Of interest in this discussion is a recent article by Tom Neumann:
     
     Neumann, Thomas W.
     1993  Soil Dynamics and the Sinking of Artifacts: Procedures for 
     Identifying Components in Non-Stratified Sites. Journal of Middle 
     Atlantic Archaeology 9:139-153.
 
 
______________________________ Reply Separator _________________________________
Subject: Re: Unstratified sites
Author:  MIME:[log in to unmask] at INTERNET
Date:    6/18/98 9:01 AM
 
 
Edward B. Jelks wrote:
     
> Hi Dan,
>         The matrix of many, perhaps most, of the prehistoric archaeological 
si
tes
> in the SE U.S. are unstratified like those you describe.  I have dug many
> such sites, and wish there had been discernible strata to permit digging by 
> stratigraphic units.  But you have to play the hand you are dealt.  If such 
> sites are all you have to work with, you should learn as much as you can
> from them in spite of their shortcomings.
>         The only recourse in such cases is to collect artifacts from 
arbitrary
> excavation units--vertical and horizontal--and to look for distinctive 
> patterns in the distributions of artifact classes and types.  And, of 
> course, sometimes there are cultural features that provide association
> contexts.  But I have found repeatedly that older types of artifacts (dated 
> from contexts elsewhere) tend to occur deeper in such homogeneous deposits 
> than later types.  This can only be demonstrated statistically by seriation 
> graphs and the like, as the absence of strata precludes the possibility of 
> stratigraphic separation of types.
>         I have never been able to determine to my satisfaction whether the 
dee
per
> occurrence (statistically) of older types is owing to their being deposited 
> at a site while soil aggradation was going on, or whether they were
> transported downward mechanically from a stable surface by bioturbation, 
> vertisol fracturing, etc., the older materials ending up deeper simply
> because they were subject to the transportational mechanisms longer than
> more recent materials.  Perhaps both processes were going on at most sites. 
>  But whatever produced them, statistical differences in the vertical
> distribution of types covering a respectable span of time in such sites can 
> usually be recognized by plotting distributional data from arbitrary
> excavation units.
>         Also, when a site was utilized by different groups of people over a 
sp
an
> of several millennia, there often are demonstrable differences in the
> horizontal distribution of different artifact types and classes, reflecting 
> different spatial utilization of the site by the respective groups.
>         So, if there is good reason to dig an unstratified site--because, 
for
> example, there are no known stratified sites containing similar cultural 
> remains in the region--the only reasonable approach is to dig it by
> arbitrary vertical and horizontal units.  And the literature is fraught
> with reports showing that such sites can produce significant interpretive 
> results.
>
>         Ed Jelks
     
 Ed,
     
Funny you should mention this! I wrote a paper on this phenomenon, which I 
calle
d
"pseudostratigraphy" many years ago. I sent it to the state society's journal 
bu
t
they thought it was kind of "crackpot."  Clearly there are plenty of 
situations
in
which you can argue that deposits have built up over time. Other times it's 
not
that
clear. Nonetheless, you have--not stratitgraphy, of course--but "vertical 
separa
tion"
of older and younger stuff. I suggested that if items were deposited on a 
stable
surface subject only to bioturbation, some of these objects would move 
downward,
pushed by roots or falling into mouseholes, etc. So what determines how likely 
i
t is
something will move downward, or how far downward? Well, clearly, size and 
shape
count. Small, long narrow flakes, for instance, would more readily be moved 
than
would large flat ones. But the other variable of significance would be time. 
The
longer something sat around, the greater the probability it would get moved 
down
ward.
Likewise, the amount it would move downward would be dependent, in part, on 
time
.
     
Whatever other factors enter into the statistics, it is probably true that 
older
objects are more likely to have moved further down in a stable deposit due to 
bioturbation (or cryo- ot other types of turbation).
None of this is likely to be of interest to historical archaeology, but I am 
heartened to see I'm not the only one with such "crackpot" ideas :-)
     
Dan
     
--
Dan Mouer
"Pioneer"
http://saturn.vcu.edu/~dmouer/homepage.htm