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

orion Orion: Scroll Dating (short)

On Mon, 13 Oct 1997 12:27:26 GMT +100, GD@teol.ku.dk writes:
>Sigrid Peterson writes:

   [... snip ... already seen ...]

>> probability falls off symmetrically from the top of the bell curve. There
>> fore, the highest probability date for 4QpPs(a), by carbon 14 dating, is 
>> 58 CE. That is the halfway point between the two dates 5 and 111 CE. The 
>> probability that the true date is between 48 BCE and 5 CE is about 11%. 
>> The probability that the true date lies between 101 BCE and 48 BCE is 
>> about 5%.
>
>This is simply not correct Sigrid.  There is no systemic higher 
>probability in the middle, because the calibration curve is 
>nonlinear.  You are confusing a linear, Gaussian bell-curve which you 
>learned in your statistics classes with this.  This is hard for the 
>radiocarbon field to explain to the public.  Let me give an example.  
>Tucson dated three economic texts whose true dates are Bar Kochba, c. 
>130 CE.  It happens that there is a steep drop in the calibration 
>curve from 125 to 135 CE, and then a flat stretch after that more or 
>less for one to two centuries.  ANY text with a true date of 135 CE is 
>going to get a two-sigma range that extends no earlier than about 130 
>CE but will extend about one or two centuries into the 2nd and 3rd CE.  
>ANY texts with a true date anywhere in that 1-2 centuries range 
>are going to show the SAME calibration calendar date range.  The highest 
>probability is not in the middle!--not for radiocarbon dates.  That is a
>fallacy.  The highest probability is the entire range, versus outside the
>range, which is low probability. 

   Thank for that paragraph, Greg.  You have just nailed down the
   point I was trying to make on the List several months ago.  Now you
   bring up a second argument I use and give a very concrete example
   of it since it seems useless to use my "tobacco smoke contamina-
   tion" analogy with an group who spend any time in the near or
   Middle East, that is, CONTAMINATION!!!

>There is an unfortunate thing with these scrolls.  The early 
>team--Cross, Allegro, etc--were dousing these things in castor oil to 
>clean them.  They didn't know any better.  But none of the cleaning 
>procedures used at Zurich or Tucson (I have checked this with 
>chemists) would have taken out castor oil, if it was there.  It would be 
>undetected, and would give an anomalously younger (not older) 
>radiocarbon date.  There are some solutions to this which I won't go 
>into here, but in principle a way must be found to test for the 
>presence of castor oil (and not date samples which have it).  In the 
>meantime we hope that samples dated happen to have been 
>ones that weren't affected.  The fact that most of the 14C 
>dates have independently converged on 1st BCE is a strong 
>argument that contamination is not affecting most scrolls dates.  
>These problems are not insoluble, but there is trial and error in 
>this.  But the only way forward is forward--learn from previous 
>batteries and do it better the next time around.  But that's the way 
>science works. 

   I have argued partly successfully that the Shroud of Turin date was
   skewed by a similar error.  The recent Texas study went part way
   but there are more techniques to be tried.  Until they come along,
   the above material has to be considered useless for Carbon-14
   dating.

   Thank you, Greg, for bringing this before the List.  It is an
   effective corrective on the debate.  It also argues that most of
   the fragments testing young may now be considered early and trying
   to argue for a First Century of the Current Era date for the
   scrolls' creation will be an exercise in frustration.

   Again, thank you Greg for making that plain.

   [... snip ... already seen ...]

>   
>Greg Doudna
>Copenhagen

Tom Simms
Saint John, N. B.  Happy Thanksgiving or Harvest Home...