by Roger Lewin

Anthropologists and archeologists must determine the age of human fossils and artifacts: without such an ability it would be impossible to reconstruct patterns of change in anatomy and behavior through evolutionary time. As we see in this chapter, the recent introduction of two new dating techniques into paleoanthropology revolutionized the interpretation of evolutionary sequences in the Near East concerning modern human origins.

Researchers have two options for seeking the age of fossils or artifacts: direct or indirect methods. Direct methods produce an age for the objects themselves - ultimately the preferred option. There are two types of problems here, however. First, for most material of interest - ancient fossils and most stone tools - there are no methods as yet available for direct dating. Some methods, such as carbon-14 and electron spin resonance, may be applied directly to teeth or young fossils, and indeed to the pigments of rock shelter and cave paintings; thermoluminescence dating may be applied directly to ancient pots or flint tools.

In practice, indirect dating methods represent the typical approach. Here, an age for the fossil or artifact is obtained by dating something that is associated with them. This may involve direct dating on nonhuman fossil teeth that occur in the same stratigraphic layer - by electron resonance, for instance. Thermoluminescence can date flint tools associated with human fossils. Ages may be attributed to fossils or artifacts through information about the evolutionary stage of nonhuman fossils associated with them, a technique known as faunal correlation.

The most common indirect approach, where it is feasible, is to date stratigraphic layers that lie below and above the object in question. Since stratigraphic layers accumulate from the bottom up, the lower layers are oldest, the upper layers youngest. An object that lies in sediments below a layer dated at 1.2 million years old and above another dated at 1.3 million years can be said to be between these two ages.

Anthropologists these days benefit from a range of dating techniques, which often depend on the steady decay of radioisotopes. The most widely used is the potassium-argon method (Figure 1), based on the decay of the radioisotope potassium-40 to the stable isotope argon-40. Minerals that contain potassium steadily accumulate argon-40, which offers a way of estimating the passage of time. A key factor in all such approaches, however, is that the clock should be set to zero at some point. This occurs in potassium-containing minerals that are ejected during volcanic eruption: the argon-40 in the crystal lattice is released at the high prevailing temperature. Argon-40 found in volcanic material is therefore that formed since the eruption. For this reason the sedimentary layers chosen for indirect dating of fossils are typically layers of volcanic ash.

Potassium-argon dating was first applied to an anthropological question in 1960, when volcanic ash above the famous Zinjanthropus skull found at Olduvai Gorge by Mary Leakey was dated: it showed the skull to be more than 1.75 million years old, rather than the 0.75 million years that had been inferred by other means. A revolution in dating had begun.

The major dating techniques initially available to anthropologists left them with a frustrating problem. Radiocarbon dating could extend to an age of 30,000 to 40,000 years ago, while the reliability of potassium-argon dating declines below about a million years ago. The origin of modern humans falls into this gap. Thermoluminescence and electron spin resonance techniques are able to cover this period, and so when they began to be applied in the mid-1980s there was a potential for another dating revolution.

Both techniques depend on the same phenomenon: the natural irradiation of the crystal structure of the target material, which may be tooth enamel or flint. The source of the irradiation is isotopes of uranium, thorium, and potassium that occur in the soil and within the target material itself. The effect of the irradiation is occasionally to boost electrons from their natural ground state to a higher, excited state. Most fall back to the ground state, but some become trapped at higher states because of imperfections or impurities in the crystal lattice. As time passes excited electrons accumulate in the same way that argon-40 builds up in potassium-containing minerals in volcanic ash. Just as the potassium-argon clock may be reset to zero by heating during a volcanic eruption, so too can the excited-electron clock - again by heating. Materials that have been dated by these techniques include pottery that has been fired and flint artifacts that were fortuitously exposed to heat (a hearth) at the time of use. The crystal lattice of tooth enamel is set at zero as it forms.

Thermoluminescence and electron spin resonance differ in the means by which they detect the quantity of trapped, excited electrons. In the former the target is heated, which causes the electrons to be kicked out of their excited level and fall to the ground state, releasing light: one photon per electron. The amount of light emitted is a measure of how many electrons were trapped at the excited level and of the time of their accumulation.

When electrons are trapped at an excited state they form paramagnetic centers and give rise to characteristic signals, known as electron spin resonance (ESR). The signal is detected by exposing the sample to a variable external magnetic field and microwaves; at a critical strength of the magnetic field microwaves are absorbed, to an extent determined by the number of trapped electrons. The strength of the ESR signal is therefore an indication of the time since the target material was set at zero.

With both techniques the extent of irradiation, internal and external, to which the sample has been exposed must be determined. This is achieved by analyzing the surrounding soil and the sample itself for the relevant radioisotopes. One complication is that uranium tends to leach into tooth mineral as it rests in the ground, thus altering the internal radiation dose. This can be accounted for in calculations, but it may add a margin of uncertainty to electron spin resonance dates. So far bone has proved unsuitable for this dating technique, partly because it absorbs more uranium than teeth, but also because its mineral content is in the range of 40 to 60 percent, compared with 96 percent in tooth enamel. The complexity of changes that go on in the mineral and amorphous phases of the bone during fossilization leads to ESR data that are too uncertain for reliable dating.

Thermoluminescence dating cannot be applied to tooth enamel because chemical changes induced by heat also release energy as light, thus confounding the measurement. Electron spin resonance dating cannot be applied to flint or pottery because the crystal lattices in these materials are insufficiently ordered for the clear production of an ESR signal by trapped, excited electrons.