How Old Is Old?

At a mammoth sandstone boulder called Jinmium, in what is today known as Australia's Northern Territory, ancient people painted figures and carved designs in the rock's crevices and sheltered areas while camping at the site. Just how old is Jinmium's art? Scientists used a complicated process called luminescence dating to find out.

Luminescence dating calculates age by measuring light energy (luminescence) released when minerals in soil sediments or old pottery are re-exposed to heat or light after being covered for long periods of time. At Jinmium, blowing sand eventually covered and buried the tools used by ancient people at the site. In 1996, scientists used luminescence dating to date soil sediments found at the same depth as the buried tools. Assuming the tools are the same age as the sediments they are buried in, scientists should have been able to determine the age of the art itself. The process, however, surprisingly dated the soil sediments at between 116,000 and 176,000 years old—that's more than twice as long as anyone thought people even lived in Australia!

At Katanga in the Democratic Republic of Congo, similarly startling results said sediments containing carved harpoon points were 80,000 years old. However, archaeologists thought people hadn't started making such tools until 40,000 years ago. In Siberia, luminescence dating gave stone artifacts an incredibly old age of 260,000 years!

Why such weird test results? What can scientists do to ensure accurate dating of ancient objects?


Luminescence dating uses heat (thermoluminescence, or TL) or light (optically stimulated luminescence, or OSL) to date pottery and soil sediments. The pottery has to have been previously heated in fire, and the soil sediments have to have been previously exposed to sunlight, and then covered or buried, for the dating process to work. Microscopic defects, which occur naturally within common mineral crystals in the pottery or sediments, act as “traps” for energy released from trace amounts of natural radioactive elements in the surrounding ground where they are buried. These traps were emptied by the original exposure to heat or sunlight. During burial, energy was again trapped. After excavation, re-heating the mineral or re-exposing it to light in the lab empties the traps, releasing light energy.

“So the longer the grain stayed buried, the more light is emitted in the lab,” explains Bert Roberts at La Trobe University in Australia, because there was more time for energy to accumulate within the traps.

Scientists measure luminescence with an instrument called a photomultiplier tube. Information about the trace radioactive elements available where the soil sediments were collected tells how much energy was stored each year. “This energy is derived from the nuclear decay of radioactive elements,” explains Roberts, referring to the rate at which radioactive isotopes of materials such as uranium, thorium, and potassium “degrade” to stable, nonradioactive elements. Total energy stored, divided by energy stored each year, equals the number of years since the last heating or burial away from sunlight.

When used correctly, luminescence dating can date from almost the present back to 150,000 years or more. In contrast, carbon dating methods can go back only about 50,000 years.

Rubble Trouble and Other Problems

But luminescence dating, like other measurement methods, has limitations. “For sediments, the main problem is whether the sediment at the time of deposition [burial or placement away from heat or light] was exposed to sunlight or heat for long enough to empty all the traps,” says Jim Feathers at the University of Washington. If the traps' “clocks” haven't been reset to zero, luminescence dating overestimates the age.

“This is the definite cause [for the error] at Jinmium,” agrees Roberts. After reanalyzing those sediments, Roberts concluded earlier this year that the real age of Jinmium's rock art is closer to—and probably even less than—10,000 years.

At Jinmium, Roberts believes unexposed grains of sediment came from chunks of sandstone that fell from the rock and were covered, along with the ancient people's tools, by windblown soil. Although the chunks' outer layers were exposed to sunlight along with the sediments they fell on, their insides might never have had their “clocks” reset by sunlight. When the covered sandstone chunks eventually crumbled, those unexposed sandstone grains mixed with the surrounding sediment being tested by thermoluminescence dating, causing wild overestimation of the sediments' age.

To neutralize this problem, which he calls “rubble trouble,” Roberts developed OSL methods that can date single sediment grains.

“It was no surprise to me when the Jinmium TL dates started tumbling as soon as I started dating single grains,” he says.

Roberts first used the method to date sand embedded in a fossilized mud-wasp nest in Australia. Since the fossilized nest lay over part of a rock painting, this enabled him to date the oldest known painting of a human figure anywhere in the world—more than 17,000 years old!

Reality Tests

How can scientists tell if laboratory results are correct? “Simply, the first ‘test’ is whether or not the results appear plausible in the light of all the other collected knowledge of the site or problem being tackled by dating,” says archaeologist Nigel Spooner at Australian National University. Is that socket wrench really from Babylonian times, or could it have been buried accidentally in older soil sediments?

Internal checks are another test. Some crystal traps are more light-sensitive than others. “The more sensitive ones are more likely to have been fully emptied by sunlight,” explains Jim Feathers, meaning their “clocks” are reset to zero more quickly. “Agreement in age determination from both kinds of traps is one indication the sample was well exposed to sun in antiquity.” Reproducibility—getting the same results when the test is repeated—is another sign of reliability.

“The best test of any dating method is to compare the answer with some independent form of evidence,” says Roberts. Carbon dating and nearby fossil evidence are two possible crosschecks.

As laboratories fine-tune luminescence dating, the technique promises to benefit archaeologists, geologists, and other scientific disciplines. “Always keep in mind the shortcomings of the various methods and the materials and situations in which they are applicable,” recommends Spooner. Indeed, with any scientific measurement, double check your results, and don't just blindly accept the numbers.

Radioactive Rocks

Low levels of radioactivity occur practically everywhere. Elements such as uranium, radium, and thorium are found in soil and rock in tiny concentrations, called trace amounts, in many areas. Other elements that are not normally radioactive, such as potassium or carbon, have some radioactive isotopes. An isotope is a form of an element that has the same number of protons but differs in its molecular weight because of one or more extra neutrons.

By giving off energy in forms including alpha rays and beta rays, radioactive materials “decay” over time to stable, nonradioactive elements. A radioactive material's half-life is the time in years for half of a quantity of that material to decay, or change. Generally, each type of radioactive material has its own constant rate of decay. For example, the half-life of U238, an isotope of uranium, is 4.51 billion years. By knowing the specific type of radioactive materials present at a site, scientists can calculate how much energy was given off and trapped in crystal defects per year.


  • isotope: One of two or more atoms having the same atomic number but different mass numbers.
  • nuclear: Of, relating to, or using energy that comes from the nuclei of atoms.
  • radioactive: Of, caused by, or having the process or property by which certain chemical elements, such as radium, give off energy in the form of rays.

Back to Top


  1. In Lesson 2, you learned about the different soil layers. In which soil layer do you think the scientists found the boulder called Jinmium? Why?
  2. What is the name of the process scientists used to date Jinmium?
  3. How does this process measure time?
  4. Why is this method of measuring an object's age not always accurate?