Sorry if I'm jumpping into this one late, but this arguement on soil
stratigraphy has completely forgotten about the soils.  For sites which have
been buried, sealed and immune to the natural processes of pedogenics will
display a wonderful stratigraphy that most archaeologist will follow.  For
these kinds of sites (pedogenically dead), the Harris Matrix is a wonderful
empirical tool.  However, most sites which have undergone pedogenic influence
(and time is one of the greatest factors from the archaeological perspective),
the Harris Matrix is near worthless.  This factor of time may be applicable to
the perceived difference in approaches used by "historic" vs "prehistoric"
archaeology.  Cultural strata are still clear and evident in the pedogenically
dead historic site, but has often come under renewed pedogenic processes with
the substantially older sites.
 
Pedogenic stratification (meaning more than the simplified A,E,Bs,Bt,Bw, and C
horizonization, but also including the process of developing and migrating
stonelines and argilic horizons, pedoturbations etc...) is not the same as
cultural stratification.  Perceived cultural stratification in a dynamical
soil system are likely wrong from the start if the processes of pedogenic
stratification are not taken into account -- and thereby one is likely to
create falacious arguements with this data.
 
I do not in any case advocate excavations by arbitrary levels -- soils,
particularly pedogenically active soils, are not arbitrary; archaeologist,
however, have been.  By the same tokin, I do not belief that the strictly
emperical Harris Matrix should be adopted in all situations.  In fact, the HM
can only work well when the soils are not in a pedogenically active context.
For these living contexts, we must become sensitive to the nature and behavior
of the soil, and not myoptically think that all stratification is the result
of human behavior.
 
Douglas Frink
OCR Carbon Dating