Curse of 2200 BC

Timo Niroma (
4 Sep 1996 15:19:41 GMT

I've discussed the 2200 BC catastrophe in sci.archaeology. Hekla 4 was suggested the
reason, but I don't think the date does not match:

In article <50gcf2$>, (Timo Niroma)
>In article <502v29$>,
>Matthusen) says:
>>In <502oii$> (Timo Niroma)
>>>Could be in the range to fit the
>>>Finnish record, but not the Libyan-Egyptian-Mesopotamian-Harappan.
>>>Indus radiocarbon can't be that early. Naram-Sin can't be that early.
>>>Pepi II can't be that early. The Mesopotamian and Egyptian data
>>>a date of 2195+-5, so the Hekla 4 eruption happened at least 100 years
>>>before the 2200 catastrophe.
>>A paper by Dugmore et al (1995, "Radiocarbon dating tephra layers in
>>Britain and Iceland," _Radiocarbon_, vol 37, #2, pp. 379-388) indicates
>>that their C-14 work and the work by Hall et al. (1994) both indicate
>>that Baillie's 2345 BC event was before the eruption of Hekla 4.
>>Dugmore et al. indicate that the compilation of dates produce a
>>calibrated age range for Hekla 4 of 2317 to 2198 BC.
>>August Matthusen
A have this early morning comment, since I have fighted against the misuse of
radiocarbon dating over 10 years. (I will comment the snipped parts later.)

I have first an accurate citation from Baillie himself ("A Slice through Time",
1995, p. 78): "More recently (than Hammer 1980/TN) new radiocarbon dates for Hekla 4
began to suggest a considerably younger age, and a high-precision wiggle-match on
stratified peat associated with the Hekla 4 layer - identified in Irish peat by Jon
Pilcher and Valerie Hall - suggests that the true rate of Hekla 4 should be in the
range 2310+-20 BC, close to another narrowest-ring event at 2345 BC (Hall et al.

Actually the radiocarbon dates measuring times 4000 years ago should have an
accuracy of only 100-120 years (if using 2 sigmas, as is more customary and also
appropriate). So Pilcher and Hall should have at least 2310+-100 and Dugmore
2260+-100 years BC.

It's a tragic error to believe that the calibration, whether done by tables or
wiggle matching, should better the accuracy, it more likely adds a risk factor more
and the accuracy gets worse.

What the calibration is hoped to do is to make the most probable date as near the
true date as the method and data allow, but more accuracy in the sense of the
exact date it doesn't give.

If we take into consideration the success story of the 1628 BC date, suggested by
Baillie already in the late 1980s, I think that we can rely that the narrowest-tree
ring event of 2345 BC is accurate to one, or at most a couple of years, which is an
accuracy never possible to achieve with radiocarbon dating.

Timo Niroma