As with the Early Pacajes phase, my definition of the later Pacajes phases closely follows the chronology formulated by Albarracín-Jordan and Mathews in the Tiwanaku Valley survey ([Albarracín-Jordan 1992,Mathews 1992]). I have found their ceramics sequence to be correct and, almost as importantly, easy to apply to mixed surface assemblages. I will extend their phase definitions only slightly to include features and decorative motifs which they did not report. In particular, I will expand the decorative repertoire of the Late Pacajes phase, of which I appear to have a much larger sample than did the authors of the original descriptions.
The identification of the later Pacajes phases relies mainly on decorated wares. Decorated diagnostic ceramics are abundant on the surface of Late Horizon and Early Colonial sites in the study area, however, so this presented no practical difficulties.