[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: orion Radiocarbon dating
According to Greg Doudna:
> Dear orioners:
> I don't know whether there has been a gap in my education, or the
> synapses just aren't firing today, but I honestly cannot understand
> what Sigrid is talking about in some of the post, the jargon and the
To a humanist, who only knows the critical examination of data, it is jargon.
To a scientist, the four kinds of data are not jargon, they are the ABCs. The
goal of a scientist is to produce a study that can be replicated, and
understanding the scientific treatment of data is basic.
I think there *has been a gap in your education. I was testing whether you
knew what you were talking about, the statistics of it, when you took me to
task for explaining probabilities according to the symmetrical bell-curve, the
term you had introduced. The probabilities of the bell-curve are as I said;
it is not the term to use for the probability distribution of C-14 data that
you described. I was using terms I thought you might recognize as terms that
*do* describe the probability distribution. That is, when graphed, they do
*not* show the shape of a bell curve. I hoped you knew the name of the
non-normal distribution you described, so I could look up its properties and
get back into the discussion of how C-14 data works.
> However what I do understand, in essence, is that there is an
> objection to interpreting radiocarbon dates differently after the fact,
Yes. This objection stands, as long as you say you are doing "scientific"
history, rather than fuzzy-minded humanistic history. I am not that concerned
which kind of historian you are; I am very concerned that you are hiding
humanistic judgments in scientific clothing. I am working so hard at this, not
to pick on you, Greg, but to push you towards accepting the necessity of
being conscious of your methods and their implications. You are grappling with
the problem of how far C-14 data can take you. You hope, in my reading, that if
you know this kind of data, understand it really well, you won't have to make
any risky (read "paleographical") judgments about dates. Instead, C-14 will do
it for you. Greg, you have to develop the courage to be wrong. Or else learn
scientific method, and develop the courage to treat data in such a way that
your studies can be replicated. Or even do both, in different studies. :->
> But so far as I can tell, Sigrid isn't objecting to a particular
> sources of tiny errors that can throw a date off. To say that a date
> that comes out at one end has the same weight as a date that agrees
> with well-established patterns is nonsense.
No, it isn't. I am insisting that you use data in a repeatable method. Let me
illustrate the kind of thinking I'm talking about.
I have just tabulated de Vaux's notes (Humbert & Chambon, 1984) with respect
to the gender of the remains found in the graves excavated at Qumran. I found
that 22 of the excavated individual remains of 44 people in 41 graves had no
gender determination to be found in de Vaux's notes. That is precise, and
someone else can repeat it. It is replicable. I did it because I was looking
for evidence of women at Qumran. I had to throw out some cases where the
remains were later identified as female, as well as some where they were later
identified as male. Some, many, or all of those identifications might be valid
and repeatable; we just don't know, because the grounds for them were not
recorded, except for the remains in T.7, who had a large pelvis, and who was
short in stature, the only measured skeleton. De Vaux later said it was the
skeleton of a woman (1973). However, it's not in his notes.
Linda Bennett Elder did a study (BA, 1994) where she searched out all the later
determinations. However, she didn't have de Vaux's notes available, and there
are some contradictions and ambiguities between her research and de Vaux's
now-published notes. I could resolve those problems by recording my decisions
about each set of remains; however, I was struck by the block of 20 graves with
no record of gender, and am very skeptical about de Vaux's later assignments of
gender to those graves in the "main cemetery." I don't think my own
determinations of gender for that block of 20 graves would have much
foundation. No one would believe a 50-50 division into 10 women and 11 men (2
were buried together), though it would agree with the population percentages.
My a priori decision was that the remains are gendered as in de Vaux's notes,
and as in Steckoll's (1967) clear and explicit results. That's a scientist's
Anyone tabulating de Vaux's notes on the graves would come up with similar
results, becuase the criteria are clear cut and comparable. I didn't call a
body female because it measured 1.6 meters, since I don't know any of the
other measurements of other bodies. Some sets of remains were reburials, and
others had been damaged by rocks, so not all could ever have been measured. I
couldn't judge that T.7 contained a woman's body based on the larger pelvis,
because a picture of the skeleton in T.5 also shows a fairly large pelvis.
Other pictures showing bodies having a small pelvis, apparently, turn out to
have been in the dorsal position, or they are reconstructions.
So I exclude the undecidable cases, but tabulate them as undecidable. There are
clear physical reasons that make them undecidable.
What I didn't do was say "I think there must have been more women,` and to show
that there were more women, I will accept some of de Vaux's later decisions,
that some of the corpses were female, and not accept other decisions, that some
of the corpses were male." Then I would have had *absolute* hard data, none of
it relative, that proved that there was a sizable presence of woman at Qumran.
> (So I wonder why Sigrid is picking on me!)
> But I do not wish to continue, and orion readers may cheer this
> sentiment also :-)
And I didn't wish to start. :-)
>Many of these issues and questions will be better understood by some resources
Thanks for the resources, Greg. I wish I were working with Qumran materials and
had some use for C-14 info. I'll try to check out the statistical parts.
> Greg Doudna
All the best,
Sigrid Peterson UPenn email@example.com