Posted by Steve Price on June 23, 1999 at 13:38:55:
In Reply to: Re: Sources of error in Carbon-14 dating? posted by James Allen on June 23, 1999 at 11:57:36:
Dear Jim,
I offer the following opinions:
1. Cases are never closed in academic circles. That's one of the things that makes us so irritating.
2. The fact that some highly respected person agrees with it is not a persuasive argument for the truth of an assertion. The basis for that person's belief might be very persuasive, of course. That's why the basis for his opinion is more important than his name.
3. There has been no argument on this board about whether pre-1700 Turkmen textiles exist. There have been inquiries about how we might know whether they exist. I believe such questions are reasonable and that they have answers that are not beyond the comprehension of some of our readers.
4. Having gotten similar results from analyzing twice is better than only doing it once, and a lot better than getting very disparate results on two tries. But twice is still a very small number of replications and no lab slob in my field would accept it as final.
5. The importance of reproducibility of results depends on the sources of error, which is why I have come back to that question several times. If the errors are from random sources (intrinsic to the instrumentation, variations between samples, weighing errors, etc.), then repeating the measurements on new samples eventually leads to a mean value close to the true value. That doesn't happen with systemic errors. For instance, my bathroom scale is about 2 pounds off, and will give you the same erroneous weight every time you step on it. That's an example of systemic error. Measurements can be very reproducible (that is, highly precise), but inaccurate.
There may a revolution in progress, but I think it is prudent to know the facts before accepting the conclusions. Those of us who think this way will miss the first couple of trains going into the future. On the other hand, most apparently revolutionary findings turn out to be flawed, so we don't spend a lot of time riding on trains that are going the wrong way, either. That's the tradeoff.
I still hope that some reader will explain what the sources of error in C-14 dating are, and give their approximate magnitudes.
Steve Price