A friend, in a discussion of the age of the Earth, recently said “I remember a study where there were some bones that were known to be only a few years old but they were carbon dated as being 8 or 9,000 years old”.
It helps to illustrate the problem: it’s someone’s recollection of something someone else told them some time ago, so there’s no citation, and few details. It’s difficult to check what actually happened. I found it then but can’t now, but there are a number of similar articles around, and we’ll get to the issues with them.
I thought today I’d talk a little bit about radiocarbon dating specifically (more on other forms of radioactive dating in coming days): how it works, what it can and cannot do, and why some of the common objections to it don’t really hold water.
Radioactive decay is a fascinating process. Unlike physical decay, it is not influenced by how hot, wet, pressured or otherwise its environment is. It is a process that appears random at the level of individual decays – it’s impossible to accurately predict when one will occur – but is highly predictable at a statistical level, when many decays are combined.
If a sample of a substance has, say, 1000 atoms of a radioactive chemical element in it, the meaning of the term ‘half-life’ is the amount of time taken for half of those atoms to undergo decay. The half-life for a particular type of decay to occur is constant. Say in our example the element has a half-life of 2 days. After 2 days, there are 500 atoms left (half the original amount). After 2 more days, there are 250 left (half as many as 2 days ago, quarter as many as were there originally, 4 days ago). After 6 days in total there are 125, after 8 days there are 62.5 (there are not really 0.5 atoms, so it would likely be 62 or 63). It will keep halving, each 2 days, until there are no atoms left.
The great majority of the carbon in our environment is carbon-12. It has 6 protons and 6 neutrons in its nucleus, for a total of 12 ‘nucleons’. Carbon-12 is stable and does not undergo radioactive decay. A very small amount of the carbon has an extra neutron for a total of 13. It’s called carbon-13 and is sometimes important in MRI scanning. Carbon-13 is also stable.
When neutrons in incoming solar radiation strike nitrogen-14 in the upper atmosphere it sometimes undergoes a tranformation into carbon-14, and carbon-14 is radioactive.
It later undergoes radioactive decay to release a beta particle and returns to being nitrogen-14. The half-life of this decay is 5730 years.
(An electron anti-neutrino is also released, and this equation isn’t properly charge balanced, and there’s an interesting reason for the minus sign for atomic number on the beta particle, but perhaps that’s too much detail for here.)
The carbon-14 in the upper atmosphere is distributed through the whole environment. Plants take it in when they use energy from sunlight to power photosynthesis, changing carbon dioxide and water into glucose and releasing oxygen. Living things either eat plants or eat things that eat plants, so all living things have carbon-14 in them. As long as they’re alive, they keep replenishing their stores of carbon-14, and so the amount in their bodies is stable. There are radioactive decays going on, but the supply is being replaced.
Once something dies, though, it stops breathing, stops eating, stops interacting with the environment. No new carbon-14 is added to its body, and what is there decays in a predictable way.
This is why radiocarbon dating can only be used on things that were formerly alive. It is not useful for dating rocks, or fossils (which are rock that’s replaced something that was formerly alive), or buildings, tools and other artifacts. Something had to be living, breathing, eating and drinking at some point in history to be able to be radiocarbon dated. There are other methods of dating other materials, that I’ll talk about in a different post.
The period of time that radiocarbon dating can stretch back is also limited. With a half-life of 5730 years, it’s very convenient for dating things that are low multiples of that, back to 25,000 years or so, but even at that point you’re 4 half-lives in and there’s only 1/16th of the original amount remaining. If you have a larger sample, so that the remnant is larger even after multiple halvings, radiocarbon dating can get you back 50,000 years or so, but not much further than that.
As a matter of perspective, there are artifacts of Aboriginal settlement in Australia that are older than that.
Radiocarbon dating is generally reliable. It makes some assumptions, but they are generally valid, or else able to be calibrated for. So, for example, if additional volcanic activity, or nuclear testing or other influences changed the amount of carbon-14 in the atmosphere at the period in which the formerly-living thing being dated was alive, that’s relevant: and can be taken into account. If additional carbon-14 has leached in or out of the sample, that’s relevant.
There are a few claims made by creationists, that are usually of the ‘gotcha’ type. They will have sent a sample to a lab with no information about what it is or where it comes from, requested dating, and then triumphantly revealed that the real known age is different. In all of these cases I have seen, including the one with which I started this piece, there is a clear, simple scientific explanation for the apparent disparity, which does not invalidate the method of radiocarbon dating for age determinations.
Often, the issue is that the organisms being dated were not in contact with the atmosphere in a ‘normal’ way. Examples include shells that grew in water from underground caves, where the carbon dissolved in the water in which they and their food grew had in many cases spent a very long period as part of limestone. The carbon-14 in it had long ago decayed already, so the carbon these shells were absorbing was depleted in carbon-14 relative to the ‘norm’ in other places. When comparing these shells to similar ones grown in fresher water, they appeared ‘older’ because they had lower levels of carbon-14.
When the science is done properly, the sample is tagged with the location where it is found, and these kinds of anomalies can be calibrated for. The ‘gotcha’ examples might give those seeking to impugn the method something to crow about, but they’re not good science.
For ease of navigation I will include links to each of the other posts in this series at the bottom of each post.
Why I think it’s important to understand evolution
Cosmogenesis, abiogenesis and evolution
Evolution and entropy
Facts, Theories and Laws
Radiometric dating and deep time
Four Forces of the Universe
Probability and evolution
Species and ‘baramin’, macro- and micro-evolution
Mitochondrial Eve and Y-chromosomal Adam
Complexity – irreducible and otherwise