Loading ...
Sorry, an error occurred while loading the content.

Re: [ANE-2] Re: C14 dating of Ancient Egypt

Expand Messages
  • Ian Onvlee
    Hi Graham, I agree that this new C-14 study is once again not an independent study. The decission whether a date is an outlier or not is purely guesswork and
    Message 1 of 12 , Jul 19, 2010
      Hi Graham,

      I agree that this new C-14 study is once again not an independent study. The
      decission whether a date is an outlier or not is purely guesswork and relies on
      what is conventionally expected, hence circular reasoning. I have enough faith
      in C14-dates within a certain margin, say +/- 50-100 years, but I don't like the
      exclusion of so-called outliers to get the wanted results. They simply try to
      save face by doing so. I want to see all the results and decide for myself
      whether the data are sufficient to conclude anything specific.  

      Ian Onvlee 
      The Hague, the Netherlands

      From: Graham Hagens <rgrahamh@...>
      To: ANE-2@yahoogroups.com
      Sent: Sun, July 18, 2010 6:32:15 PM
      Subject: [ANE-2] Re: C14 dating of Ancient Egypt

      Moderators: apologies. I neglected to add my name & address to this submission.
      Please insert.


      Graham Hagens
      Hamilton, Ontario

      --- On Sun, 7/18/10, Graham Hagens <rgrahamh@...> wrote:

      > From: Graham Hagens <rgrahamh@...>
      > Subject: C14 dating of Ancient Egypt
      > To: ANE-2@yahoogroups.com
      > Date: Sunday, July 18, 2010, 9:38 AM
      > Apologies for the late response to this string related to
      > the article by Bronk Ramesey et.al. (Science 2010:
      > 1554-1557), it took some time to study the data closely.
      > This important article reports the results and statistical
      > interpretation of 211 radiocarbon (RC) analyses of material
      > associated with various Egyptian pharaohs between the 2nd
      > and 21st dynasties, a period of over 1500 years.  By
      > integrating these data with existing reign length models the
      > reach the conclusion that the conventional dating schemes
      > are reasonably secure.  At the outset they claim that
      > the statistical method they have developed is capable of
      > reducing the error range of ancient RC dating from about 5%
      > to 2.5% (p.1554).  Later in the text however the
      > accuracy of some dates near the beginning of the New Kingdom
      > are discussed within the range of a single decade: an error
      > range close to 0.5%.  Clearly a thorough critique of
      > the methodology employed to reach such remarkable success is
      > required.   This note identifies 4 areas of
      > concern which might be included in such a study:
      > 1. Provenance
      > None of the samples analyzed derived from recent
      > discoveries.  All derived from collections placed in
      > various museums during the 19th and early 20th centuries.
      > The authors admit that they ‘were reliant on the judgment
      > of excavators and curators and on the integrity of the
      > collections.’ To some extent it is possible to test the
      > validity of this assumption by comparing results obtained
      > from different collections which were attributed to a
      > particular Pharaoh. They are not always the same.  For
      > example, sets of samples associated with Pharaoh Hatshepsut
      > from 4 different  locations give the following
      > estimates of calendar dating BCE:  Paris 
      > 1600-1540 (4), Paris  1500-1440 (7),  New
      > York  1500-1440(13), Oxford  1250-1150(1). (Note:
      > IntCal calibration using  YBP data in Table S1; number
      > of samples in each set bracketed). 
      > 2. Outliers
      > Uncertain provenance combined with numerous sources of
      > contamination may well have contributed to the always
      > difficult process of identifying outliers in large sets of
      > data.  Over 20% of the data obtained were either
      > excluded from the model, or discounted in the calculations
      > due to suspicion of being be outliers.  For example a
      > set of data associated with Ramesses/Sety I was removed from
      > consideration because the RC data yielded a calendar date
      > 200 years higher than expected.  However data linked to
      > Ramesses II were included in spite of being about 120 years
      > higher than the modeled estimate (Table S1, 3240, 3145
      > YBP).   In another example, some data were
      > rejected for being too low: in the 17 point set 
      > associated with Akhenaten, 6 were identified being
      > outliers.    Without those outliers the RC date
      > (Table S1, 3050) would yield a calendar date close to the
      > 1350 BCE predicted by the model.  When included however
      > the calendar range drops more than a
      > century to ca. 1200 BCE. 
      > Uncertainties concerning provenance and contamination are
      > exacerbated by the uneven distribution of the samples: 
      > around 40% of the total number of analyses derive from just
      > two 18th Dynasty Pharaohs, Hatshepsut and Thutmosis
      > III.  Consequently the model employs a large number of
      > very small sets consisting of 1-3 data points with
      > concomitant high standard deviations. The further one
      > deviates from the high data density 18th dynasty, the more
      > uncertain the data provided by the statistical algorithm.
      > For example, only one set of results  from the
      > 20-21st  dyns, has more than 3 data points.  This
      > is the 11 values associated with Amenemnisu.  The
      > accession dating of this Pharaoh calculated by the model
      > of  1054-1044 BCE (Table S8) is about a decade higher
      > than the conventional value 1041.  Direct calibration
      > of those data provided in Table S1 (including the 19 year
      > regional offset correction), however, yields an accession
      > date of ca 1020,  below rather than above the
      > conventional dating 
      > 3. Circular reasoning 
      > The opening lines state that the authors used a ‘Bayesian
      > model incorporating historical information on reign
      > lengths’ to resolve problems result from ‘floating
      > chronologies linked to the absolute calendar by a few
      > ancient astronomical  observations, which remain a
      > source of debate.’ They go on to caution that for this
      > reason, none of  the dates calculated by the model
      > should be used to estimate reign lengths. 
      > The problem is that the reign lengths utilized in the model
      > depend to some extent on the very astronomical
      > observations  which the authors hope to displace. For
      > example, the estimates of many of the reign lengths within
      > the 19th-20th dyns are influenced by the belief that they
      > are anchored at the high and low ends by the astronomical
      > dating of the accession of Ramesses II, and the historically
      > attested invasion of Palestine by Shoshenq I.  Such
      > circular reasoning is exacerbated by the low density of data
      > within that timeframe discussed above.
      > 4. Regional offsets
      > The authors allow for a small regional correction (-19
      > years) to take into account a small RC offset detected in
      > plant material grown in Egypt between 1700 -1900 CE. 
      > However there is no good evidence that ancient offsets can
      > be estimated with such linearity.  The nature of
      > regional offsets remains poorly understood.  Various
      > contributing factors such as latitude, integration of
      > different  sources of the calibration curve, and the
      > influence of  old carbon from oceanic upwelling or
      > volcanoes have been identified.  Moreover some other
      > studies from this region, including those related to the
      > disputes concerning the destruction of Santorini have
      > suggested the existence of much larger offsets exist 
      > (p.1557 and refs.)
      > In conclusion, it seems unlikely that this study, for all
      > its strengths, is capable of resolving differences on the
      > decadal scale within the 3000-4000 YBP timeframe.  It
      > would therefore be premature to suggest that these new data
      > might be utilized to resolving  debates such as those
      > related to the dating of  Iron I/II Syro-Palestine,
      > where the competing positions are less than 5 to 8 decades
      > apart.

      Graham Hagens
      Hamilton, Ontario

      [Non-text portions of this message have been removed]
    Your message has been successfully submitted and would be delivered to recipients shortly.