Paleolithic hunters built mammoth traps in what is now Mexico some 14,700 years ago. An unknown sea creature left footprints in sand some 550 million years ago, making them the oldest known footprints on Earth. The mysterious Denisovan humans reached southeast Asia 160,000 years ago, as evidenced by a jawbone found on the Tibetan Plateau. And 80 million years ago, a dinosaur egg never hatched.
Determining the age of fossils allows us to put the past in context and place it in proper chronological order. Archaeologists and paleontologists would be lost without this ability, yet it’s something many of us take for granted or don’t fully understand. Here’s how it’s done.
So many revelations about Earth’s history would be unknown to us without modern dating techniques. Sure, scientists might have a decent sense that the mammoth bones, squiggly lines etched in limestone, and the strange human jawbone are very old, but they wouldn’t really know their age to any meaningful precision. And without precision, scientists wouldn’t be able to place fossils in an evolutionary or geological context or perform comparative analyses, among other tasks required to reconstruct the distant past.
The dating of fossils is also important from an epistemological perspective. Biblical literalists insist that Earth is just 6,000 years old—a position that modern dating techniques show to be unequivocally false. It’s no exaggeration to say that the accurate dating of fossils is what allows us to truly know ourselves and our place in the universe.
Finding the right fossils
Scientists have been dating fossils for hundreds of years, and the techniques and methods used have become highly refined. That’s not to say the process is easy, straightforward, or without challenges, and there will always be room for improvement.
This may sound strange, but the first step in dating a fossil is to make sure the object of inquiry is actually a fossil. Many items brought to scientists for analysis are not really fossils, just things that superficially resemble them.
“These can be scratches on rocks, uneven erosion on a rock, or a ‘weird’ appearance from different minerals in the rock that make it look like it was once living,” Michael Meyer, a geologist and optical dating expert from Harrisburg University in Pennsylvania, explained in an email. “Because many people do not know how fossils are formed, and the human mind’s drive to ‘see’ recognizable shapes, there are often assumptions that a rock, which looks like an object, may be a fossilized version of it.”
Meyer was once given “fossilized” feet and ducks, which turned out to be oddly shaped rocks. The same phenomenon, he noted, makes people see weird things on Mars. As for deciding on what actually constitutes a fossil, Meyer provided two broad definitions: Fossils are any evidence of past life, typically organic material, that’s been turned to stone; or simply any evidence of ancient life.
Bridget Alex, an anthropologist at the California Institute of technology, said researchers are sometimes guilty of not finding the thing they actually want to date. For example, archaeologists wanting to date the destruction of an ancient city might date burnt charcoal and the charred remains of bones, but those may not necessarily correspond to the demise of the city. Or when wanting to date the bones of a Neanderthal, a scientist might date the bones of an animal found nearby—but “that creature could have wandered into the cave at a later date, died, but had nothing to do with the Neanderthal,” as she explained to me over the phone. The challenge, Alex said, is “finding the right fossil that answers the question you want to answer.”
Thankfully, “almost anything can be dated, but time, money, and context are the three biggest issues that prevent fossil dating,” said Meyer. “This is because it can take a lot of work to date a fossil.”
Fossils found in their original context are the easiest to date, given the typically generous amount of data available in and around the fossil. On the other hand, fossils found out of context, like a 140,000-year-old skull kept hidden in a well for 85 years, tend to be the most costly and difficult to date. As to how far back into time scientists can go, there’s virtually no limit (assuming time and money are no issue); the oldest rocks on Earth date to between 3.77 billion and 3.95 billion years old, with the oldest fossils dating back to around 3.42 billion years ago.
Scientists employ and typically combine two types of dating methods: relative and absolute. Relative dating is when fossils are arranged in order from oldest to youngest, while absolute dating pins an object down to an actual point in time.
The deeper it is, the older it (probably) is
Relative dating has been around since the 18th century, and it barely requires any technology beyond a shovel. Alex said the simplest approaches to dating often end up as being the most accurate, and they often result in the most useful data.
In the case of relative dating, the general idea is that, “as you get deeper, things tend to get older,” as Alex explained. Meyer referred to this as the Law of Superposition, “which states that older material is below younger material—much like a pile of laundry,” he said. Charles Lyell, an early geologist, used relative dating to great effect. He “used the percentages of animals with living relatives found in rocks to give a simple road map of figuring out time without dating,” Meyer explained, and it gave rise to some of the first time periods of the past, including the Pleistocene, (meaning “most recent”), Pliocene (“more recent”), Miocene (“moderately recent”), and Oligocene (“even less recent”).
Alex said relative dating can also be used when items are found together in the same context. A coin with a firm date, for example, can be used to date artifacts or fossils found around it. Or if human bones are found right next to a woolly mammoth skull, it suggests the two lived contemporaneously during the last ice age.
A major drawback of relative dating is the possibility of contamination. Responsible scientists, Alex said, “will assume that things have moved around” and that they’re “not going to get a perfectly undisturbed layer of cake with the oldest at the bottom.” Freezing, thawing, insect behavior, and human actions are among the things that can disturb the integrity of a site. Scientists need to be on the lookout for for these signs of “macrocontamination,” as Alex described it, and turn to geoarchaeologists who specialize in detecting these sorts of contamination issues.
In the case of absolute dating, it’s possible for scientists to nail down the age of a fossil to a year or a possible range of years. Absolute dating “uses chemical or physical principles to infer exact times, within a certain amount of error,” said Meyer. This approach, known as chronometric dating, relies on radioactive decay, which “happens when an element has too much energy and it spontaneously turns into another element in a predictable way,” Alex explained. This predictability serves as an accurate clock, and it’s “very reliable,” she added. Meyer said the isotopic dating of rocks or the minerals in them “is based on the known decay rates of certain unstable isotopes of elements,” as these rates have been “constant over geological time.”
The best known chronometric approach is radiocarbon dating—a technique that “revolutionized archaeology,” according to Alex. Basically, it allows for the dating of anything that comes from things that were once living—things like bones, teeth, leaves, and tree bark.
Carbon dating compares the ratio of radioactive carbon, which exists in every cell, to normal carbon. Radioactive carbon is not stable, and over time it loses its excess energy, causing it to turn into nitrogen. It “does so at a predictable rate,” said Alex, “so as soon as something dies, its radiocarbon, specifically the carbon 14 isotope, starts to decay, halving once every 5,730 years. In other words, only half of the original amount of carbon 14 will remain in an organic sample after 5,730 years. The catch is that these organic fossils have to be younger than 60,000 years old to be accurately dated, and that’s because there’s “often so little undecayed isotopic material in the sample that it gets hard to tell what is a real age signal or [just] noise,” said Meyer.
Contamination is an issue in carbon dating, as unrelated organic samples can creep into a fossil and make it seem younger than it really is, according to Alex. This “can even happen during lab work,” she said, but methods developed over the past several years can prevent this from occurring—things like cleaning and extracting collagen.
To date fossils older than around 60,000 years old, scientists can indirectly figure out their age by dating the inorganic sediments or minerals within which they were found. For example, optically stimulated luminescence (OSL) reveals the last time certain minerals in dirt were exposed to sunlight, providing a timeframe for when an item was buried (the trick with OSL to keep the samples away from any light—otherwise they’d be ruined, as Meyer reminded me). Scientists recently used OSL to show that Stonehenge began as an entirely different henge, as an example. Thermoluminescence dating indicates the last time items were heated, “such as stone tools that were either warmed up or dropped in fire,” said Alex. Uranium-series dating and electron spin resonance dating likewise measure the decay of isotopes, allowing for absolute dating of fossils.
Eleanor Scerri, an archaeologist with the Max Planck Institute for the Science of Human History, said there’s a mistaken assumption that chronometric dating “always takes place in the lab,” as she wrote in an email. Ultimately, to fully understand dates, scientists “need to understand how a given site was formed, whether the sediments have moved around and are being redeposited, or whether they are pristine.” This problem is akin to contamination issues involved in relative dating, as sediments can “move around during the thousands of years that pass before a site is excavated,” she said. Water can transport fossils and artifacts far from where they were originally buried, while animals or insects can disturb sediments of different ages around fossils and artifacts, according to Scerri. “If bad samples are given to a lab, no matter how great the chronometric method is, the outputs will be bad as well,” she said.
Indeed, chronometric techniques are powerful, but with great power comes the need for great methodological responsibility. Scerri says it’s important that scientists know what exactly is being dated and that they employ multiple dating techniques when possible.
“For example, we’ve discovered an extremely deep cave deposit full of fossil bones,” Scerri said. “We know that all those bones were redeposited in the cave in some kind of single flood event from radiocarbon dating shells all along the deposit. However, we are sure the bones are a lot older.” Accordingly, the team is currently using two different dating techniques to resolve the problem, namely uranium-series dating and electron spin resonance dating.
“If the sample of bones yields very different ages, it suggests that fossil material from all different time periods was just washed together in a big jumble during a substantial flood,” she wrote. “Alternatively, we might expect the bones to be of a similar age, since the species are pretty coherent across the deposit. They could still be substantially older than the deposition event, however. These mixed methods would allow us to test these hypotheses.”
Using multiple dating techniques is ideal, as it further strengthens and corroborates the age of the fossil being studied.
To date, these techniques have been used exclusively to date fossils found on Earth. That could change, however, given the plan to return Martian surface samples at some future juncture. So dating techniques, in addition to telling us profound things about our past, might someday tell us whether life ever existed on Mars—and when.
Correction: An earlier version of this post incorrectly described Michael Meyer as being from the University of Innsbruck.