Following is a copy of the e-mail discussion that P.J. Reimer and I had during February–March 2003. The discussion was started after M.H. Wiener sent me a draft of an article that he had written about Aegean archaeological chronology. The article contained a criticism of a report that I had published in Radiocarbon [2002], and it cited a private communication with Reimer as its reference for the criticism.






From: Douglas Keenan
To: Paula Reimer
Cc: Malcolm Wiener
Sent: Sunday, February 23, 2003 10:08 PM
Subject: Hypothesis on too-early historical radiocarbon dates

Hello Paula,

Malcolm Wiener (CC'd) has sent me a copy of his Metron paper. This briefly discusses the hypothesis of my recent paper in Radiocarbon, and claims that the works of Facorellis et al. [Radiocarbon, 1998] and Siani et al. [Science, 2001] refute the hypothesis. The Metron paper cites a pers. comm. with you on October 17th as the source of this information.

On September 20th, I'd sent you an e-mail asking for comments on my paper, but didn't receive a reply. I would have preferred learning of your ideas directly, rather than via Malcolm! Anyway, I've explained to Malcolm why I don't believe that Siani et al. have evidence against the hypothesis.

Facorellis et al. is cited in my paper. Their work provides support for the hypothesis (the support is only marginal, though, because the work only considers dates before 7000 BP). Could you explain your reasoning here?

Cheers, Doug
 


From: Paula Reimer
To: Douglas Keenan
Cc: Malcolm Wiener
Sent: Monday, February 24, 2003 8:43 PM
Subject: Re: Hypothesis on too-early historical radiocarbon dates

Dear Doug,

I'm sorry that my comments came to you indirectly. Shortly after I received your e-mail I was asked to join in a reply with Sturt Manning and others to your Radiocarbon paper . I didn't think I should continue correspondence with you on the subject until that was published. That paper is still out for review so I probably shouldn't comment on it to any great extent, but there is no evidence of an increased reservoir age prior to ca. 8200 cal BP in either the Facorellis et al. or Siani et al papers. If you look at the sediment record for the Eastern Mediterranean you'll see there have been no recent events documenting decreased ventilation since the S1 sapropel evident ca. 8200-9500 cal BP (Mercone et al.,2000, Paleoceanography 15:336-47).

Once again, my apologies for discussing this with Malcolm without informing you.

Best wishes,
Paula
 


From: Douglas Keenan
To: Paula Reimer
Cc: Malcolm Wiener; Sturt Manning
Sent: Tuesday, February 25, 2003 12:27 AM
Subject: Re: Hypothesis on too-early historical radiocarbon dates

Dear Paula,

> I'm sorry that my comments came to you indirectly. Shortly after I
> received your e-mail I was asked to join in a reply with Sturt Manning and
> others to your Radiocarbon paper . I didn't think I should continue
> correspondence with you on the subject until that was published. That

Kind thanks for explaining.

> paper is still out for review so I probably shouldn't comment on it to any
> great extent, but there is no evidence of an increased reservoir age prior
> to ca. 8200 cal BP in either the Facorellis et al. or Siani et al
> papers. If you look at the sediment record for the Eastern Mediterranean
> you'll see there have been no recent events documenting decreased
> ventilation since the S1 sapropel evident ca. 8200-9500 cal BP (Mercone et
> al.,2000, Paleoceanography 15:336-47).

Facorellis et al. state that their evidence "may imply a change in the marine reservoir effect in the region after the 8th millennium BC" of ~ 200 yrs. This contradicts your claim of "no evidence". Your claim that Sturt Manning (CC'd above) thinks otherwise is confusing, as he said the following to me recently:

... I read your Radiocarbon paper p.227 and its citation of Facorellis et
al. I then read that paper. It tells us that between ca. 8200 - 9500 cal
BP reservoir ages appear to be larger at around 515 +/- 22 14C yrs
(Facorellis et al. 1998). This increased reservoir age corresponds to ...

I can understand why some people might believe that Siani et al. [2001] have evidence against the hypothesis. I wish that I had seen their work in time for its discussion, and I've replied to Sturt and Malcolm about why it doesn't matter.

Your statement about Mercone et al. contradicts what the authors themselves say. It also ignores the extensive related discussion on this in subsection 2.3.

Your interpretation of Facorellis et al., etc., is not something I understand. I realized that many people would not like the hypothesis. I hoped that those people would still be willing to address the evidence as it has been presented.

Lastly, I find all of this truly bewildering. If the true goal is to get closer to scientific truth (?), then I would have hoped for discussion between all of us prior to submission of the reply to my paper. (I've done this thrice myself.)

Sincerely, Doug
 


From: Paula Reimer
To: Douglas Keenan
Cc: Malcolm Wiener; Sturt Manning
Sent: Tuesday, February 25, 2003 2:05 AM
Subject: Re: Hypothesis on too-early historical radiocarbon dates

Dear Doug,

I'm afraid my hasty reply has confused things. I agree that Facorellis et al. show a larger than modern reservoir age around the 8th milennium which is approximately the time of "the most recent sapropel event" according to Mercone et al..2000. There is no evidence for a more recent period of decreased ventilation in the sediment Ba/Al ratio which is an indicator of productivity either. Measurements on foraminifera from ash layers (Siani et al. 2001) and a paired shell and terrestrial material (Siani 1999) show the reservoir age had decreased to near modern values during the period from about 5400 to 3800 14C yrs BP. There is unfortunately no reservoir correction data between 3800 14C yrs and ca. AD 1800, but it seems improbable that the reservoir age would increase significantly again without affecting the Ba/Al ratio or leaving a record of more recent sapropel formation.

Sincerely,
Paula
 

(Reimer's responses seemed very confused to me. My suspicion was that her understanding of the issues was poor. To test this, I next asked a basic question, which could not be easily talked around.)


From: Douglas Keenan
To: Paula Reimer
Sent: Tuesday, February 25, 2003 10:02 PM
Subject: Re: Hypothesis on too-early historical radiocarbon dates

Paula,

I'd like to clarify something. What do you think is the end date for S1, and how do you think this differs from the date given in my paper?

Cheers, Doug
 


From: Paula Reimer
To: Douglas Keenan
Sent: Friday, February 28, 2003 1:21 AM
Subject: Re: Hypothesis on too-early historical radiocarbon dates

Doug,

According to the Ba/Al data of Mercone et al., 2000, the sapropel event was over by 5300 14C yr BP (marine) which with Delta R of zero calibrates to cal BC 3800 -3600 (2 sigma) or if I use the average Delta R value calculated by Facorellis for the 8th millenium, to cal BC 3660 - 3400. The date you quote from Thomson et al., 1995 of 4500 BC is from the visible dark bands and high organic carbon content, so, because of carbon degradation may actually underestimate how long the event lasted. However, I would argue that the reservoir age decreased to modern values even as the sapropel event was ending. Siani et al., 2000, show the reservoir age back to modern by 6000 14C yr BP (terrestrial) or cal BC 5030 - 4730.

Cheers,
Paula
 


From: Douglas Keenan
To: Paula Reimer
Cc: Malcolm Wiener; Sturt Manning
Sent: Saturday, March 01, 2003 12:00 PM
Subject: Re: Hypothesis on too-early historical radiocarbon dates

Paula,

> According to the Ba/Al data of Mercone et al., 2000, the sapropel event was
> over by 5300 14C yr BP (marine) which with Delta R of zero calibrates to
> cal BC 3800 -3600 (2 sigma) or if I use the average Delta R value
> calculated by Facorellis for the 8th millenium, to cal BC 3660 -
> 3400. The date you quote from Thomson et al., 1995 of 4500 BC is from the
> visible dark bands and high organic carbon content,

The date quoted from Thomson et al. is based solely on Ba/Al. The same holds for van Santvoort et al. (also cited). These two papers were the first to use Ba/Al, and they demonstrated why Ba/Al holds as the redox front progresses. Mercone et al. rightly cite both as their primary references for the technique.

The date given in my paper is "after 4500 BC--possibly many centuries after". If it is as many centuries as you say, then that strengthens my hypothesis.

You haven't cited evidence from Mercone et al. that is against the hypothesis.

It is amazing that you, Sturt et al. would publish a reply to my paper without first reading either my paper or the other papers that you cite against it.

Sincerely, Doug