On Tue, 16 Aug 2005 14:49:09 +0100, Irene Schrufer-Kolb 
<[log in to unmask]> wrote:

>Edgar,
>
>pottery can be dated more accurately than +/- 100 years, for example Roman
>Samian ware, which was subjected to a quickly changing fashion. Samian
>ware from Italy or southern France often shows potters stamps and the
>potters are known from securely dated (coins, C14) archaeological contexts
>to have operated only for a few decades in total and a particular Samian
>style was in fashion only for a couple of years. However, what Time Time
>try to get across is that pottery may have been in use for quite some time
>before it was broken/deposited, and this is especially true for the fine
>wares which were used only occasionally or were mostly for display.
>
>Metal objects on the other hand are mostly dated archaeologically, i.e.
>typologically, by archaeological context/stratigraphy, closely datable
>associated finds such as coins or certain types of pottery, or by
>scientific dating methods applied to other associated materials/features
>such as thermoluminescence dating, archaeomagnetic dating or C14 dating.
>It is however generally possible to date iron-based artifacts using
>radiocarbon dating. Have a look at
>http://www.tms.org/pubs/journals/JOM/0305/Cook-0305.html
>for this method. The obtained date ranges can be even greater than with
>conventional archaeological dating. The observed physical make-up and
>production techniques of metal artefacts are rather problematic for dating
>purposes as these functional factors changed only little over time. This
>is equally true for the typology of metal tools. Chemical composition can
>be a broad indicator to some degree as it is known for example that mainly
>Irish Copper and early Bronze Age artefacts can have a rather high arsenic
>content. This is however complicated by the reuse of scrap metal in later
>periods.
>
>So, overall, pottery remains more closely datable than metal objects and
>it is for this reason that pottery is one of the cornerstones of dating in
>archaeology. I think dating metal objects more closely is one of the
>archaeometallurgical research questions of the future and the attempts
>made with radiocarbon dating are an interesting move in this direction.
>
>Irene
>
>
>
>
>
>Dr Irene Schrufer-Kolb
>Associate Lecturer in Classical Studies
>The Open University
>Southern Region, Foxcombe Hall, Boars Hill, OXFORD OX1 5HR
>
>Tel./FAX: +44-(0)1280-823102

Thankyou for the interesting reply.
I should learn to put the brain in gear properly before posting.
What I was thinking about was the problem of dating other objects
with certainty from accurately dated finds within the same strata.
The URL I included mentioned that (as you say) some pottery 
can give very accurate dates such as samian ware, and coinage too, 
but one cannot be sure how long an object was in use before the 
owners either threw it away or buried it with some bones.
I suppose that this applies to any find.
An iron sword might have spent several generations in the hands of 
it's owners before it was smashed and hurled into some water or bog.
Nevertheless, the degree of accuracy remains important, especially 
when comparisons between objects of just a few years difference in 
age can be deduced. 

Thanks again,
Edgar