Discussion

I argue in Chapter 6 that in the Middle Formative access to exotic goods and extra-local trading partners was incorporated into a local system of status competition on the Taraco Peninsula. The peninsula's position on a probable trade route from the western Titicaca Basin to the lowland valleys or yungas (see Figure 6.8) was used by local leaders and groups to obtain exotic objects which were then strategically redistributed in public ceremonial or commensal contexts. Around 250 B.C., at the beginning of the LF1, the lake rose again to nearly modern levels. This would have interrupted the existing trade route, which would then have shifted to the south through Desaguadero and the Tiwanaku Valley (see Figure 7.5).

While this relocation of the late MF trade route would certainly have had an impact on the Taraco Peninsula communities, there is some evidence which suggests a more general and profound disruption of the regional exchange system at this time. Though firm evidence is lacking, it would seem that the olivine basalt hoes - the principal archaeological marker of the late MF exchange system - does not merely become less common in the LF1, but appears to vanish entirely. The dramatic increase in the frequency of quartzite and slate hoes noted by Janusek represents nothing other than an increased use of local lithic materials to replace the preferred but unobtainable basalt.

Of course, Incatunahuiri - the possible source of this basalt (see Chapter 6) - was not abandoned at this time, nor did hoe production cease there. This is clearly shown by the fact that at the nearby site of Camata, hoes of this material actually become most common (almost 90% of all lithic material) in the Pucara 2 phase, around 100 B.C. - 100 A.D. ([Steadman 1995]: 32). What happened to Incatunahuiri is that it fell under the influence of the Titicaca Basin's first urban center: the site of Pukara, located some 60 km northwest of Lake Titicaca. Incatunahuiri in the Late Formative Period was closely affiliated with Pukara. The hill on which it is located was remodeled, an imposing sunken court was placed at its summit, and a number of large stone sculptures in the Pukara style were either produced at the site or imported from elsewhere. The nature of the Pukara polity is poorly understood, though research is under way to address the many questions with which it is surrounded ([Cohen 2001,Klarich and Craig 2001]). What does seem clear, though, is that Pukara was deeply involved in exchange networks connecting the Titicaca Basin to the Arequipa area (the principal source of obsidian; see [Burger et al. 1998,Burger et al. 2000]), the southern Peruvian ([Bandy 1995,Feldman 1989,Feldman 1990,Goldstein 2000]) and northern Chilean (cf. [Conklin 1983]) coasts, and the eastern valleys ([Plourde and Stanish 2001]; more generally see [Mujica 1978,Mujica et al. 1983,Mujica 1985]). It is possible that Pukara exercised a monopoly on the obsidian trade, at least for a period of its history, and it may have possessed or aspired to similar control of other commodities. Thus it may be that Pukara actually discouraged exchange relations between its trading partners and third parties. Put plainly, it is possible that in the LF1 period exchange between the southern and western areas of the Titicaca Basin was actively suppressed by the Pukara polity. Such a situation could potentially explain the sudden disappearance of olivine basalt from the southern basin at this time. However, this scenario must remain hypothetical pending further research.

This disruption of the late MF trade route and possibly of contact with the western Titicaca Basin would have created a crisis for leaders whose power strategies had come to depend on easy access to imported goods and non-local social contacts. If we assume that the leaders of the various communities had all adopted broadly similar competitive strategies in the late MF, then we can hypothesize a generalized crisis of political legitimacy on the Taraco Peninsula at this time. What would such a crisis look like archaeologically? First of all, we might expect artifacts associated with the older strategies to disappear. This is precisely what happens with the Late Chiripa ritual feasting artifact complex. Second, we would expect local leaders to search for other more appropriate strategies, and therefore the appearance of new artifacts associated with these new institutions. This is exactly what happens when the Kalasasaya ceramics appear in the southern basin; a new set of artifacts, associated ideas and activities almost certainly resulting from imitation of Pukara ceramics and sculpture from the northern Basin.

In addition, we would expect some leaders to emerge from the crisis more successfully than others. This would mean that each community might have fewer leaders than before, and that some sites might rise to local prominence, subjugating or extracting tribute from their neighbors. This also seems to have taken place. Sometime early in the LF1, the site of Kala Uyuni began to grow very rapidly, drawing population from the remainder of the sites on the peninsula. Simultaneously, a movement of population took place away from the large villages and into smaller farmsteads and hamlets. I interpret these facts, and the resulting emergence of a three-tier site size hierarchy, as indicative of the emergence of a multi-community polity centered at Kala Uyuni. This polity was quite small-scale; it's eastern border seems to have fallen between the sites of Chiripa (T-1) and Chiripa Pata (T-4). In other words, there were many surrounding villages which were not tributary to Kala Uyuni.

This political development was accompanied by a wholesale transformation of the entire ceremonial complex which had characterized MF ceremonialism on the Taraco Peninsula and in adjacent areas. The characteristic Chiripa decorated ceramics, in continuous use for over 600 years, were abruptly abandoned, replaced by the Kalasasaya complex. The latter includes significant imitation of the northern basin Pukara style. At the same time, the sunken court complex was reinvented, incorporating for the first time the large upright pillars that come to be characteristic of later Tiwanaku architecture. The association of stone tenon heads with sunken courts also begins in this phase. The overall impression is of a complete reinvention of ceremonial practices and political strategies. I propose that this was in response to the political crisis resulting from the lake level rise and concomitant shift in the trade route.

It was apparently also at this time (early in the LF1) that the site of Tiwanaku was first substantially occupied. The Tiwanaku Valley had been home to a low-density population from the Early Formative period ([Albarracín-Jordan 1992,Albarracín-Jordan 1996a,Albarracín-Jordan et al. 1993,Mathews 1992]). However, in the LF1 the site of Tiwanaku was founded on the valley floor and grew very rapidly to cover - by the end of the LF1 - something like 20 ha. This was comparable in size to the center of the Taraco Peninsula polity. We may suppose that it was the center of a contemporaneous and competing polity, employing the same basic ceramic, architectural and sculptural styles.

It is also informative to consider what happened in the Lower Tiwanaku Valley at this time. This area lies exactly between Tiwanaku and Kala Uyuni and had been home to a small but significant population in the MF. However it was almost completely abandoned in the LFI ([Albarracín-Jordan 1992,Albarracín-Jordan and Mathews 1990,McAndrews et al. 1997]). This seems to be a clear case of the formation of an unoccupied buffer zone between competing polities, as has been documented in many parts of the world.

The founding and rapid success of Tiwanaku and the Tiwanaku polity probably had everything to do with the lake level rise and the shift to the south of the late MF trade route. As Figure 7.5 makes clear, Tiwanaku was located precisely on the new trade route. Though the trade passing along this route may have been much reduced relative to its MF level, as I discussed above, whatever exchange was taking place between the southwestern Titicaca Basin and the yungas in the early LF1 passed through Tiwanaku. The misfortune of the Taraco Peninsula communities was a positive boon to the inhabitants of the Upper Tiwanaku Valley, and a crucial element in the founding and early growth of the Tiwanaku polity. In fact, the site of Tiwanaku was quite possibly founded in a bid to control the new trade route.

Around 100 A.D. the level of Lago Wiñaymarka dropped suddenly, and the majority of the little lake again became dry (the shoreline was roughly the same as that indicated in Figure6.8). Once again, the shortest route between the western Titicaca Basin and the yungas passed through the Taraco Peninsula, as it had during the late MF. Though the magnitude of this trade was possibly reduced by the influence of the Pukara polity, a strategic position on the trade route was probably still desirable to local elites.

Such a consideration was probably the reason that the center of the Taraco Peninsula polity was relocated sometime during the LF1 from Kala Uyuni to the Santa Rosa group. This group of sites, including Sonaji (T-271), Kumi Kipa (T-272) and Kollin Pata (T-322), is located precisely along the low lake stand trade route (Figures 6.8 and 7.3). The Santa Rosa group eventually grew to cover 21 ha by the end of the LF1, roughly the same size as Tiwanaku.

It may be that the Taraco Peninsula experienced a period of regional preeminence with its capital at Santa Rosa. If so it was short-lived. Around 300 A.D., at the beginning of the LF2, the level of the lake rose again, this time to modern levels. At this time, the site of Tiwanaku began to grow very rapidly, averaging around 1% annually for the LF2. The reasons for the process of urbanization are beyond the scope of the present study. We may note, however, several factors which fall within the terrain covered in this chapter. First, due to the rise in lake level, the trade route between the western basin and the yungas again passed through Tiwanaku. Second, the collapse of the northern basin Pukara polity, which took place around this time, would have removed any barriers to exchange with the western Titicaca Basin which may have existed in the LF1. Thus, activity on the trade route which now passed through Tiwanaku was probably greatly intensified relative to LF1 levels.

Whatever the reasons, Tiwanaku began to urbanize rapidly in the LF2. The Taraco Peninsula was one of the regions from which population flowed into the new city. In fact, it was the principal region, with as much as 46% of the population influx into Tiwanaku in the LF2 coming from the Taraco Peninsula communities. This significant loss of population to Tiwanaku suggests that at some point in the LF2 the Taraco Peninsula was incorporated into the Tiwanaku polity.

What do we mean when we say ``incorporated into the Tiwanaku polity?'' The immediate image that is evoked is one of military and political domination. That is, that Tiwanaku dominated the Taraco Peninsula militarily and extracted tribute from it. The flow of population from the Taraco Peninsula would then be caused by either forced relocation at the behest of the conquering polity or a natural response to a shift in the availability of tribute redistribution, or some combination of the two. This is an easy scenario to envision.

However, the military conquest scenario is incapable of explaining the population relocation I have been discussing. This is so because at the end of the LF1, before the population relocation began, the population of the Tiwanaku polity was probably less than $\frac{1}{3}$ that of the Taraco Peninsula polity. It is difficult to imagine a polity of 1500 persons overcoming one with a population of 5500 by force of arms, especially since we have no evidence of any change in military technology at this time. In other words, Tiwanaku military superiority only became possible after the population relocation which it was supposed to have explained.

All this is to say that military superiority was an effect rather than a cause of the population movement in question. The actual cause of this population movement cannot at present be established. Any hypothesis, however, must now take into account the fact that people were moving to Tiwanaku in large numbers from surrounding areas in the absence of military coercion. My favored explanation is that the leaders of the Tiwanaku polity invented a new type of political economy in the LF2. This new political economy used raised field agriculture and a polycyclic strategy of staggered production cycles ([Bandy 1999c]) to allow the extraction of a greater amount of surplus labor per person without interfering with the dryland subsistence activity of the commoner population ([Bandy 2000]). This increased surplus extraction could then have been used to underwrite an intensification of public ceremonialism and large scale feasting at the site of Tiwanaku itself, providing further inducement to the inhabitants of adjacent areas to change their residence. This process could have produced accelerating rates of immigration throughout the LF2, and the model can account satisfactorily for the urbanization of Tiwanaku.

This scenario is a hypothesis which is in need of testing. Whatever its validity, it is clear that by the end of the LF2, around 500 A.D., Tiwanaku was a city, and had become capable of dominating the entire Titicaca Basin politically, economically and militarily. At some point during the LF2 the Taraco Peninsula ceased to be an autonomous polity and became a heartland province of the Tiwanaku state.