While I agree with the point Bill Adams made (namely it can be difficult, if
not impossible, to use a small sample to make inferences about an entire site
or community), I'm continuously amazed at the ways in which privy deposits,
middens, and changing landscapes can be microcosms of much broader processes
that characterize a region or trends of a nation.  To provide an example, at a
site we recently investigated that was occupied from the 1830s-1980s, we found
a labor union pin.  Although we have had limited luck finding out much about
its owner, etc., other associated objects suggest that the pin was
used/discarded in the 1940s.  Now admittedly, this is a single object and we
probably haven't recovered a 1% sample from the site.  Nevertheless, I felt
compelled to state in the final report that the pin and the
affiliation that it represents clearly point to the need for wage labor outside
of the household, a trend that began in the 19th century in this part of MI.
 Moreover, this symbol is also evidence of the vitality of collective
bargaining and labor organization among
 industrial workers in the first half of the 20th century.  Labor union
participation in some industries during this period was a significant
proportion of the
work force, whereas in more recent years it has declined to a 20th century low.
Then, I really go out on a limb by suggesting that
declining union membership has come to be associated with larger proportions of
 homeless in the population and greater disparities in wealth and income
throughout America. (We had already made the point that the house was
transformed from an owner occupied residence to a tenemant in the 1930s.
 Thus, the plight of working class people nationwide is indirectly
 mirrored by the abandonment and ruin of the Shepard house.
 
All of this (the ills and ruin of late 20th century capitalism!) all from a
single union pin in a 1940s deposit?  You judge for yourself.  Does the
argument violate the logic of extrapolation?  I would say not. Economic
processes and social relations are, after all, created and reproduced at
multiple scales.  Why should the microscale be any less informative about the
whole than grandiose, multistage, multi-year, extraregional research?  Of
course, many of us might insist on the latter, but in the meantime, there is
much to be learned from the former, which are in effect, those small things
forgotten.
 
Michael Nassaney