A scientific procedure that is extremely useful for ancient historians is carbon dating, also known as radiocarbon dating. Sometimes in our internet travels, we stumble upon claims made about carbon dating that are misleading, or even inaccurate. So I want to take some time here to explain how carbon dating works and explode some of the myths about it. Hopefully, in so doing, I can dispel some of the clouds surrounding the process.
Carbon, as you may know, is an abundant natural element that can be found in our atmosphere, oceans, and in all living creatures. An extremely rare carbon isotope is carbon-14 (sometimes abbreviated 14C or C-14), which amounts to about one in a trillion carbon atoms. It is produced in the upper atmosphere when cosmic rays bombard nitrogen-14 atoms.
Carbon gets into both plants and animals. After it oxidizes, plants absorb it through photosynthesis, and animals absorb it indirectly by eating plants (or eating other animals that eat plants). Plants and animals absorb both carbon-12 and carbon-14, and while they are alive, they generally have one trillion times more carbon-12 than carbon-14 in their bodies. After they die, an interesting change takes place. They retain the carbon-12, but lose the carbon-14. Carbon-14 is radioactive and unstable, so it decays over time, gradually converting back into nitrogen-14.
Here’s the interesting thing: we can measure how much carbon-12 and carbon-14 is in organic matter. It is therefore possible to measure how long a plant or animal has been dead based on the ratio of carbon-14 to carbon-12 within it. When a creature dies, the ratio of carbon-12 to carbon-14 should be a trillion to one. When half of the carbon-14 has left the body (we call that a “half-life”), the ratio would then be 2 trillion to one. To arrive at such a ratio, it takes approximately 5,730 years. Then, for it to halve again, it takes another 5,730 years, and to halve again, another 5,730 years. You see now how, by measuring the ratio, we then can determine the time of the creature’s death.
Three principal methods are used to carbon date: gas proportional counting, liquid scintillation counting, and accelerator mass spectrometry. The first two methods count beta particles, which are created by radiocarbon decay. The last, which counts the radiocarbon content directly, is considered the most efficient.
Now that we have the basics, let’s look at specifics and address several common (and not so common) misunderstandings about the process, sometimes propagated in popular books and articles written by those with limited knowledge of the subject.
Myth #1: Carbon dating can be used to date stones or bricks.
Generally, if it’s not organic, the procedure is useless. Some inorganic matter like minerals can be dated, as long as the mineral’s formation involved assimilation of carbon-14 in equilibrium with the atmosphere. Stones or bricks would not work. So if we find some ancient ruins, and there are no remains of once-living organisms (like bone, flesh, hair, blood, leather, pollen, or wood), we are out of luck. How about clay pots? Nope, but sometimes pots have residue of organic matter left in them, and this we can date.
Myth #2: Carbon dating can be used to date organic matter millions of years old.
The technique can be used to date specimens that died tens of thousands of years ago, but not millions. Why? After that long, the amount of carbon-14 becomes too miniscule for us to measure, because it can no longer be reliably distinguished from the carbon-14 that is created by the irradiation of nitrogen by neutrons from the spontaneous fission of uranium, which can be found in trace quantities in almost everything. Instead we must rely on other dating techniques, such as potassium-argon dating or rubidium-strontium dating.
Myth #3: Carbon dating can be used to determine the exact date that the living organism died.
Alas, no. It is not that precise. Carbon dating has a certain margin of error. The older the object, the greater the margin of error. A measurement includes the use of the half-life standard, comparison with a modern radiocarbon standard (oxalic acid), and correction for isotopic fractionation (basically that means expected fluctuation in the carbon amounts caused by the sample’s biochemistry). Standard errors are reported with a “±” value, which is calculated through statistical means.
Myth #4: When a radiocarbon measurement is reported as years BP (“before present”), this indicates the number of years before the measurement was taken.
Actually, no. The “present” is assumed to be the year 1950. The reason is so that a person doesn’t need to know the exact year that the measurement was taken. All reports assume 1950 as the present.
Myth #5: The ratio of carbon 12 to carbon-14 has remained constant in the atmosphere throughout time.
The level of atmospheric carbon-14 is not constant. Changes in the earth’s magnetic field over thousands of years have affected the amount of carbon-14 that can get in, and the sunspot cycle causes fluctuations in the magnetic field that can vary the level over decades. More recently, the amount of carbon-14 in the atmosphere has been affected by the combustion of fossil fuels by humans and the above-ground testing of thermonuclear weapons.
Myth #6: The amount of carbon-14 in the atmosphere has progressively increased over time (or progressively decreased over time).
Actually, it has fluctuated both ways. We have solid independent evidence for the fluctuations and reversals of the planet’s magnetic field, which affects the amount of carbon-14 that can get in. In 1967, Czech geophysicist Václav Bucha calculated the strength of the earth’s magnetic field in different periods of history by measuring the thermoremanent magnetization of baked clay artifacts from archaeological sites. What he discovered was that the earth’s magnetic field was 1.6 times stronger in 400 B.C.E. than it is today, 1.5 times stronger in 1 C.E. than today, and has been in a slow decline since then. Even more interesting is that, before 400 B.C.E., it was rising: the field’s strength rose steadily between 4000 B.C.E. and 400 B.C.E., peaking in 400 B.C.E. before dropping. Further studies have confirmed his findings. Moreover, study of paleomagnetism on the sea floor has conclusively proven that the earth’s magnetism has reversed itself dozens of times throughout earth’s history. Thus the amount of C-14 in comparison to C-12 has gone up and down.
Myth #7: At some point, the amount of carbon-14 forming in the atmosphere will match the amount of carbon-14 decaying, thus reaching an equilibrium.
This theory has been used by some to attempt to calculate the age of the earth, but it assumes that the formation of carbon-14 in the atmosphere occurs at a steady rate with no fluctuations. Unfortunately, that’s not the case. As we know, the earth’s magnetic field has an effect on the rate of carbon-14 formation, so sometimes the production rate will exceed the decay rate, and other times the decay rate will exceed the production rate. An equilibrium, can never therefore be reached.
Myth #8: It is impossible to know truly how much carbon-14 was in a creature when it died.
The amount of carbon-14 in a plant or an animal matches closely the amount in the atmosphere, because plants take in carbon from the atmosphere, and animals eat plants. Plants cannot distinguish between carbon-12 and carbon-14, so the carbon is absorbed wholesale. The only notable exception would be in the case of some mollusks (see Myth #14 below). To know what the ratio of carbon-12 to carbon-14 is in a creature, we need know only what the ratio is in the atmosphere, and measurement of the earth’s magnetic field has allowed us to determine this mathematically. Small deviations do not pose a problem for carbon dating.
Myth #9: There is no way to be certain a sample has not been contaminated.
If fresh organic material that is still rich in carbon-14 diffuses into a sample, either while it is in the ground, or in the lab, then it becomes contaminated. Such contamination would make the sample appear younger (not older) than it actually is. So contamination is a concern. Fortunately, we do have techniques for identifying and correcting contamination. These techniques include careful selection of samples, various methods of cleaning the material, while measuring each rinse, running “blanks” through the process, and dating different portions of the same sample. It has happened that different parts of a single sample can yield different carbon-14/carbon-12 ratios, because cracks, partial decay, or insect burrows might occur unequally throughout the sample. If the material is hopelessly contaminated, it will simply be pitched. The point is, we do have a decent idea when a sample is contaminated, and since collectors of specimens do their best to collect a large number of samples from a site, we are sure to have many readings to compare.
Myth #10: The rate of carbon-14 decay has varied considerably throughout earth’s history.
The checks we have done through dendrochronology (see Myth #11 below) have shown that the decay rate has remained relatively constant.
Many experiments have been performed in an attempt to see what could alter the rate of decay, such as atmospheric pressure, magnetism, gravity, and other environmental conditions, but none of them have produced any significant changes to the decay rate of carbon-14. Some experiments with alpha and beta emissions have produced very slight changes in decay characteristics of carbon-14, but not in its decay constants. Electron capture has produced slight deviation in the decay rate of Be-7, and deviations in the decay rate of other isotopes have been caused by a change in chemical conditions, but none of these concerns carbon-14 or isotopes of geologic interest that might be germane to carbon dating methods. Recent experiments have also disproved the hypothesis that neutrinos can affect carbon-14 decay.
Myth #11: Scientists use carbon dating by itself without any checks or balances.
Carbon dating can provide us with reliable and accurate dates, but not entirely on its own. A carbon date is not a calendar date. Knowing, as we do, that the fluctuations in earth’s magnetic field, and other factors, have an effect on the amount of carbon-14 in the atmosphere, calibration is necessary, so that we may convert a carbon date into a calendar date.
A reliable method of calibration we use is dendrochronology (a.k.a. “tree-ring dating”). Many tree species reliably lay down exactly one tree ring each year. During the entire life of the tree, the rings, once laid down, do not change. Since the tree is or was once alive, we can measure the carbon-14 concentration in each ring and this determine how much of it was in the atmosphere that year. This can be done for each ring in the tree, so that each ring can be matched with a calendar year
In addition, we know that the widths of the tree rings are affected by changing weather patterns. By measuring the widths of the rings in one tree, we can then match it to the tree rings in another tree, which came into being earlier and is already dead. Very old trees can provide us with radiocarbon concentration measurements for each calendar year going back thousands of years. We’ve done this using Bristlecone Pines in the USA and waterlogged Oaks in Ireland and Germany, and Kauri trees in New Zealand, so that we have records that extend back more than 8,000 years.
Let’s say we want to calibrate a carbon date we received on a sample. We try to find a tree ring in the general vicinity of the carbon date that has the same proportion of radiocarbon. Since we know the calendar dates of the tree rings, we can then obtain the date of the sample.
This procedure is complicated by the fact that radiocarbon dates are imprecise, and there are sometimes several possible tree ring matches to a radio carbon dated sample. The more precise the carbon date, the more precise the calibrated date can be.
You can identify a calibrated date either because it is explicit (the document might say “calibrated range”), or if you see the abbreviation “cal” before a date (such as “calBP,” “calAD,” or “calBC”). In calibration, a range of years is provided (e.g., 4040-3713 BC), never a single year.
Sometimes dendrochronology has assisted us in narrowing a date down. For example, a wooden walkway that was buried in a peat bog in England was carbon dated to about 4000 BCE. Tree ring dating found that the walkway was built from trees in the winter of 3807-3806 BCE. Carbon dating of seeds and wood buried beneath volcanic ash on the island of Thera pointed to a date around 1600 BCE for the eruption of the volcano. As this was a gigantic eruption, it cooled the earth’s atmosphere and had a great effect on tree growth around the world. As expected, oaks found in Ireland’s bogs showed an unusual cooling of the atmosphere between 1628 and 1618 BCE, and the bristlecone pines in California showed the same thing. Geologists independently determined the volcano erupted in 1645 BCE, give or take twenty years. These cross-checks have shown the general reliability of carbon dating.
For periods older than 8 or 9,000 years ago, we have other methods to check our carbon dates. For example, we can use information gathered from lake sediments in which organic carbon compounds have been preserved. Layers of sediment are formed at the bottom of lakes over time, when small light-colored algae called diatoms die and cover the lake floor each winter. They are, in turn, covered by a dark layer of sediment every summer. If a lake is very still, is low in oxygen, and has not been disturbed by glaciers or geologic activity, these microscopic layers will form an annual record in a similar manner as tree rings. The amount of carbon in the layers can be measured. We have a record of lake sediments going back 22,000 years, and recently scientists from the University of Oxford have discovered layers of sediment in Lake Suigetsu in Japan that will provide us with a record going back 52,800 years. Interestingly, lake sediment measurements have shown that the rate of carbon-14 decay hasn’t varied by more than 10% throughout those years.
We can see, therefore, that the variation in the rate of production of carbon-14 can be accounted for in carbon dating by using calibration. On occasion we might be able to use other means to verify a date. Stonehenge was determined by carbon dating to have been built between 1900 and 1500 B.C.E. Astronomers were able to determine that the arrangement of the stones aligned with the positions of the sun and moon as they were close to 4000 years ago. This confirms that the carbon dating is giving us accurate estimates.
Myth #12: Calibration has shown generally that carbon dates for samples have been too old.
For the period after 400 B.C.E., because the strength of the earth’s magnetic field has been on the decline, dendrochronology has shown that carbon dates have generally made samples seem slightly older than they were. When it comes to carbon dates before 400 B.C.E., dendrochronology generally has shown carbon dates to be too young. In other words, the object died longer ago than what the carbon date indicated. But the farthest the carbon date strays from the tree ring data is about 500-900 years.
Myth #13: Tree rings sometimes produce more than one growth ring per year, which would make the trees used for calibration look older than they actually are.
While it is true that some species of tree tend to produce two or more growth rings per year, that is not the case with the Bristlecone Pine and other trees used in dendrochronology, especially at the altitude from which samples are being taken. In their case, it is more likely that tree rings might be missing. A typical Bristlecone Pine is missing 5% of its rings. That being the case, the tree would appear younger than it actually is. This is one of the reasons why we always provide a calibration date range instead of an exact year.
Myth #14: Carbon dates taken from shellfish have proved inaccurate, thus showing carbon dating to be unreliable.
Scientists are well aware that readings taken from certain shellfish, such as freshwater clams and mussels, have been much older than they should. The reason for the anomalous readings is because these organisms build their shells from their water environment, which can contain carbon atoms from dissolved limestone. The limestone is so ancient that the carbon consists almost entirely of carbon-12 atoms and almost no carbon-14. The shellfish thus “inherit” some of the old age of the limestone, and the dates provided by carbon dating will be artificially older. We call this the “reservoir effect.” Because it is well known that certain types of marine life can be affected by these reservoirs of carbon in sedimentary rocks (seals are another example), we simply need to be cautious when interpreting the data (not all shellfish are affected in this way). But it is not correct to say that carbon dating as a whole is invalidated because of these special cases. Wood, for example, doesn’t absorb any limestone carbon, because it gets all of its carbon from the atmosphere.
Myth #15: Carbon dates derived from coal and diamonds have suggested that they were younger than 20,000 years.
Geologists, as you may know, say that coal, diamonds, oil, and natural gas are millions of years old. Carbon dating does not work well on samples that old. In the early days of radiocarbon dating, any sample that was older than 20,000 years was going to have such a small amount of carbon-14, that it became too difficult to separate the carbon-14 in the sample from the carbon-14 in the research environment. This issue prompted scientists at the time simply to say that the sample was older than 20,000 years and leave it at that. Even samples with no carbon-14 at all would register as having some, thereby making samples look younger than 20,000 years. Today we can be more precise, going back 70,000 years with accuracy and without risk of contamination from the laboratory environment, but sometimes people will cite these older results in an attempt to discredit carbon dating. Anthracite has been carbon dated today in excess of 70,000 years.
Another factor to consider is that carbon-14 can be created underground through the decay of uranium and thorium. We call this “cluster decay,” and it can produce carbon-14 in substances like coal and diamonds, which would cause them to appear younger than they actually are.
Myth #16: Volcanic eruptions could drastically affect the ratio of carbon-12 to carbon-14 in objects around the globe, thus causing them to appear older than they actually are.
There is no doubt that volcanoes eject huge amounts of carbon dioxide into the atmosphere, which would increase the amount of carbon-12 absorbed by plants and animals, this causing them to appear older when carbon dated. However, this would affect only objects a short distance away, so a worldwide effect cannot be expected.
If you didn’t know much about carbon dating before, I hope you found this article informative. Maybe a few misunderstandings have been cleared up. In researching the subject, I learned a few things myself. Having methods such as these to help us date material from the past is extremely important for the study of ancient history, as I am sure you have gathered.