The author calls the predicament faced by the Historic American Engineering Record "para- doxical" (lines 36-37) most probably because