Late Formative 2

Janusek dates the LF2 from 300 A.D. to 500 A.D. ([Janusek 2002]). I have retained these dates in the present analysis.

Considered in ceramic terms, the shift from LF1 to LF2 consisted of the gradual disappearance of the Kalasasaya red-rimmed bowls, the abrupt replacement of the Kalasasaya zoned-incised complex with a new set of decorated ceramics grouped under the name ``Qeya'', and a partial change in plainware pastes.

I consider three groups of ceramics to be diagnostic for the LF2. They are:

1. Qeya polychrome ceramics,
2. Qeya incised ceramics, and
3. a distinctive thin, reduced micaceous ware I called Paste Group 13 in my ceramic analysis.

Qeya polychrome ceramics were first noted by Bennett as a stylistic component within his Early Tiahuanaco phase ([Bennett 1934]). The name ``Qeya'' was apparently attached to the style by Dwight Wallace ([Wallace 1957]). It refers to the site of Qeya Qollu Chico on the Isla del Sol, where Bandelier collected some specimens of the style early in the century ([Bandelier 1910]; photographs of some specimens collected by Bandelier have been published in [Bauer and Stanish 2001], Figures 4.4-4.8). The Qeya style is quite elaborate, with a wide range of unusual vessel shapes and iconography. Forms include burnished, scalloped annular bowls often with modeled feline heads on the rim, kero-like cups, and tall, flaring-rim bowls called escudillas. These vessels are frequently decorated with a variety of geometric, zoomorphic and anthropomorphic motifs, including a front-faced figure with rays radiating outward from the face, clearly prefiguring the Tiwanaku Gateway God motif.

Qeya polychrome ceramics were formally defined by Dwight Wallace ([Wallace 1957]). Wallace characterized them as having a fine, light-colored paste, sometimes micaceous, unslipped and painted in a variety of colors. Qeya ceramics also exist in a variety of unusual forms. From the perspective of the survey archaeologist, however, Qeya polychrome ceramics may be summarized simply as ``polychrome on unslipped buff''. Fragments of these vessels are quite rare on the surface, and they seem to be restricted both regionally and locally (that is, on both the intra- and inter-site scales), as was the case in the LF1. Examples recovered from the survey are illustrated in Figure 7.1m-q.

In addition to the polychrome on buff sherds, some black on red sherds are also present in this phase (see [Albarracín-Jordan 1996a]: 131 and Figure 7.7; Figure 7.1r-s). One yellow on black rim was recovered which is probably LF2 in date (Figure 7.1t). These sherds do not fall within Wallace's definition of Qeya, but they do seem to pertain to the LF2.

Qeya incised ceramics were also identified by Wallace. These sherds display a very closely-spaced pre-firing incision, typically comprised mostly of parallel lines. This is a very widely distributed style, having been found in the Tiwanaku heartland ([Bermann 1990]: Figure 77b, [Bermann 1994]: 135, Figure 9.5b; ), in the Juli area ([Stanish and Steadman 1994]), at Camata ([Steadman 1995]: 392, Figure 70d), on the Santiago de Huatta Peninsula ([Lémuz Aguirre 2001]: Figure 8.60d-f), and in the northern Titicaca Basin and as far away as Cuzco ([Chávez 1985]). They are even more rare than Qeya polychrome ceramics. Only one example was recovered during the survey (Figure 7.1u). Flattened or ``shelf'' rims incised with parallel or zig-zag lines appear to be distinctive of this phase as well (see Figure 7.2p; also [Burkholder 1997]: 170 and Figures 6.10, 8.2), as do lobes on the rims of hyperboloid bowls, incised with 2-4 parallel lines ([Lémuz Aguirre 2001]: Figure 8.55a-b).

Since Qeya polychrome and Qeya incised ceramics are so rare, they cannot be used to reliably define LF2 occupations in Taraco Peninsula sites. For this purpose it was necessary to identify a plainware ceramic paste distinctive of the phase. Paste Group 13 serves the same purpose for the LF2 as Paste Group 6 did for the LF1.

Paste Group 13 consists of thin, reduced sherds, brown to dark gray in color, mostly of cooking ollas, which are tempered with a great quantity of mica. The density of mica in these sherds is very high, and may be considered distinctive of the phase. Paste Group 13 did not comprise the majority of the LF2 ceramic assemblage. However, it was sufficiently common to be useful as a temporal marker in surface collections. For example, Janusek found that a ``brown'' paste and a ``dense mica'' temper characterized 14% of the LF2 assemblage at the site of Kirawi on the Pampa Koani. He also notes that at this time ``ollas with thin walls and high densities of fine mica became common'' ([Janusek 2002]). A similar pattern may be observed at Tilata, in the Tiwanaku Valley. Further, Mathews notes that the blackware of the ``Kallamarka style'' ([Portugal Ortíz and Portugal Zamora 1977]: 260) represents ``a probable Tiwanaku III domestic ware'' ([Mathews 1992]: 223). This last assemblage may or may not correspond to what I have termed Paste Group 13. Finally, Carlos Lémuz has stated that nearly 20% of LF2 (his `Pana Tardío' phase) ceramics on the Santiago de Huatta Peninsula can be classified as his Paste 12, which is ``very dark'' and contains significant quantities of mica ([Lémuz Aguirre 2001]: 159).

Summarizing the LF2 ceramic evidence, I am forced to admit that this phase is the most tenuously identifiable. The truly diagnostic ceramics (the Qeya polychrome and incised wares) are very rare indeed. The only apparently diagnostic plainware type (Paste Group 13) is uncommon, comprising - if I may extrapolate from Janusek's excavation data - no more than 10-20% of the total plainware assemblage. I believe that my definition of the phase has been adequate to satisfactorily identify LF2 occupation components using mixed surface assemblages. However, it is abundantly clear that much more work is required and that the LF2 is the most problematic phase in the regional sequence.