Phase definition

The Late Chiripa phase is the last of the phases defined by Steadman in the Chiripa sequence ([Steadman 1999]). In my analysis, Late Chiripa occupations were identified using the frequency profile analysis technique, as described in Chapter 3. Essentially, the Late Chiripa phase is characterized by high frequencies of large angular quartz fragments as temper. Over 50% of sherds from this phase have this temper. This paste is quite easily identifiable with the naked eye. Flat-bottomed bowls become common in this phase, represented by both vertical- and slightly flaring-walled examples. Two new forms include vessels with low ring bases and ceramic trumpets, both central to the definition of Yaya Mama religious tradition ([Chávez and Mohr Chávez 1975,Chávez 1988]), and both illustrated in Figure 6.1g and k. Additionally, ceramic figurines seem to appear in this phase, though they are exceedingly rare (only one has been reported from the Taraco Peninsula to date; it is illustrated as Figure 6.1l).

Another oddity that merits mention, if only for its rarity, is the sherd illustrated in Figure 6.1d and e. This is an example of a ``zoned-excised'' sherd. The term is intended to convey the fact that the different color zones are placed on different surface levels. In the illustrated example, on the exterior of the vessel (Figure 6.1e), the ``sunken'' area is slipped cream, and the ``raised'' area is slipped red.

Figure 6.1: Middle Formative ceramics

a-c) painted, d-e) ``zoned-excised'', f) olla, g) ring base, h) applique fillet, i-i) modeled decoration, k) ``trumpet'' with post-fire paint in incisions, l) anthropomorphic figurine with quartz inlay

Decorated ceramics also become common for the first time in the Late Chiripa phase. The most common decorated wares are bowls manufactured in a micaceous paste. The most common color combination is cream on red (e.g. Figure 6.1b), though other colors include black and a reddish-yellow or orange. Zoned incised wares are rare but not unknown. In all these respects, Steadman's Late Chiripa phase is identical to Chávez's Late Chiripa ([Mohr 1966,Chávez 1988]) and Browman's Mamani ([Browman 1978a,Browman 1978b,Browman 1980]) phases. All of these seem to be describing the same general ceramic assemblage.

There is also some support, for the first time, for the hypothesis that decorated ceramic serving wares and public architecture are somehow indicative of competitive feasting and commensal politics. Steadman notes that the frequency of decorated Late Chiripa ceramics on the Chiripa mound is more than three times their frequency in the site considered as a whole, and as much as twenty times their frequency in certain domestic middens ([Steadman 1999]: 66).

There are many indications that the Taraco Peninsula villages participated in a Titicaca Basin-wide exchange system in the Middle Formative. This topic will be discussed more thoroughly later in this chapter. For now, however, we may note the - admittedly sparse - presence of ceramics relating to the northern Titicaca Basin Qaluyu tradition. Some examples are illustrated in Figure 6.2. Most of these are classifiable as Qaluyu Wide-line Incised ([Steadman 1995], see also [Stanish and Steadman 1994]: Figure 97). This style dates to the Late Qaluyu 1 and 2 phases at Camata ([Steadman 1995]: Table 35), or approximately 850-400 B.C. ([Steadman 1995]: Table 3). These sherds (Figure 6.2a-e, g) are decidedly non-local, manufactured as they are in an exotic mineral-tempered paste. The one painted example bears a common Qaluyu Polychrome motif (see [Steadman 1995]: Figure 45 for an example). This style dates to the same time period as Qaluyu Wide-line Incised, which is to say the Late Qaluyu 1 and 2 phases ([Steadman 1995]: Table 34). This sherd (Figure 6.2f) may also be imported. Though its paste is fiber-tempered, the red paint applied over a cream background sparkles and probably contains powdered hematite. This is definitely a northern basin trait. However, the particular colors used (red as opposed to brown on cream) are more southern than northern, and may identify the piece as a local imitation of a Qaluyu Polychrome design.

Figure 6.2: Qaluyu-related Middle Formative ceramics from the Taraco Peninsula

e-e) and g) Qaluyu Wide-line Incised, f) Qaluyu Polychrome-related

See Figure 6.1 for color coding.

The chronological placement of the Late Chiripa phase remains somewhat problematic. The ``official'' dates of the Taraco Archaeological Project are 800-100 B.C. ([Whitehead 1999]: 20). Fixing the beginning of the Late Chiripa phase is problematic due to an extended plateau in the radiocarbon calibration curve ([Whitehead 1999]: 21), which gives dates in the range of 800-400 B.C. or so extraordinarily large error ranges. Nothing at all can be done about this, but it does seem that 800 B.C., give or take a bit, is a legitimate point of demarcation between the Middle and Late Chiripa phases.

The end of the Late Chiripa phase is another matter altogether. 100 B.C. is a conventional date and has been used by many commentators on Chiripa, including Browman ([Browman 1978a]: 809) and Chávez ([Chávez 1988]: 18). This conventional date is primarily based on dates from Kidder's excavations at Chiripa, and intends to date the time of the burning of the Upper House Level. All investigators seem to agree that this event marks the end of the Late Chiripa phase, since the superposed mound structure seems to be Late Formative in date ([Browman 1978a]).

One of the principal contributions of the Taraco Archaeological Project's excavations at Chiripa has been the improved dating of the Chiripa phases. Over 40 AMS dates on annual plant remains have been run ([Whitehead 1999]). Of the more than 15 of these dates that pertain to the Late Chiripa phase, only one has a mean calibrated age more recent than 250 B.C. This one more recent outlier is clearly an aberration, since it is form a Lower House structure (Locus 1386) and is bracketed by a whole series of dates older than 350 B.C. Additionally, we have three dates on floor materials from House 5, firmly dating the burning and abandonment of that structure to approximately 270 B.C. ([Bandy 1998]), plus or minus a few decades. It should also be noted that there are very few dates from other projects which would be later than this. The two Kidder dates that fall after this time (P-117 and P-118; see [Ralph 1959]), are clearly erroneous, since they are identical or prior to the very same event - the burning of House 5 - as are the three TAP dates mentioned in my discussion of the Chiripa architectural sequence (section 6.3).

All of the above data indicate that there is no evidence placing the end of the Late Chiripa phase later than 250 B.C. In the present analysis, therefore, I place Late Chiripa at 800-250 B.C. It should be noted, though, that no good stratigraphic column has yet been published from the southern Titicaca Basin that spans the Late Chiripa/Kalasasaya transition. This is clearly a priority for future research in the region.