Prehistoric chronologyPerhaps the most essential, most difficult, and most contentious aspect of archaeology is chronology. How old is the site you are reading about or, even better, excavating? Until the early 1950s archaeologists could only do this by comparing the evidence from one site with that from another. Suppose you had unearthed some decorated pottery. How did that compare with pottery from other sites, both locally and further afield?
The accepted theory was that peoples, building and tomb styles, and ideas in western Europe generally all had their origins in the east Mediterranean area. This is called diffusionism. Early archaeologists had studied Greek and Roman authors, and it was to the east that they looked for the origins of prehistoric monuments in Western Europe.
Libby's development of radiocarbon datingAll this was upset by the introduction of radiocarbon dating, developed by an American, Willard Libby, who produced the first dates in 1949. These astounded archaeologists by showing that many field monuments in the western Europe predated the Egyptian pyramids.
This method of dating measures the quantity of carbon-14 (C14) in organic matter such as wood or bone. All living matter takes in C14 when it is alive, but ceases to when it dies. The amount of C14 then diminishes, but at a set rate - it has a half life, now known to be 5730 years. This means that there is half the amount of C14 in the sample after that period.
Problems with C14
This was a huge leap forward in dating sites. But researchers soon realised there was a problem, in that, despite pushing the Neolithic, for instance, back by over a thousand years, dates were still coming out too young. Studies using samples which could be dated by other means, such as the study of tree rings or dendrochronology, showed that Libby had underestimated the half life. A calibration curve was developed and is still used to produce a more accurate date.
This means that a date of 3000 BC, for example, will be calibrated to about 3700 BC. In addition labortories accept an element of error, expressed as either a plus or minus, ie 3000 BC +/- 150, or a percentage, ie 95% probabilty the date falls between 3200 BC and 2800 BC.
Other problems ahve arisen. The amount of C14 taken in is affected by increased sunspot activity. Some periods such as the firts half of the first millenium give very wooly dates.
Bayesian Statistics
Although this gives a wonderful idea of how our dates fit into the broader pattern of prehistory, it only gives rather fuzzy dates. Can we get any more accurate? Recent work using methods originally developed by an eighteenth century clergyman, Thomas Bayes, have taken numbers of dates, and, using modern computer technology and 'number crunching', assessed them together. Using a system of inverse probability developed by the clergyman, some remarkable results have been achieved. Dates from excavated long mounds in the south of England have shown that burial stopped at four out of five in one year, 3625 BC. Exact dates are new in British prehistory. What happened in this year? Archaeology can't yet tell us, but perhaps, in the future, it might.
Sources
Colin Renfrew and Paul Bahn (2004) Archaeology: Theories, Methods and Practice
Current Archaeology 209 (2004), 9-20. The New Radiocarbon Dating Revolution
Ian Arthur Colquhoun - Bronze Age swords are my main area of interest. I was the co author of The Bronze Swords of Britain, a catalogue of all the Bronze Age ...
