Late Formative 1

Janusek dates the LF1 from 200 B.C. to 300 A.D. ([Janusek 2002]). I have used these dates, while moving the initial date to 250 B.C. to accord with my hypothesized terminal date for the Late Chiripa phase (see Section 6.1).

I consider three groups of ceramics to be diagnostic of the LF1 phase. These are:

1. Kalasasaya style zoned-incised ceramics (Figure 7.1a-l),
2. Kalasasaya style hemispherical bowls with red-painted rims (Figure 7.2r-u), and
3. a distinctive fiber-tempered paste, termed Paste Group 6 in my analysis.

The Kalasasaya zoned-incised group is the most familiar, and the rarest, of the three. It was first identified by Bennett at Tiwanaku ([Bennett 1934]). The best-known examples of the style were excavated by Ponce from several tombs underlying the Kalasasaya structure at Tiwanaku ([Ponce Sanginés 1976]); thus the name. Examples have been found at other sites in the Tiwanaku heartland area by Mathews ([Mathews 1992]), Bermann ([Bermann 1990]), Janusek ([Janusek 2002]), and myself (Figure 7.1a-l). These vessels present a high burnish on the exterior, and the incisions delimit color zones. Representations of geometric motifs - most commonly the ``step-fret'' - predominate, and zoomorphic and anthropomorphic motifs exist as well. Designs are usually painted in black. The rest of the vessel is commonly unslipped, but red slips are relatively common, and other slip colors are known.

Figure 7.1: Late Formative ceramics

a-l) Kalasasaya Zoned-Incised (LF1), m-q) Qeya polychrome on unslipped buff (LF2), r-s) black/red (LF2), t) yellow/black (LF2), u) Qeya incised (LF2), v-w) ceramic ``buttons'', one cane-stamped and painted (LF1 and LF2), x-aa) incised ``trumpet'' fragments (LF1), ab-ac) annular bowls (LF1 and LF2).

Kalasasaya zoned-incised ceramics are very rare. The examples illustrated in Figure 7.1 represent the majority of the specimens recovered in the survey, out of a sample, it will be recalled, of almost 100,000 sherds. They are also very spatially restricted. Examples were recovered from only four sites, of the 54 sites with documented LF1 occupations. Most of these are from T-232 and T-272, the two largest sites during this phase. Another two are from T-141, which seems to have included a relatively high-status LF1 cemetery. This observation agrees with that of Janusek, who found that at the Katari Basin site of Kirawi, 90% or more of elaborate decorated LF1 sherds were located in a single outdoor midden ([Janusek 2002]). Thus these materials were concentrated at a regional scale - they are found almost exclusively in the largest and most important sites - and on an intra-site scale as well - they are located in very restricted contexts within individual sites. This is an important observation, and a major departure from Middle Formative precedent.

The second group of diagnostic LF1 ceramics are small hemispherical bowls with red-painted rims (Figure 7.2r-u). These are commonly referred to as ``Kalasasaya bowls.'' Bermann calls these ``chestnut bowls'' (see [Bermann 1994]: Figure 5.4 for a photograph). The examples I encountered on the Taraco Peninsula have a very fine, compact and light-colored paste, often with small red mineral inclusions. In this they resemble the bowls from Tiwanaku more than they do those from Lukurmata, which seem to be more brown. They are often highly eroded, as the paste is usually very soft, with an almost soapy texture. Rims are usually rounded and direct, though one flared-rim specimen was recovered (7.2r). The bowls are typically painted on both the exterior and interior rim, though significant variation exists in this regard. The paint employed is normally a very dark red, almost purple (7.5 YR 3-4/4-6).

Kalasasaya bowls are much more common than are the zoned-incised sherds, and they were found on virtually every site with a LF1 occupation. There is some evidence that they were used in ritual or public consumption contexts. At Lukurmata, Bermann found that in the second occupation Kalasasaya bowls were predominantly associated with a single, elaborate outdoor hearth, to the exclusion of other more quotidian hearths nearby ([Bermann 1994]: 72). Janusek makes a similar observation with respect to his excavations at Kirawi ([Janusek 2002]). Bermann suggests the bowls were employed in ``offering/prestation activities'' ([Bermann 1994]: 76).

Figure 7.2: Late Formative ceramics

a-f) appliqué fillet (LF1), g-h) appliqué fillet, cane-stamped (LF1), i) appliqué fillet, obliquely incised, j) appliqué fillet, punctate, k) grooved rim bowl (LF1), l-n) thickened rim, punctate (LF1), o) punctate rim (LF2), p) incised rim (LF2), q) horizontal handle on rim (LF1), r-u) red-rimmed buff bowls (LF1).

See Figure 7.1 for color coding.

The third group of diagnostic LF1 ceramics are those manufactured of my Paste Group 6. This is a semi-compact paste with a very finely-chopped fiber temper. This is easily distinguished from Middle Formative fiber temper, which is invariably much more dense and coarse. Sherds of this paste are normally oxidized - though they commonly have a slightly reduced core - and are a red-brown color. Small gold mica fragments are also present in quantity. Most distinctive, however, is the use of significant amounts of biotite in the temper. The tiny black flecks in the red-brown matrix impart a very distinctive appearance. Sherds of this paste are typically extremely eroded when encountered on the surface, and are broken into small fragments.

This paste is extremely common in the LF1. All sites with Kalasasaya sherds also have great quantities of Paste Group 6 on the surface. No site was located which simultaneously had Kalasasaya ceramics and lacked Paste Group 6 sherds. I have used it as a diagnostic for this reason. It is easily distinguishable from Middle Formative ceramics. While fiber-tempered ceramics continued to be used in small quantities in the subsequent LF2, and even into the Tiwanaku Period, they were much rarer in these later phases, and seem to have been restricted to large storage vessels and annular bowls. In the LF1, by contrast, Paste Group 6 was employed for ollas and jars, as well as for annular bowls. Further, the distinctive biotite temper is lacking in the later phases.

If present at Lukurmata, Paste Group 6 has probably been included in Bermann's ``Lorokea Fiber'' type ([Bermann 1990]: 78). Paste Group 6 is much more specific than Lorokea Fiber, however, since this type seems to simply aggregate all fiber-tempered sherds recovered in Bermann's excavations. Lorokea Fiber is a generic type and is therefore of limited utility.

Finally, Carlos Lémuz's Paste 18 is I think very similar to my own Paste Group 6. Paste 18 is, according to Lémuz, ``the paste most representative of the Early Pana phase'' ([Lémuz Aguirre 2001]: 162), his local LF1 phase designation for the Santiago de Huatta Peninsula. It apparently occurs only in Early Pana contexts, disappearing in the LF2 and Tiwanaku period. Paste 18 contains fiber temper, mica and, significantly, biotite, and is fired in an oxidizing atmosphere. Paste 18 makes up less than 10% of the ceramic sample from Lémuz's excavations ([Lémuz Aguirre 2001]: 162); however I believe Paste Group 6 to be much more common than this on the Taraco Peninsula.

Several other ceramic attributes may be diagnostic of the LF1 phase, though they were not employed as such in the present study. These include triangular (``pointed'') appliqué fillets on the exterior of vessels (Figure 7.2a-h). These fillets are frequently decorated with oblique incision, cane-stamping (Figure 7.2g-h;l see also [Lémuz Aguirre 2001]: Figures 8.53a-b, 8.54a, d) or punctation. Many vessels with these fillets - apparently predominantly large bowls - are also painted on the rim with the same dark red paint which is used on the Kalasasaya bowls (Figure 7.2b-e).

Also common is a thickened vertical rim with punctation (Figure 7.2l, n), horizontal handles on the rim of large bowls (Figure 7.2q), and latitudinally-grooved bowl rims (Figure 7.2k). These last three traits seem to also be present in the LF2, however, and are therefore probably not diagnostic for the LF1.

Finally, it should be noted that the manufacture and use of ceramic ``trumpets'' continued into the LF1 (7.1x-aa; also [Bermann 1994], Figure 5.5c). These may easily be distinguished from the earlier MF trumpets by their lack of fiber temper and their very compact paste. These do not seem to extend into the following LF2 period (witness their complete absence in the LF2 levels at Iwawe [[Burkholder 1997]]), but this is not sufficiently certain at present to allow them to confidently be employed as LF1 diagnostics.