[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: orion-list Radiocarbon discussion

To the list:

I was asked to comment on some statements made by Greg Doudna and maybe others
on the question of the radiocarbon measurements of the scrolls.

I think it is important that we all understand that the radiocarbon dating of
the scroll samples was limited by the amount of material which could reasonably
be removed (as was discussed by both Bonani et al 1991 and Jull et al 1995) and
further by the amount of pretreatment (cleaning) which could be done without
destroying the small samples we had for dating.

Since I am coming at this problem from the dating end, I will concentrate on
that angle.
I would comment specifically on some assertions made on the Orion list by
Doudna and Simms.

> > > Similarly the 4QQahat
> > > radiocarbon date is rejected by most Qumran scholars: it
> > > is significantly earlier than all other Qumran texts, and
> > > earlier than Qumran scholars' expectatations; so it is not
> > > regarded by most as accurate.  The principle is a sound
> > > one: in any battery of lab data, data points at one end or
> > > the other of distributions _must_ be regarded skeptically.

In a series of data on the same sample, this assertion is correct, but one
cannot apply it to a series of different samples and just reject the ones which
appear inconsistent with some other hypothesis.  I will return to the question
of age distributions later.  In my opinion, this result should be considered
skeptically because of the wide difference in ages reported by Bonani et al
(195) in their table 2 on this scroll and so has a suggestion of some problems
with pretreatment.

> > > In the Tucson battery in which 4QpPs(a) received its 1st
> > > century CE radiocarbon dating, 4QS(d) received a 2nd-3rd
> > > century CE radiocarbon dating.  There was nothing in the
> > > radiocarbon date itself, viewed in isolation, which would
> > > indicate this as suspicious.  The only reasons it was
> > > regarded with suspicion are because (a) it was later than
> > > all other Qumran literary text C14 datings, and (b) it
> > > did not agree with existing theories on the Scrolls.  It was
> > > retested, and the suspicions were justified.

I agree that in this case the unusual result precipitated the retesting to some
extent.  However, you will note that this sample also was one of those listed
as being contaminated with perspex glue.  If it were not removed, this glue
would be expected to be 14C-free, from petroleum, and hence give an apparent
"old" age.   However, one might also consider that those scroll fragments which
have been glued to rice backing paper may also have had other things done to
them.   The sample of 4Q171 pPs was considered to be cleaner in our report
(Jull et al., 1995)

> The second
> > > sample gave a dating in agreement with the other Qumran
> > > texts.  There are now two reported C14 dates for 4QS(d),
> > > and both cannot be correct.  There is no reason to reject
> > > the 2nd century CE dating other than that it is outside
> > > the pattern of other Qumran text radiocarbon datings, as
> > > well as outside of expectations on other grounds.

I think this indicates that we need to develop better criteria for cleaning
these samples.

>     In the earlier discussion noted, we learned that the
> >    modern text examiners often brushed the fragments with
> >    modern oil to highlight the writing.  No consistent
> >    record of who did this exists other than that it DID
> >    occur.  IOW, we have no consistent record of the type
> >    of handling of the fragments.

True, this would be helpful.

> >    There is a further explanation not discussed the first
> >    time around.  Some fragments may have been shielded
> >    from the ambient contamination of the centuries.  No
> >    review that I know of has addressed this issue for it
> >    can easily account for the width of the distibution
> >    curve of results.

I find it hard to believe that this is a problem.  If by ambient contamination
you mean physical, then these should be removed on pretreatment.  If you mean
chemical contamination (e.g. by oils) we can use solvent extraction to remove

>  But it is wholly
> > > inappropriate to assume no regional offset until proven
> > > otherwise.  It is appropriate to assume lack of knowledge
> > > on this point until there is knowledge.

You make a true statement, but it hides the fact that no regional effects in
this kind of radiocarbon sample has been observed with an offset in radiocarbon
age of more than 20 yrs.  Indeed (and this is an important point) this is about
the uncertainty in the calibration curve itself.

> > > In fact, the existence of a regional offset, possibly
> > > significant (i.e. up to the 20 years estimated possible),
> > > might already be suggested by existing Scrolls data.
> > > I refer to the five dates on the Bar Kochba texts.
> > > The distribution of dates was Zurich had two one way
> > > from the Seattle-Belfast curve, the Seattle-Belfast
> > > curve was in the middle, and all three of Tucson's dates
> > > were in the _other_ direction.  Making sense of those
> > > Bar Kochba dates is very difficult.

All of these samples were papyrus and not parchment.  I think this may be
important to understand these differences.

> >    The variability noted above would convince most re-
> >    searchers to set aside these dates in assessing the
> >    overall dating of the Scrolls' deposit anciently.
> >
> >    When two different labs contadict each other on 3 SD
> >    items, don't use the results.

I think we should consider what is reported.  The radiocarbon date provides an
uncalibrated age and a calibrated age range (using the tree-ring chronology).
The calibrated age ranges are quoted for 1 sigma and 2 sigma, which are the
estimated error range where 68% and 95% of all measurements of the same sample
from some central intercept of the uncalibrated "radiocarbon age" (which is
actually measured) and the tree-ring age on the calibration curve.  Hence, 5%
of measurements must fall outside the quoted 2 -sigma limits.  Hence, a quoted
calibrated age range is NOT infallible, but gives a best estimate of the
radiocarbon lab of the results.

There is in such things no substitute for more measurements, which if they
agree allow us to reduce the error, and if they do not, should tell us that
there is some problem.  Also, better sample preparation protocols might be
developed for this type of material.

I hope these comments are helpful,
Timothy Jull

Research Scientist, Uinversity of Arizona
and Editor, Radiocarbon

For private reply, e-mail to "A.J.T. Jull" <ajtjull@physics.arizona.edu>
To unsubscribe from Orion, e-mail to majordomo@panda.mscc.huji.ac.il with
the message: "unsubscribe Orion." For more information on the Orion Center
or for Orion archives, visit our web site http://orion.mscc.huji.ac.il.