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

Re: C14 dating of Ancient Egypt

Expand Messages
  • Graham Hagens
    Moderators: apologies. I neglected to add my name & address to this submission. Please insert. Thanks Graham Hagens Hamilton, Ontario ... Graham Hagens
    Message 1 of 12 , Jul 18, 2010
    • 0 Attachment
      Moderators: apologies. I neglected to add my name & address to this submission.
      Please insert.

      Thanks

      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
      >
      >
      >
      >
      >
    • 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 2 of 12 , Jul 19, 2010
      • 0 Attachment
        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.  

        Regards,
        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.

        Thanks

        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.