Wari

A valley in contact with the coast

formationDrafted from Tung, The Oxford Companion to Archaeology, 2nd ed. (2012)

Long before Huari existed as a city, the Ayacucho basin in the central Peruvian highlands was home to a distinct pottery tradition archaeologists call Huarpa, spanning roughly the first six centuries of the common era. Toward the end of that period, the valley's own people came into sustained contact with the Nasca culture of the south coast - a contact visible in shared iconography and in a new practice that would become one of the most distinctive markers of the state that followed: the taking of trophy heads.

A practice borrowed and then redesigned

Nasca trophy heads, prepared with a hole through the forehead for a carrying cord, had circulated in the region for generations before Huarpa and early Wari sites began producing their own version. The Wari form was different enough to be told apart from its Nasca model on sight: a perforation through the crown of the skull instead of the forehead, designed so the head could be displayed upright and facing forward rather than hung. Only men and children were treated this way, a selectivity that suggests the practice carried specific meaning about who counted as an appropriate trophy rather than being applied indiscriminately to any enemy taken in war.

Metropolitan Museum of Art · Tung, The Oxford Companion to Archaeology, 2nd ed. (2012) · Metropolitan Museum of Art · CC0

A tradition inherited rather than invented

What Huarpa and early Wari communities took from their Nasca neighbours, they did not copy wholesale. The redesign of the trophy head - a small technical change with real symbolic weight - is one of the clearest signs archaeology has of a highland society actively reworking outside influence into something distinctly its own, in the same period that the settlement at Huari itself was beginning the rapid growth that would turn it from a valley town into a capital.

A city with no gaps between its walls

formationDrafted from Tung, The Oxford Companion to Archaeology, 2nd ed. (2012)

Early in the sixth century, Huari grew from a valley settlement into a genuine urban centre, eventually covering several square kilometres and laid out to follow the local terrain rather than any single rigid plan. A secondary site, Conchopata, developed some twelve kilometres to the south, and from this two-centre heartland Wari's own economic and political reach began extending across the central Andes.

Tung, The Oxford Companion to Archaeology, 2nd ed. (2012) · Randal Sheppard · CC BY-SA 2.0

Rooms that let people in but not through

What made Wari architecture instantly recognizable, at Huari and at every site it later built, was its own distinctive orthogonal form: great rectangular enclosures laid out on a rigid grid and subdivided into square or rectangular patio groups, each with a central open patio ringed by long narrow rooms called galleries, sometimes stacked two or three storeys and rooms deep. Doorways connected each patio to its own surrounding galleries freely enough - but doorways between one patio group and its neighbour were kept deliberately scarce. Movement within a single enclosure was easy; movement across the settlement as a whole was not.

An architecture built to be recognized from a distance

That same rigid, compartmentalized layout became archaeology's own most reliable fingerprint for Wari presence anywhere in the Andes - a building style distinctive enough that its appearance at a site, whether a small provincial outpost or Huari itself, is taken as reasonably solid evidence of Wari's own hand in it, even where no other evidence of political control survives. A state that left no written record of its own borders had, without meaning to, left behind walls that still mark them.

An empire or a shared idea?

aside?What aside meansDrafted from Tung, The Oxford Companion to Archaeology, 2nd ed. (2012)

Whether Wari was a genuine centralized empire, exercising direct political and military control over the territory its architecture and artifacts turn up in, or something looser - a shared religious and economic system that spread through trade, ritual authority and imitation rather than conquest - is a question archaeology has not settled, and the evidence pulls in both directions depending on where it is examined.

Metropolitan Museum of Art · Tung, The Oxford Companion to Archaeology, 2nd ed. (2012) · Metropolitan Museum of Art · CC0

Reasons to see an empire

The case for centralized control rests on real coercive infrastructure: military raiding that produced trophy heads from captured men and children, walled provincial centers built to a standardized plan and clearly designed to restrict movement and control access, valleys like Jincamocco's whose own terracing and maize production were reorganized under what the evidence suggests was direct Wari direction, and a road network substantial enough that the Inca simply inherited large parts of it centuries later rather than building their own from nothing.

Reasons to doubt it

Against this, several of Wari's most significant sites resist a simple imperial reading. Construction at Viracochapampa in the north began and was never finished - evidence, perhaps, that the surrounding polity negotiated cooperation rather than being conquered outright, or perhaps that Wari's own reach never actually extended that far in practice. Pachacamac, a major religious center clearly related to Wari in style, cannot be confidently classified as either an independent shrine complex with its own wide influence or as a Wari province proper. And burial practice at the Wari heartland itself, compared directly against Tiwanaku's, shows no shared funerary customs tied to the Staff God imagery the two states otherwise held in common - a sign that even where Wari and its neighbours shared religious symbols, they did not necessarily share the same underlying institutions.

Tung, The Oxford Companion to Archaeology, 2nd ed. (2012) · Unknown photographer · CC0

A debate the evidence keeps reopening

Coastal control is the clearest example of how thin the evidence for outright conquest can be: the collapse of the Moche polity on the north coast happened around the same time Wari influence first appears there, but nothing establishes that Wari caused it rather than simply outlasting or replacing it after an unrelated failure. Wari artifacts found along the coast could as easily reflect trade between independent neighbours as they could tribute extracted from conquered subjects, and nothing yet excavated settles which explanation is correct.

The largest compound the Andes had seen

expansionDrafted from Tung, The Oxford Companion to Archaeology, 2nd ed. (2012)

Wari's own provincial centers were never simply smaller copies of the capital. Where Huari sprawled across several square kilometres and bent its own layout to the local terrain, each provincial center was typically a single, self-contained rectangular enclosure - a scaled-down but still unmistakably Wari design, dropped into a landscape far from Ayacucho and clearly meant to project the same authority there that the capital projected at home.

A wall eight hundred metres on a side

The largest of these provincial centers, Pikillaqta, stood just southeast of what is now Cuzco - a single great rectangular enclosure some eight hundred metres on a side, subdivided internally into dozens of patio groups. Its central portion saw a century or more of real occupation, but other parts of the site were still under active construction when the wider Wari state collapsed, work simply halted and never resumed.

Tung, The Oxford Companion to Archaeology, 2nd ed. (2012) · KimonBerlin · CC BY-SA 2.0

A valley reorganized for maize

Further south, in the province of Jincamocco, Wari's own presence took a different, more agricultural form. Investigations in and around the site show the surrounding valley extensively terraced and, by most readings of the evidence, reorganized under Wari direction specifically to increase maize production - not merely a garrison imposed on existing farmland but an active reshaping of how the land itself was worked, expanded further as Wari's own occupation there continued. Between a fortress-scaled administrative center at Pikillaqta and an agriculturally re-engineered valley at Jincamocco, the provincial network shows two rather different faces of the same expanding state - one built to hold territory, the other built to make it produce more than it otherwise would have.

Neighbours on the same mesa

peakDrafted from Williams, Latin American Antiquity, vol. 12, no. 1 (2001)

In the Moquegua Valley of far southern Peru, Wari built an administrative and ceremonial center at Cerro Baúl, expanding agricultural output in the upper valley with new canal systems and terracing. The location was no accident: Moquegua was the one place Wari's own territory came into direct contact with the other great power of the Middle Horizon Andes, the contemporary state centered at Tiwanaku on the shore of Lake Titicaca, which held its own settlements in the same valley.

Williams, Latin American Antiquity, vol. 12, no. 1 (2001) · Randal Sheppard · CC BY-SA 2.0

A question archaeologists long assumed had an easy answer

For decades, scholars who noticed how much iconography Wari and Tiwanaku shared - both venerated a version of the same Staff God figure

• assumed the simplest explanation: that Tiwanaku, as the older and more established power, had shaped Wari religion and preceded it in time. Excavation at Cerro Baúl itself has complicated that assumption considerably. Radiocarbon dates recovered there give no support to Tiwanaku predating Wari's own presence in the valley; instead, the evidence indicates both states occupied Moquegua simultaneously for roughly the last three centuries of the Middle Horizon.

Neighbours who never came to blows

What excavation at Cerro Baúl has not found is any sign that this shared frontier produced open conflict. Wari built in its own distinctive orthogonal style and Tiwanaku in its own, each recognizably itself even while occupying the same valley and drawing on the same religious imagery, and nothing in the archaeological record points to violence between them. Two of the largest states the pre-Columbian Andes had yet produced shared a border for centuries and appear, on the evidence available, to have simply lived beside one another.

Pots killed with a blow to the chest

peakDrafted from Tung, The Oxford Companion to Archaeology, 2nd ed. (2012)

Wari's own expansion carried a distinctive religious practice with it wherever its influence reached: the ritual offering of large quantities of finely made, brightly decorated ceramics. A pit would be dug, and elaborate polychrome pots - some depicting supernatural beings or human figures - were deliberately smashed and thrown in, the blow aimed first at the head or chest of whatever figure the vessel portrayed, as though the pot itself were the thing being killed rather than merely broken.

Cleveland Museum of Art · Tung, The Oxford Companion to Archaeology, 2nd ed. (2012) · Cleveland Museum of Art · CC0

Offerings that travelled with sacrifice

At Conchopata, these ceramic-killing deposits appear alongside sacrificed camelids and human trophy heads, tying the practice directly to the same violent expansion visible elsewhere in the archaeological record. Similar offering deposits turn up at Nawinpukio near the capital and, far from the Wari heartland, on the south coast at Maymi in the Pisco Valley and Pacheco in the Nasca Valley - concrete evidence that this specific religious practice, and not merely generic Wari-style objects, had spread well beyond the Ayacucho basin.

A destroyed site that still changed a valley

Pacheco's own offering deposit was investigated in the 1920s, before the surrounding site was destroyed by modern agriculture, but even that partial record shows the deposit's impact: Nasca settlement patterns in the area were completely disrupted around the time Pacheco was established, and new burial customs appear alongside it - evidence, however read, of intense interaction between Wari and the existing Nasca population, and quite possibly of Wari authority being established there directly rather than merely traded for.

A collapse no one has dated precisely

endDrafted from Tung, The Oxford Companion to Archaeology, 2nd ed. (2012)

Around 1000 CE, Huari and most of the provincial network built to project its own authority across the central Andes were abandoned. Construction that had been underway at Pikillaqta simply stopped, leaving parts of the site permanently unfinished; the provincial centers that had linked Cajamarca to Cuzco fell out of use along with the capital that had directed them. A state that had grown rapidly from a single valley town into a network spanning much of highland Peru came apart within roughly the same span of time it had taken to build.

A cause still missing from the record

What ended it is, at present, genuinely unknown. Unlike neighbouring Tiwanaku, whose own collapse has recently been dated with real precision through large radiocarbon sampling programs and tied to specific, closely studied social changes in the decades beforehand, no comparably rigorous study of Wari's own decline has yet settled the question of when, exactly, the collapse began or what combination of internal strain, environmental change, or external pressure actually caused it.

A road network too useful to abandon

What did not disappear, even after Huari itself emptied out, was the physical infrastructure Wari had built to hold its own territory together. The road network linking the capital to its provinces survived the state that built it by centuries, absorbed wholesale into the Inca highway system that would eventually connect an empire built, in more than one sense, on foundations Wari had already laid.

Tung, The Oxford Companion to Archaeology, 2nd ed. (2012) · Ordzonhyd Rudyard Tarco Palomino · CC BY-SA 4.0