Are the Pacajes phases temporally discrete?

As long as I am discussing the Pacajes ceramic phases, I am compelled to address an issue that was raised by the original formulators of the Pacajes ceramic sequence. In their survey report, Albarracín-Jordan and Mathews suggested that the Pacajes-Inka phase may in fact have overlapped temporally with both the Early Pacajes and Late Pacajes phases ([Albarracín-Jordan and Mathews 1990], Figure 14). In this view, Pacajes-Inka phase ceramics represented a foreign stylistic intrusion into the southern basin, which coexisted with both late Early Pacajes ceramics and early Late Pacajes ceramics. Mathews argues that certain ceramic attributes, such as disk bases and everted rims, are present in both Early and Late Pacajes ceramics, but not in Pacajes-Inka ceramics.

In essence, this argument requires the assumption of the contemporaneous manufacture of two distinct ceramic types, emerging from two separate cultural traditions, in the Tiwanaku Valley. Specifically, we need to posit that the Inka ceramic material represented a tradition that was superimposed over over the existing cultural matrix, rather than one which replaced it completely. Inka-Pacajaes could have represented an elite or sumptuary class of ceramic which was available only to certain elements of the indigenous Pacajaes society, possibly determined by position within the Inka administrative hierarchy of indirect rule. At the same time, commoners continued to produce the diagnostic Early Pacajes wares described above, for everyday domestic consumption. [[Mathews 1992]: 194]

This is actually a straightforward hypothesis to test. If Mathews's suggestion is correct, then a high percentage of sites with Pacajes-Inka ceramics will also have Early Pacajes ceramics on the surface. On the other hand, if Pacajes-Inka ceramics are partially contemporary with Late Pacajes ceramics, then a high percentage of sites with Pacajes-Inka ceramics will also have Late Pacajes ceramics on the surface. If neither one of these it true, then no such relationships will be observed.

Table 10.1: Percent of Pacajes-Inka sites also having other Pacajes occupations

	Taraco Peninsula	Lower Tiw Valley	Middle Tiw Valley
Early Pacajes	25	88	94
Late Pacajes	40	58	36

Table 10.1 displays these data for the Taraco Peninsula and for the Lower and Middle Tiwanaku Valley. The Tiwanaku Valley data were derived from the dissertations of the Tiwanaku Valley researchers ([Albarracín-Jordan 1992] Appendix 2, [Mathews 1992] Appendix A). The Early Pacajes value for each area is equal to the complement of the site founding index for the Pacajes-Inka phase. The Late Pacajes value for each area is equal to the occupation continuity index for the Pacajes-Inka phase. See Chapter 4 for a discussion of these indices.

It is easy to see how Mathews could think that Early Pacajes and Pacajes-Inka ceramics were partially contemporaneous. In his sample (the Middle Tiwanaku Valley) 94% of Pacajes-Inka sites also have Early Pacajes ceramics on the surface. However, only 36% of Pacajes-Inka sites have Late Pacajes ceramics on the surface, clearly showing that Pacajes-Inka and Late Pacajes ceramics are sequential and not contemporaneous. The Lower Tiwanaku Valley data are similar, though the relationships, both positive and negative, are weaker.

However, the Taraco Peninsula data show the opposite relation. That is, only 25% of Pacajes-Inka sites on the Taraco Peninsula have Early Pacajes ceramics on the surface. And only 40% of Pacajes-Inka sites on the Taraco Peninsula have Late Pacajes ceramics on the surface. This means that the Taraco data that Early Pacajes and Pacajes-Inka ceramics are sequential and not contemporaneous. The association between Early Pacajes and Pacajes-Inka occupations in the Tiwanaku Valley must be due to some other factor. My personal belief is that there are so many Early Pacajes sites in the Middle Tiwanaku Valley (over 500), that there is an Early Pacajes phase occupation of the vast majority of the sites of any phase. That is, there is an Early Pacajes site in virtually every habitable location. Whatever the case, this exercise has conclusively demonstrated that the three Pacajes phases are in fact sequential and have no appreciable temporal overlap.