors.

However, archaeomagnetism in Africa has recently been rapidly expanding (see figure 3). Geophysicists are increasingly interested in the region because of its proximity to an important geomagnetic feature, the South Atlantic Anomaly, which is a rapidly weakening region of the Earth’s magnetic field. Archaeomagnetic dating should now in principle be possible for the southern African Iron Age (from c. 350 ce onward) using data from Hare et al., and an appropriate numerical procedure26, although further data are needed to better refine the chronology.26 It will not be possible to perform archaeomagnetic dating in other regions until more directional data are obtained from west and east Africa for the construction of proper secular variation curves.

Rehydroxylation (RHX) as a Potential Dating Technique for Pottery and Burnt Clays

The rehydroxylation (RHX) dating method is based on the phenomenon of mass gain and moisture expansion in fired clay ceramics. When fired to moderate temperatures (500–600oC), common clay minerals, such as those found in pottery, begin to lose structural hydroxyl (OH) groups. Upon removal from the kiln, the ceramic begins to chemically recombine with atmospheric moisture, and structural hydroxyl groups are once again chemically bonded to vacant sites in the ceramic matrix. This slow rehydroxylation reaction continues over the lifetime of the ceramic, causing the material to expand and gain mass. The older the ceramic, the more hydroxyl groups are chemically bonded to sites within the clay mineral matrix. If the amount of chemically bonded hydroxyl groups can be measured, along with the rate of this reaction, then the ceramic provides an “internal clock” that can be read to determine the elapsed time since it was last fired. Some authors have reported uncertainties of <1 percent of the age of the ceramic using this method, but this is almost certainly unrealistic, and the technique can probably achieve no better than about 5 percent, in the most ideal cases.27 Several authors have encountered significant analytical difficulties with the technique, particularly in cases where significant amounts of organics are present in the ceramic. Additionally, there is uncertainty over the underlying mechanisms of rehydroxylation, and how this reaction varies with underlying clay mineralogy. These relationships are not yet sufficiently well understood to allow robust dating of several different ceramic types. Efforts are currently under way to address these issues, and this dating method may ultimately move away from gravimetric protocols toward IR spectroscopy or similar techniques. Because of these difficulties, the rehydroxylation dating technique is currently regarded as under development.28

Tephrochronology as a Complementary Dating Technique

Tephrochronology is a method based on using geochemically well-characterized deposits of volcanic ash, correlated with an eruption of known age, to form a powerful chronological framework that links different sites and different regions. It is therefore not so much an analytical technique based on a rate of decay, or a change in a quantity through time, but rather a technique based on correlation. Ideally, the volcanic ash is released in a single eruption, and transported over an area of thousands of square kilometers. Most tephras have a unique geochemical fingerprint that is associated with a particular eruption. Tephra is not usually directly dated itself, although its age can be established by a combination of conventional methods such as fission track dating, 40Ar/39Ar dating, U-series dating, or stratigraphic constraints imposed from 14C and OSL dates. Rather, it is an age-equivalent technique whereby the presence of this particular tephra horizon in other peat sequences, lakes, aeolian sediments, and archaeological sites allows the sequence to be dated indirectly. And given that the eruption and deposition of the tephra is fairly rapid, within the space of a year or so, tephrochronology is often highly precise.

An advantage of the dating technique is that a tephra layer is often relatively easily identified in sediments and other excavation sequences, but its presence depends on the proximity to the volcanic source. In recent years, much research has focused on cryptotephra, which is very fine volcanic glass, invisible to the naked eye: being lighter it is more widely dispersed during an eruption.29 Geochemical fingerprinting of tephra and cryptotephra is usually established using ion microprobe or laser ablation inductively coupled mass spectrometry (LA ICP-MS) measurements. Potential problems with the technique include miscorrelation of one tephra horizon with another geochemically similar eruption, and alteration of tephra chemistry over time.30

Tephrochronology is an important dating tool in east African hominin sites, throughout both the Pliocene and the Quaternary, due to its association with the East African Rift System.31 In these contexts, tephrochronology is frequently most effective when combined with U-series dating, magnetostratigraphy (another age-equivalent method based on the geomagnetic polarity timescale), and OSL dating.

Dating Case Studies

Case Study: Thulamela, South Africa

Thulamela is a Late Iron Age site near Pafuri in the Kruger National Park of South Africa. It is a sizable hilltop settlement, enclosed by stone walls, with a large structure identified as a chief’s residence or palace at the highest point. The site has been excavated, revealing two distinct phases of occupation, the first consistent with habitation of the site before the construction of the stone walling, with pottery stylistically consistent with a pre-Khami group.32 The second occupation is associated with the extensive stone walling, and Khami pottery.Vogel (2000) obtained a suite of eighteen 14C dates, which distinguish three chronological phases.33 Three of these dates are plotted in figure 4A. The first date, Pta-7307, is from charcoal associated with a midden at the outskirts of the large palace wall, and calibrates to a likely age range 1290–1420 ce (2σ‎) according to the Southern Hemisphere curve, in agreement with five other dates from the pre-walling archaeological deposits (not shown in figure 4A). The second date, Pta-7243, is from a male skeleton, buried inside the enclosure, and calibrates to a slightly later age range 1390–1480 ad (2σ‎), consistent with the second Khami phase. The third date, Pta-7103, is on charcoal from a midden that accumulated after the construction of the main walls, and is likely to postdate the construction and habitation of the main enclosure. This last date calibrates to a relatively large age range of 180 years between 1450 and 1630 ce (2σ‎). Note the reason for the lack in precision in this date is the plateau in the radiocarbon curve between c. 1450 and 1600 ce, shown in figure 4A.

Scientific Dating Methods in African ArchaeologyClick to view larger

Figure 4. A: Three radiocarbon dates from the Iron Age site Thulamela (Vogel 2000), plotted against the Southern Hemisphere calibration curve SHCal13. Note the large uncertainty of 180 years in the calibrated date Pta-7103, caused by the plateau in the calibration curve c. 1450–1600 ce. B: Predicted secular variation of the geomagnetic field at Mapungubwe (22.212oS, 29.387oE) according to the pfm9k model of Nilsson et al. Most geomagnetic field models predict large variations in declination and inclination in southern Africa. Large variations are supported by direct measurements (Hare et al., 2018, not shown), which is promising for future archaeomagnetic dating in the region. Illustrations by the authors.

The conclusion drawn from this suite of dates is that the site had been occupied in the 14th century ce, and by the early 15th century the Khami-style settlement had been constructed. By 1600 ce the site was likely abandoned, but this may have happened before. The uncertainty in the radiocarbon date Pta-7103, and others like it from the post-wall midden, shows that it is impossible to say whether Thulamela was abandoned slowly, over 150 years, or whether it was a relatively sudden event. Elsewhere, radiocarbon dates from settlements in Zimbabwe show that Khami-style settlements persisted until the end of the 17th century. The period between 1450 and 1650 ce is an important time in the Late Iron Age of southern Africa, with debate among archaeologists about the collapse or transition of Late Iron Age societies, and therefore the issue of this plateau in the radiocarbon curve limits the ability of archaeologists to answer important questions about this time period. However, one technique that might be useful in this period is archaeomagnetic dating. In figure 4B we show the predicted secular variation close to the site of Thulamela according to one geomagnetic field model. This model shows strong changes in southern Africa during this particular time period, declination varying by about 5o, and inclination varying by about 10o. Direct measurements of other Iron Age sites suggest even higher rates of directional change. These changes could be utilized to provide archaeomagnetic dates for Late Iron Age sites that contain burnt features such as hearths and earthen floors. The rapidity of these changes suggests that archaeomagnetic dating could provide age uncertainties comparable with, or better than, radiocarbon.

Case Study: The Haua Fteah, Northeast Libya

Increasingly, different dating techniques are sometimes combined with one another, as a powerful means to corroborate chronology, or to tease apart different aspects of site formation history. The sensible use of Bayesian modeling software designed for chronological analysis can help to compare and combine dates in a rigorous way.

The Haua Fteah is a large cave situated in northeast Libya with a long sequence of human occupation, spanning tens of thousands of years. Given this long sequence, the site is central in discussions of North African prehistory from the original dispersal into the region through to debates about the movement of domesticated plants and animals throughout the Holocene and historical periods. The site was originally excavated and dated in the 1950s and 1960s; recent excavations have focused on improving the chronology with a combination of methods, including radiocarbon dating of multiple materials, single-grain OSL dating of sediments, tephrochronology, and electron spin resonance dating of tooth enamel (a trapped charge technique with similar principles to luminescence dating).34 The diversity of methods employed is a strength, with each method acting to corroborate the others. However, particular care should be taken to ensure that different dating measurements are combined on a correct, common age scale. For example, radiocarbon dates are given in radiocarbon years before 1950 (bp), while OSL ages are given in years before measurement. The example of Haua Fteah shows that care should be taken when combining dating methods, if the goal is a highly resolved chronology.

Bayesian methods are particularly powerful in such situations, where the archaeologist wishes to analyze groups of dates from a stratigraphic sequence. The mathematics of Bayesian statistics allows for the complicated probability distribution functions that describe the likely age range of a calibrated radiocarbon date to be handled with relative ease. In particular, these methods enable probabilistic ranges to be ascribed to chronological breaks or boundaries in a stratigraphic sequence. Other advantages are that these methods provide relatively easy quantitative identification of outliers, and different dates can be synchronized on a common age scale. None of these useful analyses would be straightforward with more conventional statistical methods.

Bayesian models based on about fifty dates were constructed for the Haua Fteah, constrained by “priors” that incorporate the stratigraphic structure of the depositional sequence (e.g., the relative order of each facies, the insertion of chronological breaks or boundaries where evident in the depositional sequence). These models evaluated the correspondence between dates derived from different methods, within alternative sedimentological and cultural stratigraphic frameworks. The statistical models also formally identified inconsistent dates, thereby highlighting possible stratigraphic ambiguities or inaccurate dates, and discounted such dates in the overall model. The updated chronology can be meaningfully contrasted with local and global climate records to unpick the technological consequences of environmental shifts, including the appearance of microlithic Oranian technologies at the peak of the Last Glacial Maximum, at c. 21.5 kbp when global climates were at their coldest during the last glacial cycle, and a microlithic Capsian during the Younger Dryas, an abrupt, and marked cold event in the Northern Hemisphere c. 12.9–11.7 kbp. The new chronology also emphasizes the early development of Upper Palaeolithic Dabban technologies c. 43–40 ka bp in this region.

The Status of Dating Methods and Laboratories in Contemporary African Archaeology and History

Despite some of the limitations of radiocarbon, such as the issue of calibration plateaus, it remains the most useful and accurate dating technique in African contexts. Unfortunately, there has in the past been a shortage of analytical expertise on radiocarbon dating in Africa, largely due to the expense of establishing and operating a radiocarbon laboratory. Traditional liquid scintillation counting techniques are currently employed at the Laboratoire Carbone-14, Institut Fondamental d’Afrique Noire (IFAN), in Dakar, Senegal, and the Laboratoire de Datation par le Radiocarbone, in Cairo, Egypt. However, the only facility capable of making radiocarbon AMS measurements on the continent currently is in South Africa, and it is establishing the laboratory procedures to produce precise radiocarbon measurements. This represents a very promising development for the continent.

OSL ages require considerably more instrumental time, are typically several times the cost of a radiocarbon date, and often require the direct involvement of a luminescence dating specialist in project design, sampling and data interpretation.35 Moreover, archaeological sites present some of the more challenging conditions for luminescence dating, with thin sedimentary features, complex site histories, and sampling restrictions related to heritage preservation. Consequently, OSL ages are not as common in the archaeological literature as radiocarbon dates, despite their wider age range, and greater applicability to geological and sedimentological materials. However, the development of single-grain methods has seen rapid increases in the application of luminescence dating to archaeological sites, and indeed, many methodological advances have been developed to address archaeological problems. There exist a number of laboratories capable of performing OSL measurements in Africa, including a laboratory at Rhodes University, and an active OSL dating laboratory at the University of the Witwatersrand, South Africa.

Applications of OSL dating in Africa have frequently demonstrated the great antiquity of archaeological contexts, far older than alternative dating methods had established. In the Nubian Desert, a novel application to sediment that had blown over and partially covered rock engravings verified that the engravings dated to the terminal Pleistocene, earlier than had been previously accepted. Older evidence for early art, including incised shell beads, ostrich eggshell fragments and ochres, and pigment processing tools, comes from rock shelters and caves in southern Africa and north Africa, extending beyond c. 100 ka and upturning previous assumptions about the beginnings of art.36 The application of OSL dating to Middle Stone Age deposits in Africa helped to establish the extreme antiquity of Middle Stone Age technologies and particular lithic industries, including the Howieson’s Poort and Still Bay in southern Africa and the Aterian in north Africa.37

U-series dating has been extensively applied to African archaeological contexts and the hominin record, with particular success at cave sites, and in combination with palaeomagnetic constraints (e.g., the geomagnetic reversal timescale) or OSL dating.38 Such combined approaches have recently been effectively employed to constrain the ages of stunning new hominid discoveries in southern Africa, that is, Australopithecus sediba and Homo naledi.39 The development of a uranium-series dating facility at the Department of Geology, University of Cape Town, South Africa, the first of its kind in Africa, presents a particularly exciting development for the continent.

Frontiers of Dating in African Archaeology

It is important to consider that scientific dating methods are an active area of fundamental research. The refinement of techniques is currently enabling new approaches to archaeology and the study of the African past. Several new developments are particularly exciting. Single-compound radiocarbon dating is being developed to target individual molecules, such as hydroxyproline, a component of the protein collagen found within bone, and new techniques are also allowing the non-destructive extraction of organics for radiocarbon analysis (e.g., lipids preserved in pottery).40 The principal advantage of these refined 14C dating methods is that they should allow more accurate estimates of the age of residues, because individual molecules can be better associated with particular practices such as dairying, are better preserved in the archaeological record, and are unlikely to be present as a result of contamination. Another promising development is the rapid advance in new OSL methods such as K-feldspar pIRIR (post-infrared infrared) or TT-OSL, and the combination of U-series dating with electron spin resonance dating (U-series-ESR). On older timescales, archaeologists and palaeoanthropologists are increasingly combining a variety of complementary techniques, including palaeomagnetism and tephrochronology, to place constraints on the complex histories of hominid-bearing localities. This practice should be welcomed, along with collaboration between archaeologists, geologists, and dating specialists. There is a pressing need for technical skills and facilities on the continent, a challenge that Africa is well-positioned to meet. Another frontier is the development of archaeomagnetic dating in Africa, which has the potential to resolve longstanding and complex questions around the end of the Late Iron Age in southern Africa. A direct dating method applicable to rock art remains the ultimate technical challenge (because radiometric techniques are often out of the question) but one that probably awaits future generations.

American Chemical Society National Historic Chemical Landmarks. Discovery of Radiocarbon Dating.

Oxford Radiocarbon Accelerator Unit. Radiocarbon Dating.

Further Reading

Aitken, Martin J. Science-Based Dating in Archaeology. London: Longman, 2014.Find this resource:

Hus, Joseph, R. Geeraerts, and Simo Spassov. Archaeomagnetism and Archaeomagnetic Dating. Dourbes, Belgium: Centre de Physique du Globe Institut Royal Météorologique de Belgique, 2003.Find this resource:

Malainey, Mary E. A Consumer’s Guide to Archaeological Science. New York: Springer, 2011.Find this resource:

Nilsson, Andreas, Richard Holme, Monika Korte, Neil Suttie, and Mimi Hill. “Reconstructing Holocene Geomagnetic Field Variation: New Methods, Models and Implications.” Geophysical Journal International 198, no. 1 (2014): 229–248.Find this resource:

Tauxe, Lisa, Subir K. Banerjee, Robert F. Butler, and Rob van der Voo. Essentials of Paleomagnetism. 5th Web ed., 2016.

Taylor, R. E., and Ofer Bar-Yosef. Radiocarbon Dating. Walnut Creek, CA: Left Coast Press, 2014.Find this resource:

Vogel, John C. “Radiocarbon Dating of the Iron Age Sequence in the Limpopo Valley.” Goodwin Series 8 (December 2000): 51–57.Find this resource:

Walker, Mike J. C. Quaternary Dating Methods. Chichester, UK: John Wiley & Sons, 2005.Find this resource:

Notes:

(1.) T. W. Linick, et al., “Accelerator Mass Spectrometry: The New Revolution in Radiocarbon Dating,” Quaternary International 1 (January 1, 1989): 1–6. Also see Rachel Wood, “From Revolution to Convention: The Past, Present and Future of Radiocarbon Dating,” Journal of Archaeological Science 56 (April 2015): 61–72. Many dates for African archaeological sites can be found in the regional date lists, which were semi-regularly compiled and published by various authors in the Journal of African History from the early 1960s to the 1980s. John Vogel also published much of the output of his laboratories (both Groningen and Pretoria) in date lists in the journal Radiocarbon [John C. Vogel, “Groningen Radiocarbon Dates IX,” Radiocarbon 12, no. 02 (July 18, 1970): 444–471; John C. Vogel and M. Marais, “Pretoria Radiocarbon Dates I,” Radiocarbon 13, no. 02 (July 18, 1971): 378–394; John C. Vogel and Ebbie Visser, “Pretoria Radiocarbon Dates II,” Radiocarbon 23, no. 01 (July 18, 1981): 43–80; John C. Vogel, Annemarie Fuls and Ebbie Visser, “Pretoria Radiocarbon Dates III,” Radiocarbon 28 (1986): 1133–1172]; and in the South African Archaeological Bulletin [e.g., John C. Vogel, “Radiocarbon Dating of the Iron Age Sequence in the Limpopo Valley,” Goodwin Series 8 (December 2000): 51).].

(2.) David K. Wright, “Accuracy vs. Precision: Understanding Potential Errors from Radiocarbon Dating on African Landscapes,” African Archaeological Review 34, no. 3 (September 30, 2017): 303–319.

(3.) It is now known that this value is very slightly inaccurate, but radiocarbon dating laboratories continue to use the original estimate to maintain comparability.

(4.) Paula J. Reimer, et al., “IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years Cal BP,” Radiocarbon 55 (2013): 1869–1887. Also see Alan G. Hogg, et al., “SHCal13 Southern Hemisphere Calibration, 0–50,000 Years Cal BP,” Radiocarbon 55, no. 4 (November 3, 2013): 1889–1903.

(5.) Caitlin E. Buck, et al., “Combining Archaeological and Radiocarbon Information: A Bayesian Approach to Calibration,” Antiquity 65, no. 249 (December 2, 1991): 808–821. Also see Christopher Bronk Ramsey, “Bayesian Analysis of Radiocarbon Dates,” Radiocarbon 51 (2009): 337–360.

(6.) Eduardo Q. Alves, et al., “The Worldwide Marine Radiocarbon Reservoir Effect: Definitions, Mechanisms, and Prospects,” Reviews of Geophysics 56, no. 1 (March 1, 2018): 278–305. It is also possible to calculate local marine reservoir offsets at Calib database.

(7.) For example, Ricardo Fernandes, et al., “Food Reconstruction Using Isotopic Transferred Signals (FRUITS): A Bayesian Model for Diet Reconstruction,” ed. Luca Bondioli. PLoS ONE 9, no. 2 (February 13, 2014): e87436.

(8.) Bone carbonate is not considered a suitable material for dating due to the high likelihood of diagenesis and contamination, although see Antoine Zazzo, “Bone and Enamel Carbonate Diagenesis: A Radiocarbon Prospective,” Palaeogeography, Palaeoclimatology, Palaeoecology 416 (December 15, 2014): 168–178; and Zazzo, “Bone and Enamel Carbonate Diagenesis: A Radiocarbon Prospective,” Palaeogeography, Palaeoclimatology, Palaeoecology 416 (December 15, 2014): 168–178.

(9.) David Collett and Peter Robertshaw, “Problems in the Interpretation of Radiocarbon Dates: The Pastoral Neolithic of East Africa,” African Archaeological Review 1, no. 1 (1983): 57–74.

(10.) Austin Long, R. B. Hendershott, and P. S. Martin, “Radiocarbon Dating of Fossil Eggshell,” Radiocarbon 25, no. 2 (1983): 533–539. For a contrasting view also see John C. Vogel, Ebbie Visser, and Annemarie Fuls, “Suitability of Ostrich Eggshell for Radiocarbon Dating,” Radiocarbon 43, no. 1 (2001): 133–137.

(11.) Jürgen C. Freundlich, et al., “Radiocarbon Dating of Ostrich Eggshells,” Radiocarbon 31, no. 3 (July 18, 1989): 1030–1034. Also see Vogel, Visser, and Fuls, “Suitability of Ostrich Eggshell for Radiocarbon Dating,” 133–137.

(12.) Katerina Douka, R. E. M. Hedges, and Thomas F. G. Higham, “Improved AMS 14C Dating of Shell Carbonates Using High-Precision X-Ray Diffraction and a Novel Density Separation Protocol (CarDS),” Radiocarbon 52 (2010): 735–751.

(13.) Fiona Brock, et al., “Current Pretreatment Methods for AMS Radiocarbon Dating at the Oxford Radiocarbon Accelerator Unit (ORAU),” Radiocarbon 52 (2010): 103–112. Also see Michael I. Bird, et al., “Radiocarbon Dating of ‘Old’ Charcoal Using a Wet Oxidation, Stepped-Combustion Procedure,” Radiocarbon 41 (1999): 127–140; and Thomas F. G. Higham, Roger Jacobi, and C. B. Ramsey, “AMS Radiocarbon Dating of Ancient Bone Using Ultrafiltration,” Radiocarbon, 2006.

(14.) Brock, et al., “Current Pretreatment Methods for AMS Radiocarbon Dating”; Bird, et al., “Radiocarbon Dating of ‘Old’ Charcoal”; and Higham, Jacobi, and Ramsey, “AMS Radiocarbon Dating of Ancient Bone Using Ultrafiltration.”

(15.) John Hellstrom and Robyn Pickering, “Recent Advances and Future Prospects of the U–Th and U–Pb Chronometers Applicable to Archaeology,” Journal of Archaeological Science 56 (April 1, 2015): 32–40.

(16.) Jon Woodhead and Robyn Pickering, “Beyond 500 Ka: Progress and Prospects in the UPb Chronology of Speleothems, and Their Application to Studies in Palaeoclimate, Human Evolution, Biodiversity and Tectonics,” Chemical Geology 322–323 (September 5, 2012): 290–299.

(17.) See reviews in Ioannis Liritzis, et al., Luminescence Dating in Archaeology, Anthropology, and Geoarchaeology: An Overview (Heidelberg: Springer, 2013).

(18.) TL methods have changed little since originally developed, see M. J. Aitken, Thermoluminescence Dating (Orlando, FL: Academic Press, 1985). For an example concerning burnt flints, see Abdeljalil Bouzouggar, et al., “82,000-Year-Old Shell Beads from North Africa and Implications for the Origins of Modern Human Behavior,” Proceedings of the National Academy of Sciences 104, no. 24 (2007): 9964–9969; For an example of iron-smelting furnaces, see D. I. Godfrey-Smith and J. L. Casey, “Direct Thermoluminescence Chronology for Early Iron Age Smelting Technology on the Gambaga Escarpment, Ghana,” Journal of Archaeological Science 30 (2003): 1037–1050

(19.) Ages back to 1 Ma were obtained by isolating only the slowest component of the six that constitute the OSL signal: E. J. Rhodes, et al., “New Age Estimates for the Palaeolithic Assemblages and Pleistocene Succession of Casablanca, Morocco,” Quaternary Science Reviews 25, no. 19–20 (October 1, 2006): 2569–2585. TT-OSL methods rely upon traps that are less light sensitive and that accumulate electrons more slowly, so saturating at considerably greater ages. However, these require a long bleaching period, and so are applicable only in certain geological contexts; see Robyn Pickering, et al., “Paleoanthropologically Significant South African Sea Caves Dated to 1.–1.0 Million Years Using a Combination of U–Pb, TT-OSL and Palaeomagnetism,” Quaternary Science Reviews 65 (April 1, 2013): 39–52.

(20.) For example, the dating of early hominid sites in Eurasia: Gunther A. Wagner, et al., “Radiometric Dating of the Type-Site for Homo Heidelbergensis at Mauer, Germany,” Proceedings of the National Academy of Sciences 107, no. 46 (November 16, 2010): 19726–19730.

(21.) A. S. Murray and Ann G Wintle, “Luminescence Dating of Quartz Using an Improved Single-Aliquot Regenerative-Dose Protocol,” Radiation Measurements 32, no. 1 (February 1, 2000): 57–73.

(22.) Zenobia Jacobs and Richard G. Roberts. “Advances in Optically Stimulated Luminescence Dating of Individual Grains of Quartz from Archeological Deposits,” Evolutionary Anthropology: Issues, News, and Reviews 16, no. 6 (December 19, 2007): 210–223. Also see Geoffrey A. T. Duller, “Single-Grain Optical Dating of Quaternary Sediments: Why Aliquot Size Matters in Luminescence Dating,” Boreas 37, no. 4 (November 1, 2008): 589–612.

(23.) Rex F. Galbraith and R. G. Roberts, “Statistical Aspects of Equivalent Dose and Error Calculation and Display in OSL Dating: An Overview and Some Recommendations,” Quaternary Geochronology 11 (August 1, 2012): 1–27.

(24.) Joseph L. Kirschvink, “The Least-Squares Line and Plane and the Analysis of Palaeomagnetic Data,” Geophysical Journal International 62, no. 3 (September 1980): 699–718.

(25.) I. Zananiri, et al., “Archaeomagnetic Secular Variation in the UK during the Past 4000 Years and Its Application to Archaeomagnetic Dating,” Physics of the Earth and Planetary Interiors 160, no. 2 (February 16, 2007): 97–107.

(26.) For example, in Fco. Javier Pavón-Carrasco, et al., “A Matlab Tool for Archaeomagnetic Dating,” Journal of Archaeological Science 38, no. 2 (February 1, 2011): 408–419. For southern African data see Vincent J. Hare et al., “New Archeomagnetic Directional Records From Iron Age Southern Africa (ca. 425–1550 ce) and Implications for the South Atlantic Anomaly,” Geophysical Research Letters 45, no. 3 (2018): 1361–1369.

(27.) Compare Moira A. Wilson, et al., “Rehydroxylation (RHX) Dating of Archaeological Pottery,” Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences 468, no. 2147 (November 8, 2012): 3476–3493 with; and Vincent J. Hare, “Theoretical Constraints on the Precision and Age Range of Rehydroxylation Dating,” Royal Society Open Science 2, no. 4 (April 1, 2015):140372.

(28.) Sarah Blain and C. Hall, “Direct Dating Methods,” in The Oxford Handbook of Archaeological Ceramic Analysis, ed. Alice M. W. Hunt (Oxford: Oxford University Press, 2017), 671–690.

(29.) Christine S. Lane, et al., “Advancing Tephrochronology as a Global Dating Tool: Applications in Volcanology, Archaeology, and Palaeoclimatic Research,” Quaternary Geochronology 40 (May 1, 2017): 1–7.

(30.) David J. Lowe, “Tephrochronology and Its Application: A Review,” Quaternary Geochronology 6, no. 2 (April 1, 2011): 107–153.

(31.) Harald Stollhofen, et al., “Fingerprinting Facies of the Tuff IF Marker, with Implications for Early Hominin Palaeoecology, Olduvai Gorge, Tanzania,” Palaeogeography, Palaeoclimatology, Palaeoecology 259, no. 4 (March 31, 2008): 382–409; Alan L. Deino, et al., “40Ar/39Ar Dating, Paleomagnetism, and Tephrochemistry of Pliocene Strata of the Hominid-Bearing Woranso-Mille Area, West-Central Afar Rift, Ethiopia,” Journal of Human Evolution 58, no. 2 (February 1, 2010): 111–126; Catherine M. Martin-Jones, et al., “Glass Compositions and Tempo of Post-17 Ka Eruptions from the Afar Triangle Recorded in Sediments from Lakes Ashenge and Hayk, Ethiopia,” Quaternary Geochronology 37 (February 1, 2017): 15–31; and Nick Blegen, et al., “Distal Tephras of the Eastern Lake Victoria Basin, Equatorial East Africa: Correlations, Chronology and a Context for Early Modern Humans,” Quaternary Science Reviews 122 (August 15, 2015): 89–111.

(32.) Maryna Steyn, et al., “Late Iron Age Gold Burials from Thulamela (Pafuri Region, Kruger National Park),” South African Archaeological Bulletin 53, no. 168 (December 1998): 73; and M. M. Kusel, “A Preliminary Report on Settlement Layout and Gold Melting at Thula Mela, a Late Iron Age Site in the Kruger National Park,” Koedoe 35, no. 1 (September 24, 1992): 55–64.

(34.) Katerina Douka, et al., “The Chronostratigraphy of the Haua Fteah Cave (Cyrenaica, Northeast Libya),” Journal of Human Evolution 66 (January 1, 2014): 39–63.

(35.) Good user guides provide details for site selection and sampling in the field: Michelle S. Nelson, et al., “User Guide for Luminescence Sampling in Archaeological and Geological Contexts,” Advances in Archaeological Practice 3, no. 2 (May 16, 2015): 166–177.

(36.) Zenobia Jacobs, “An OSL Chronology for the Sedimentary Deposits from Pinnacle Point Cave 13B—A Punctuated Presence,” Journal of Human Evolution 59, no. 3–4 (September 1, 2010): 289–305.

(37.) R. N. E. Barton, et al., “OSL Dating of the Aterian Levels at Dar Es-Soltan I (Rabat, Morocco) and Implications for the Dispersal of Modern Homo Sapiens,” Quaternary Science Reviews 28, no. 19–20 (September 2009): 1914–1931. Also see Zenobia Jacobs, et al., “Ages for the Middle Stone Age of Southern Africa: Implications for Human Behavior and Dispersal,” Science 322, no. 5902 (October 31, 2008): 733–735.

(38.) Darryl J. de Ruiter, et al., “New Australopithecus Robustus Fossils and Associated U-Pb Dates from Cooper’s Cave (Gauteng, South Africa),” Journal of Human Evolution 56, no. 5 (May 1, 2009): 497–513. Also see Robyn Pickering and Jan D. Kramers, “Re-Appraisal of the Stratigraphy and Determination of New U-Pb Dates for the Sterkfontein Hominin Site, South Africa,” Journal of Human Evolution 59, no. 1 (July 1, 2010): 70–86.

(39.) Robyn Pickering, et al., “Contemporary Flowstone Development Links Early Hominin Bearing Cave Deposits in South Africa,” Earth and Planetary Science Letters 306, no. 1–2 (June 1, 2011): 23–32; and Paul H. G. M. Dirks, et al., “The Age of Homo Naledi and Associated Sediments in the Rising Star Cave, South Africa,” ELIfe 6 (2017).

(40.) Anat Marom, et al., “Single Amino Acid Radiocarbon Dating of Upper Paleolithic Modern Humans,” Proceedings of the National Academy of Sciences (2012): 201116328. See also Thibaut Devièse, et al., “New Protocol for Compound‐Specific Radiocarbon Analysis of Archaeological Bones,” Rapid Communications in Mass Spectrometry 32, no. 5 (2018): 373–379; and Thibaut Devièse, et al., “Supercritical Fluids for Higher Extraction Yields of Lipids from Archeological Ceramics,” Analytical Chemistry 90, no. 4 (February 20, 2018): 2420–2424.