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

Re: orion Re: Date of Scrolls Deposit--55 BCE?

According to fred cryer:
> Please, Sigrid: 

This sounds patronizing to me.

> palaeographical date assignments are *relative*
> assignments, dependent on the baselines scholars have agreed on. The C-14
> dates are *absolute* date information; 

No. Barbara Thiering made that mistake at the Congress last July, taking
the single year suggested by C-14 as a fixed point of reference, and
dating other events in relationship to the C-14 date of her text. Read
Greg's post, or mine. C-14 gives a probability that the true date
lies within a certain range, not an absolute single date. The ranges are
getting smaller, an indication of greater reliability.

> they can indeed falsify claims,

When properly understood and properly used. The critical element is not
the *absolute or *relative nature of the data, but the way in which the
hypothesis is thought through and stated.

> whereas the relative dates based on the opinions of a few scholars cannot.

The first lesson in scientific method is to fix firmly in your mind the
four kinds of data. They are nominal, ordinal, interval, and ratio.
Relative dates based on the opinions of a few scholars are ordinal data,
similar to a consensus decision to classify pottery according to certain
gradations in size.

> The whole notion of the validity of the palaeography has not been tested by


Whether data is nominal, ordinal, interval, or ratio it can be separately
assessed as to its reliability, and also be assessed as to its validity. 
Reliability refers several things. Paleographical datings would be
reliable if a) there were inter-rater reliability--two or three
paleographers have good agreement among themselves as to dating individual
mss; or if b) the same group of paleographers showed a high correlation
between the ratings they made at time 1, and the ratings they made at time
2--called repeated measures reliability; or if c) ratings were internally
consistent. Smaller ranges for C-14 and AMS testing indicate somehting
like higher internal consistency of the test. Reliability is getting at
whether a measurement is consistent, and can be applied repeatedly. The
question of contamination of scrolls is a reliability question. 

Validity gets at whether a measure actually measures what it says it
measures--the lapse of time, for C-14 and paleography. Validity is
assessed in several ways. In general, both paleographical dating and
C-14/AMS testing have been externally validated using mss which have
internal dates. Tabulating the paleographers' datings and
correlating them with C-14 or AMS dating would be a test of the validity
of the two measures against each other. 

Such a correlation would give a single figure for the agreement of
datings given by two different measures. However, it would need to be
done with reliable measures, which can be repeated with the same results. 
Otherwise the correlation would not represent any of the "true" variance
in the underlying measures. It would shift with the testing of
contaminated vs. uncontaminated portions of the same piece of leather or
papyrus, as Greg has explained. 

Data is *all* relative, even the rate of decay of C-14. The "hardness" of
your *absolute data can't save us. To be more apodictic about this than I
feel, it is necessary to think through the problems of knowledge and how
we know what we know. 
 > best regards, 
 > Fred Cryer 

Sigrid Peterson  UPenn   petersig@ccat.sas.upenn.edu