Absolute dating: More physics tricks

Introduction to Archaeology: Class 6

Absolute dating: More physics tricks and dating in historical archaeology

ã Copyright Bruce Owen 2002

More physics-based dating methods

Potassium-argon dating

Like radiocarbon, this is a radiometric method

Depends on the decay of radioactive isotopes
This is absolutely regular, not affected by temperature, humidity, etc.

Potassium in nature is partially ⁴⁰K, which decays to ⁴⁰Ar (stable, inert Argon gas)

Potassium is a very common constituent of many rocks
In molten rock, the argon gas is free to diffuse away
So when the rock solidifies into crystals, it starts off with no argon
The ⁴⁰K decays, producing ⁴⁰Ar that is trapped in the crystal
So the ratio of trapped ⁴⁰Ar to remaining ⁴⁰K increases over time, allowing calculation of the time since the crystal solidified

Unfortunately for archaeologists, people have rarely lived in solidifying molten lava

But if a site was covered by a lava or volcanic ash fall, that would give a date of the site's destruction
Or if an artifact is found in the sediments above one basalt flow but below another, they will bracket the artifact

But ⁴⁰K has a very long half-life (1.3 billion years), so it takes a very long time for enough ⁴⁰Ar to build up for a reliable measurement
So this method is good for old samples, starting at about 1 million years old, up to beyond the age of the earth's crust (lunar rocks at over 5 billion years old)

Geochron lab will not date samples expected to be under 2 million years old
While Thomas says it can be used up to just 200,000 years ago
He is probably thinking of some very experimental attempts to stretch the method to younger samples
Currently not very useful to archaeologists other than those working with pre-Homo sapiens sites

Cost

Geochron lab: $400-$600 per sample, depending on preparation

Under 5 million years expected age has 50% surcharge for extra care in handling, since contamination is more of a problem with such tiny amounts of argon

Argon-argon

A clever modification, but not widely mentioned

Obsidian hydration dating

A wonderful tool in that date is tied to a specific human behavior we might be interested in: flaking stone to make tools
review the method; Thomas's description is OK
The hydrated "rim" is also sometimes called the "rind"
Thomas says obsidian "one day may rival ceramics as archaeology's most useful artifact for controlling time"

Not a chance
In many places obsidian is relatively rare
"dates" are pretty approximate

the uncertainty is often not even mentioned, because it is hard to estimate
results are usually not even converted to dates in calendar years, but are given as rim thicknesses in microns

with one or more estimates of the relevant rate (years per micron)

typical examples:
220 ą44 years/micron
458 ą92 years/micron

the reader can calculate calendar ages, but the writer does not imply that these are necessarily right

say, 2.0 microns X 220 ą44 years/micron = 440 ą88 years old (68% chance that the true age is between 352 and 528 years)

rim thicknesses are often used for rough age bracketing and relative chronology
in short, generally thought to be much less precise than radiocarbon dates

inherent imprecision of any method that is based on chemical reactions

these are affected by temperature, often by humidity, and other factors
it is rarely possible to assess these, or to know that any two artifacts experienced the same conditions

so their rates of reaction may not have been quite the same

say two piece of obsidian are flaked at the same time, but one lays on the surface in the sun for years, while another is quickly scuffed under a little soil

the one that got hot a lot hydrated more quickly, and will give an older date
but they may both now be buried in the same site, near each other, with no trace of their different histories

it is now routine to chemically "source" the obsidian before calculating the age of the rind

the chemical compositions of different sources or quarries vary

although the composition within a given deposit (a single source) is relatively uniform

in many places, like California and increasingly Peru, most or all the major sources are known and have been chemically characterized
and the hydration rate is more or less known for each
so a piece from a site can be matched to its source

sometimes it is possible to do this visually; often it is not

and the appropriate hydration rate used to calculate the date
if you don't chemically or visually identify the source of the obsidian, you have no idea what hydration rate to use - and they vary widely

another method (which has been used here at SSU, I believe) involves measuring the hydration rate of each piece experimentally

called "hydrothermal induced hydration"
first you measure the rind on an ancient broken surface
then you break off a fresh surface and put it in a humid chamber at known high temperature and pressure
it will hydrate rapidly
there is a pretty solid model of how temperature and pressure affect the hydration rate
by measuring the thickness of the hydration rind after a known time at a known high temperature and pressure, it is possible to calculate what the hydration rate would be at a more typical temperature and normal atmospheric pressure
then you use this rate

still another method for getting the hydration rate of a given fragment is the "intrinsic water" method

based on a model of the variables that affect the hydration rate
in this, the main variables are the "intrinsic water" of the obsidian piece, the relative humidity, and the effective hydration temperature
by measuring the "intrinsic water" of the obsidian and measuring or estimating the other two (which are conditions of the site, not the artifact), the hydration rate is calculated
the relative humidity and effective hydration temperature for a given archaeological context is measured by burying probes in the layers that the obsidian artifacts came from and returning months or a year later to collect the readings.

cost

measuring the rind thickness is cheap

at Diffusion Lab, $30 - $20 each (the price goes down for larger numbers of samles)

sourcing or determining the hydration rate may not be, but it may only be necessary to do this for a few samples, then apply the result to many others

at Diffusion Lab, $700 for induced hydration
at Diffusion Lab, $50 each for measuring intrinsic water

since each date is cheap, but not very precise, a common strategy is to run lots of them

and use statistics to cut down on the uncertainty
being able to date many sites and layers poorly may still be useful for getting the overall picture

Amino acid racemization dating

A chemical process, rate depends on temperature, possibly other variables
Thomas's description is OK
works (if it does!) on well-preserved organic materials that have been well enough studied to estimate the racemization rate under different circumstances

that is, mostly bone and ostrich egg shells

This method was a spectacular failure in the early 1980s, when it was used to "date" human bone from burials in Sunnyvale, San Diego, and elsewhere to long before the end of the Pleistocene

When AMS radiocarbon dating was developed, small bits of those same skeletons were dated and gave much more recent, expected results

Presumably the racemization rate was affected by something, but the original authors had been careful to make reasonable estimates of temperature in the surrounding soil

Who knows what the problem was?

But given that history, people are likely to be skeptical of racemization dating

Thermoluminescence ("TL") dating

Crystals have an orderly arrangement of atoms in them
If a high-energy particle from cosmic rays or radioactive decay passes through the crystal, it can disorder the crystal's structure
this damage or trapped energy remains permanently in the crystal, unless it is heated to a fairly high temperature
If the crystal is heated, the extra vibration of the atoms allows them to return to their orderly arrangement

And the trapped energy that is released escapes as a very faint flash of light

with the proper instruments, it is possible to measure the brightness of this light

hence the amount of damage or trapped energy that has accumulated in the crystal either since it formed or since it was last heated

If you know the amount of radiation that the crystal was exposed to per year, then the amount of light emitted tells you how many years the crystal has been accumulating damage since it formed or was last heated
This is useful for archaeologists, because some human activities involve heating crystals

Firing ceramics that contain mineral grains in the clay
Heating certain kinds of chert or other stones to improve their flaking qualities before making a stone tool out of them

only certain cultures at certain times have done this

in theory, this should allow us to tell how long ago a pot was made, or a stone was heated before flaking it.
but we also need to know the annual "dose" of radiation that the object received

this is estimated by going to the excavated site and burying probes in the same layers that the artifacts came from
you return a year later and measure how much radiation they received from the soil around them

this works OK for objects that have been buried in the same conditions ever since they were made

but if a potsherd lay on the surface and got blasted with solar radiation for a while before being buried...
then the total amount of damage to the crystals will reflect an unknown number of years on the surface (with an unknown annual "dose" during that time)
plus some time in the ground at the known dose rate.
in this case, there is no way to calculate the age

it can be difficult to be sure that the object really has been buried in constant conditions all along
the actual uncertainty of TL dates is hard to estimate

but by comparing to other methods, it seems that TL dates on single sherds from good contexts are said to be roughly ą 15% of their age

like 500 ą75 years, or 1000 ą150 years

by averaging many sherds from a single context, TL dates may be as good as ą7 to 10%

like 500 ą50 years, or 1000 ą100 years

but again, this method is more problematic and less trusted that radiocarbon

and there are others that work in certain special circumstances

Paleomagnetism
Fission-track dating
Electron-spin resonance dating
etc.

More on dating in historical contexts

Pipe stems

Note that Binford's version is specific to two decimal places: 3.65 days!

nails

a lot more informative than Thomas indicates
due to a series of technical advances in manufacturing, mostly documented with patents and sales literature
the problem is that each improved nail machine did not cause the older ones to be abandoned
so again, nail types only provide a terminus post quem (date after which they must have been made)

"Documentary evidence", in Thomas's terms

Thomas focuses on pre-industrial artifact types

Reconstructing relative and absolute dates from paintings

But also from factory records, sales catalogs, patents, etc.

Back into the 1700s, if not earlier
Especially factory-made ceramics with painted, printed, and/or molded patterns that changed rapidly with popular fashion, or technical innovations
Often small factories would put a mark on the bottom of plates and other vessels that would change a little every few years

As owners changed, etc.
These can often be dated to within just a few years