Good question (and point).  Functional labels are fine as long as they
are used in context and with appropriate caveats.  We only run into
problems when we become entrenched in one or another view (but this
never happens to good scholars, right?).  We had a long thread on
histarch not long ago about moving buildings, which could open up the
whole question of how good is landscape analysis if the landscape
keeps changing; how good is any material analysis if materials are
reused, sometimes for other than originally intended purposes.  We
don't know much if we don't work out the context and sometimes we're
wrong anyway.  Like I tell my seven year old, being wrong is part of
being right, so we may as well enjoy it.
 
Sorry about your picture.
 
Richard Kimmel