(Intro)

    In 1946, chemist Willard Libby proposed a new way to scientifically estimate how old a dead thing is by measuring the amount of radioactive carbon left in it.  We've used this method to date a bunch of stuff: from the wood of an Egyptian pharoh's boat, to fossilized human poop.  But, carbon dating is only reliable for objects younger than 50,000 years or so.  And what if the thing you wanted to date never had any carbon to begin with?  What then?

    Here are four clever ways scientists figure out how old stuff is. Carbon dating belongs to a larger group of techniques known as radiometric dating.  These methods rely on measuring the ratio of abundance of two different atoms.  Sometimes, these atoms are two different elements.  Other times, it's the same element with different numbers of neutrons, called isotopes.  For carbon dating, it's the ratio of radioactive carbon-14, a carbon with eight neutrons, and stable carbon-12, a carbon with six neutrons.

    Because the rate at which the radioactive isotope decays is consistent over time, scientists can calculate how long ago a living thing died by measuring the ratio of carbon-14 to carbon-12, at least until too much of the carbon-14 has decayed for us to measure it.  That works best with living things, which continually accumulate both kinds of carbon over their lifetimes.

    But, we can still measure the age of other things, like rocks, if we use different elements.  One of the most common is potassium-argon dating, which is based around potassium-40 decaying into argon-40.  But that method operates under a few assumptions that don't always hold.  For one, it assumes no extra argon-40 was incorporated into the rock when it formed, and that you're only measuring argon that came from potassium.  It also assumes that none of the argon that was formed had a chance to escape the rock.  But those things can happen, and that messes with your date. 

    Some scientists have turned to a related, updated test.  And while it's a bit less direct, it does get rid of a lot of uncertainties.  This one compares the same argon-40 made from potassium decay against a lighter isotope of argon: argon-39.  They actually make the argon-39 by putting a sample of the rock they're dating inside a nuclear reactor, converting some of it's potassium-39 atoms into argon. Then, they determine the amount of argon-40 by heating up the rock so it releases the argon as a gas.

    This might sound weird, but what they're doing is really clever. See, the amount of argon-39 produced is directly related to the amount of potassium-39 it came from.  And, at least on Earth, the ratio of potassium-39 to potassium-40 is constant.  So, by measuring argon-39, you can do some math and ultimately work out how much potassium-40 was in your rock.  It's an indirect way of providing the same age that we would've used in potassium-argon dating, but it's a lot more accurate. 

    For one, you can measure both argon isotopes at the same time, rather than having to separately measure the argon and the potassium.  And, you can also check if your rock actually started with extra argon or lost some over time by slowly heating up the rock in stages and dating each stage.  If you get different ages, then one of your fundamental assumptions was wrong, and neither potassium-argon nor argon-argon, is going to give you a precise age.

    But this new method does have weaknesses of its own.  Because the amount of argon-39 created in the reactor depends on factors like how much potassium there was to begin with and how long the sample was irradiated, there are a bunch of variables that could throw off your measurement.  So, scientists have to compare it against a standard mineral they already know the age of.  Also, both potassium-argon and argon-argon dating methods require you to destroy your sample in order to date it, which isn't always ideal, but argon-argon requires a smaller sample.

    That makes this method better for stuff like moon rocks and asteroids, where we can't exactly hop back for more on short notice. In fact, this method is what actually helped us figure out that the rocks astronauts brought back from Apollo 11 are almost four billion years old.

    Now, other methods are less invasive, like electron spin resonance, or ESR, which uses magnets.  So, let's say you've got something a bit smaller than a pot, a human tooth perhaps, and your fossil site is too old for carbon dating, but doesn't have the right materials for other dating methods.  Tooth enamel is made of minerals, and minerals have a crystalline structure, meaning their atoms are locked into a rigid pattern.  That structure can trap electrons- some of which are jolted loose from their usual position by radiation, and some of which are part of that radiation.  The radiation can come from the sediment, from the tooth itself, and even from space. 

    They build up over time, so the longer the tooth has been buried, the more trapped electrons it has.  These trapped electrons collectively cause the tooth to react if you expose it to a magnetic field.  A moving charge creates its own magnetic field, and electrons are charged particles constantly moving around, so each is actually also a subatomic magnet.  Physics is just like that.

    When paramagnetic materials are put into an external magnetic field, all those lone electrons snap into the same position, aligning their own magnetic fields parallel to the external one.  It gets weirder.  Here's how scientists actually date that tooth. 

    Inside that magnetic field, they shoot microwaves at it.  At a specific wavelength, the energy of those microwaves will get absorbed by the trapped electrons, causing them to flip their magnetic field in the opposite direction.  That change of state is referred to as resonance- hence the name "electron spin resonance."

    By varying either the strength of the external magnetic field or the microwave frequency, you can measure the resonance signal given off by those electrons, and the intensity of that signal tells you how many trapped electrons you have, and how old your tooth is.  If the tooth's not too important, scientists actually grind the enamel into a powder and create a bunch of samples.  But if you'd rather not do that, you can take a small solid piece and perform the experiment over and over again.  

    ESR also works for other minerals associated with living things, like seashells, and even non-living things that grow, like stalactites.  ESR is best for dating minerals from between 50,000 and 800,000 years ago.  But, in ideal conditions, can go back two or three million years.  

    All of the methods we've covered so far provide absolute dates. There's some amount of error, but you get an actual number for how old something is.  But sometimes the best you can do is figure out is if something is older or younger than something else.  That's known as relative dating, and the most popular version is called stratigraphy.

    Stratigraphy relies on a couple of surprising common sense ideas. First, sediment tends to build up in horizontal layers called strata over time.  Consequently, anything found in lower strata generally has to be older than anything found in the strata above it.

    It's not quite that easy though- sometimes geologic events can disturb the layers so that sections of rock, including many strata, get shifted up or down relative to another section.  Erosion and volcanic activity can complicate things as well.  Assuming you've taken any of those events into account, there are a couple of ways you can use your strata for dating. 

    If you have a fossil in one layer sandwiched between two others, you have an age range that fossil has to be: younger than the fossils below it and older than the fossils above it.  Or, if you have known and unknown fossils from the same layers, you can infer that they lived about the same time, and get an age for the unknown species. 

    Stratigraphy isn't just used for geology or paleontology, of course.  It's used at archaeological sites as well.  For example, it's really hard to date Australian Aboriginal rock art, because the pigments the artists used don't have the isotopes used in radiocarbon dating.  So, in one instance, scientists determined the ages of rock art in western Australia by dating wasps' nests. Specifically bits of charcoal that had been incorporated into the nests.

    Since some nests had been buried beneath the painted rock walls, and some hadn't, they could get an age range.  They couldn't pin down an exact age, but the data suggested the art across 14 different sites was around 12,000 to 13,000 years old. 

    Finally, what if you can't date your sample by digging around a bunch of layers, or grinding it up and putting it in a machine?  Like stratigraphy, seriation can at least hint at how old a piece of art is, relative to a similar piece, made by the same culture at some other point in time.  Dating objects by their style was proposed back in the mid-1700's.  If you knew what features, for example, classical Roman statues tended to share at different points in time, it'd be fairly easy to assign a newly found statue to the appropriate time frame. 

    But what if you don't have that background knowledge?  Like, say you're in the 1800's with artifacts from a bunch of different prehistoric Egyptian graves with no clear ages.  Carbon dating hasn't been invented and stratigraphy doesn't apply.  Well, if you assume a culture's style changes gradually over time, you can compare how different style elements of a certain artifact overlap across the grave sites and place them in chronological order. 

    There's no way to tell which end of the order was the oldest and which was the youngest, but it's a start.  By necessity, seriation has to be extremely focused.  You can only compare the same kind of object against other versions because the way an ancient artist decorates a pot isn't necessarily going to be the same way another decorates a sword, even if they're made in the same year.

    You can also only compare the different versions of that object within not just one culture, but a small geographical region.  Also, you have no way of knowing if there was some outside influence by another culture that could have changed the style of your artifact, or if the culture itself went through a huge period of nostalgia for an earlier style. 

    But, for prehistoric art, seriation can be an important tool. Like stratigraphy, it's been used to date cave paintings, because what palaeolithic humans chose to represent, and the materials they used to do so, changed over time: dots and spirals, hand stencils, people, and animals drawn with varying levels of detail. 

    And sometimes those animals that are represented go extinct, so if you've got an old painting with a mammoth in it, it's got a good chance of being painted before the end of the last ice age. 

    Until humanity invents a time machine to actually go back and see when the things we want to study were made, we'll have to rely on clues like these.  So it's a good thing that we've been able to get so creative with the evidence we have.  It brings us closer to the past.

    Thanks for watching this SciShow list show, and thanks to our patrons for helping to make it happen.  If you'd like to join our awesome community of supporters, check out patreon.com/scishow.

(Outro)