Thanks to Jeanette for mentioning my participation in the early 1980s work
in Phoenix.  It may be useful to contact Cory Breternitz of SSI to see if
any copies of the reports on this work are still available (probably not).
My copies are in deep storage and not accessible.  Pat Garrow, who was also
involved in some of this work, may be able to provide some insights.

Bottom line, from my recollection, is that using this conceptual approach
requires a focus on artifact function, and South's general categories can
still be useful for the early 20th century.  However, as noted, some
artifacts served multiple functions, sometimes sequentially, sometimes
simultaneously, and attempts to incorporate these into the function
analyses can get complicated.  Since this is an analysis of function, it
would seem productive to run the analyses with these multi-function
artifacts placed in their various categories and examine the differences.
For example, in the Phoenix work on an early 20th-century bottling works,
we placed the bottles recovered (lots!) in the Kitchen category, and that,
obviously, really skewed the results.  Since we knew this was a small-scale
light industrial factory (beverage bottling), we placed the bottles in the
Activities Group, which produced what we called an "industrial pattern" (I
hope Jeanette will correct me if my memory is imperfect).

Good luck on your analysis!

Sue Henry Renaud, RPA