The Rise of Native Lordships at Pashash, a.d. 200–600, North Highlands of Ancash, Peru

ABSTRACT This article examines the rise of native, segmentary lordships in the highlands of north-central Peru. It reports on new excavations and mapping at the seat of a prehispanic polity, Pashash (Recuay culture), a large hilltop center that developed after the collapse of Chavín civilization. Fieldwork revealed monumental constructions and two special activity contexts radiocarbon-dated to ca. a.d. 200–400: an offering area in a large palatial compound and a room-complex with chambers closed off and sealed with feasting refuse. Multiple lines of evidence help reconstruct a regional picture for the establishment of wealthy local elites. Cultural innovations explicitly link new leaders to roles in defense and warfare, economic production, and early burial cult within a high-status compound. The current data underscore a major break from earlier systems of authority and elite material culture, comprising an organizational pattern that was a precursor to the ethnic polities that predominated in later Andean prehistory.

An enduring topic in Andean archaeology centers on early complex societies and the process of social differentiation that gave rise to nobles and political leaders (Burger 1992;Moseley 1975;Shimada 1999;Silverman and Isbell 2008;Stanish 2001). This contribution focuses on the development of segmentary polities headed by figures known as curacas or señores.
Broadly, we understand these as chiefly leaders or lords who drew authority from their role as heads of their respective social segments, typically extended kin groups, what in later prehistory and colonial times were styled as "ayllu" (referring to a corporate group beyond the nuclear family, whose members reckon descent from a common founding ancestor and have ritual as well as living and work obligations together). Historical accounts often called the heads of such groups curacas, principales, and señores, among other terms. Segments were differentiated through kinship ties and moiety divisions and internally ranked against the others. The social arrangements also built in potential for internal conflict, cooperation, and confederation. Despite their autonomy, their modular quality also meant that they could be incorporated, and nested, into existing and other political structures at different scales (Espinoza Soriano 1978;Netherly 1977;Rostworowski and Remy 1992;Zuloaga 2012).
Relevant regional political agglomerations were known under various terms, often in reference to the lead individual (e.g., curacazgo, señorio, cacicazgo); these and similar arrangements (e.g., étnia, huaranga, parcialidad) had local distinctions and finer etymological and temporal distinctions linked to Inca and later Spanish era reorganization. A gender-neutral term, such as "liderazgo," may be useful to describe such forms of leadership in the past, since women may have occupied ruling positions, as well.
Leadership was particularly marked in the domains of collective work, warfare, and festive ritual. Far from mere aggrandizing and wealth extraction, the activities and distinction of leaders concerned group well-being and representation in relation to the other segments. Being identified as living embodiments of a privileged, ancestral bloodline, leaders also maintained key roles in the corporate rituals that legitimated the group's social memory and power. Leadership arrangements of this kind comprised the traditional, if not the predominant, indigenous form of social organization across the Andean region before and during the succeeding Inca and Spanish colonies (Earle 2002;Patterson 1987;Ramírez 2005;Salomon 1986), not least for the highlands under consideration (Chocano 2016;Cook 1977;Mariscotti 1972;Topic and Topic 2020;Zuidema 1973). It remains unclear, however, when and how the emergence of such arrangements took hold and flourished in the Andean past.
This article details one trajectory for this process in Peru's North Central highlands (Figure 1). We report on 2019 fieldwork at the Pashash archaeological site and its main occupation between A.D. 200 and 700 (ages based on calibrated ranges). New survey and mapping detail the large hilltop ceremonial center and its defensive fortifications and monumental enclosures. Excavations revealed two high-status activity contexts: 1) an offering area within a sealed chamber near the corner of a large palatial structure and 2) a room complex with corridors and chambers intentionally closed off and sealed with feasting refuse. Radiocarbon and stylistic evidence date the primary occupations and assemblages to the early centuries A.D. Overall, the research illuminates a northern Peruvian context and the diachronic process of innovations by which secular elites came to power. The new patterns explicitly link emergent leaders to roles in defense and warfare, economic production (as evidenced by camelids, feasting, and labor accumulation), and burial and offering contexts in a palatial context. The record underscores a major break from priest-oriented systems of authority of the preceding period and comprises an early instance of the kind of ethnic polities that predominated in later Andean prehistory.

Regional Culture History and Andean Complexity
From 1000-400 B.C., many of the salient developments in northern Peru were associated with the Chavín culture of the Early Horizon (Menzel and Rowe 1967). The eponymous temple complex of Chavín de Huántar promulgated an extensive regional cult based on veneration of cult images, pilgrimage, and interregional exchange (Burger 2019). A handful of coeval temple settlements are known in central and northern Peru which shared its cosmology and material style. These were peer centers, perhaps branch oracles for the Chavín cult, whose elites actively drew valuable resources, both physical and ideological, from the wider religious and economic network. The far-flung system espoused a lowland shamanistic worldview, and its practices extolled hallucinogen-induced rituals, public congregation, and display of exotics and offerings (Burger 1992  At some centers (e.g., Pacopampa and Kuntur Wasi), high status burials with luxury items have been found directly in the temple platforms (Onuki 2017;Seki et al. 2019). These comprise some of the best early evidence for social differentiation in northern Peru. That elites were buried in the temples is crucial: the personal adornments, grave items, and imagery suggest that their power and status were intimately tied to their connection to the cult. They may have been priests or officiants. As we will see, the explicit link to priestly roles and esoteric knowledge/ritual is far less marked in later Recuay times.
In and around its heartland in highland Ancash, Chavín also began to exert influence on smaller villages and hamlets in neighboring drainages and hilltop communities located to exploit lands for herding and cultivation. The current record indicates widespread but fairly limited quantities of Chavínrelated articles such as pottery, obsidian, and bone items with distinctive cultic imagery (Bria 2017;Burger 1982Burger , 1985Nesbitt, Ibarra, and Tokanai 2020;Ponte 2001;Tello 1960). Broadly, this suggests that local agro-pastoral groups remained largely autonomous but nevertheless engaged in networks of trade and religious interaction with Chavín. Villages nearer to the centers probably also contributed labor and resources to Chavín's monumental building programs and ceremonial activities (Burger 1982;Nesbitt, Ibarra, and Tokanai 2020;Rick 2008).
As a pilgrimage place and engine for long-distance interaction, Chavín began to decline in the centuries following the final building programs during the 5th century B.C. (Burger 2019;Rick 2013). The ensuing time was characterized by major cultural transformations and reorganization at Chavín de Huántar (Lumbreras 2007) and in many communities across its heartland (Bria 2017;Burger and Nesbitt in press;Gero 1991;Terada 1979).
The initial post-Chavín developments are usually associated with a white-on-red pottery and archaeological complex known as "Huarás." Precise dating remains uncertain, but conservatively, it ran from the temple's abandonment (ca. 400 B.C.) to the early centuries A.D. Stratigraphic mixing occurs because Huarás groups directly built on and reused earlier buildings, including at Chavín de Huántar and in its hinterlands (Bria 2017;Burger 1985;Kato 1979;Lau 2004;Lumbreras 2007). Certain aspects of Huarás pottery (e.g., shapes, modeling, painted designs, and use contexts) indicate that its later manifestations also comprised an early component of the Recuay cultural tradition (Gero 1991;Lau 2004Lau , 2011Lau , 2016. The current record, including our recent work at Pashash (below), indicates that white-on-red cultural groups do not demonstrate much social differentiation or variability in indicators of wealth.
By around A.D. 200, this picture began to change quickly. By no means was it a homogeneous process, but some communities in highland Ancash, such as Pashash, began to show greater wealth accumulation, as reflected in much fancier items and major efforts in monumental stone architecture, stone sculpture, and metalwork. These media and associated funerary practices are the primary diagnostics of the Recuay culture and its socio-political reconstruction as a "commonwealth" (Lau 2011, 16-17).
The most widespread, tangible change was the manufacture and use of fine kaolinite ceramics. The light colored clay, coupled with other diagnostics (modeling, resist painting, and shapes), mark the beginning of the Recuay pottery style. Kaolin ceramics are almost always very finely elaborated, and their manufacture required new raw materials, techniques, and knowledge (Grávalos 2021;Reichert 1977); they also began to appear frequently as grave goods and offerings with other items (e.g., metalwork, lapidaries, and textiles), suggesting that people saw them as forms of wealth and objects strongly linked to high status identity.

Fieldwork at Pashash
In 2019, the PIARP (Proyecto de Investigación Arqueológica Región Pallasca) began a multi-year project focused on the rise of segmentary polities in north-central Peru (see Figure  1). This case study is explicitly designed to compare Pashash's developments with another regional center, Cerro San Isidro, based in the Río Nepeña's midvalley Moro area (Chicoine andNavarro 2020, 2021). Because of the overlap in time (ca. 500 B.C.-A.D. 500), the joint project aims to document variability in coeval trajectories of sociopolitical development in adjacent areas of the highlands and coast. More comprehensive comparison is planned after additional fieldwork and analysis.
Our general proposal is that during the early centuries A.D., segmentary lordships became an important political form that proliferated during the socio-political vacuum following the demise of Chavín and related ceremonial centers of the Early Horizon. Prior work points to four key variables, and the case studies seek to examine the impact of each variable for the emergence of native polities. The first is the intensification of economic production (Billman 2002;Earle 2002;Quilter and Castillo 2010;Schaedel 1985), including long-distance trade (Goldstein 2000); the formation of nucleated and heavily differentiated centers (Brennan 1982;Kosok 1965;Makowski 2008;Millaire 2010;Schaedel 1966;Shimada 1994); warfare, competition (Arkush and Tung 2013;Chamussy 2009;Ikehara 2015;Wilson 1988), and ideologies of dominance (Bawden 1996;Bourget 2016;Donnan 2004;Swenson 2007); and, a fourth pattern, the ancestralization of leaders (Isbell 1997;Lau 2010bLau , 2013Zuidema 1973). Our framework recognizes these as pathways and contingent processes for widespread socio-organizational transformations.
Probably the least recognized and evidenced is the fourth process. Essentially, we postulate that a critical factor in the development of polities in northern Peru centered on the newfound status of chiefly leaders as potent ancestral beings (Helms 1998;Hill and Hageman 2016;Houston and Stuart 1996;McAnany 1995). Historically, group recognition of such leaders was often pegged to special high-status goods and ideologically potent imagery and materials. Funerary ritual was also key, for their veneration engaged their being as living instantiations of the privileged bloodline; special housing for and limited access to cult objects (including the ancestral embodiments themselves) ensured the continuity of legitimacy (Isbell 1997;Shimada and Fitzsimmons 2015) and distinctive lifestyles (Couture 2004). If so, we propose that one key diagnostic of such arrangements should be the appearance of monumental residences with exclusive burial cults directly linked to and operating with the activities of the house, not unlike the Inca "panaca" or Chimu "ciudadela" palatial compounds of later cultures.
Pashash (Pallasca province, Ancash) ( Figure 2) has already seen important study. Travel explorations in the late 19th century A.D. reported the site's major monumental constructions (Raimondi 1873;Wiener 1880). It was a convenient quarry for ashlars and other stone to build colonial and Republican structures in the nearby town of Cabana. Also known from the ruins were carved monoliths, some of the finest examples of ancient Andean stonecarving (Schaedel 1952). These included carved slabs and lintels/ friezes that almost certainly adorned the edifices at Pashash. Like Chavín, its walls featured tenon-heads, blocks with projections cantilevered into walls; however, these usually took the form of human heads, often in warrior headgear, not snarling mythical, part-zoomorphic typical of Chavín.
Schaedel contended that the Pallasca region was a major area for Recuay culture and that Pashash comprised the regional center of a large northern señorio (or "chiefdom") (Schaedel 1952;1985, 443, 446;also Smith Jr. 1978). From 1969-1973, a team led by Terence Grieder and Alberto Bueno conducted excavations in two areas of the site, the La Capilla main mound and El Caserón structure (see below). A very rich burial and offerings featuring hundreds of intact grave items (including fine ceramics, metalwork, lapidaries, and stone bowls) were found on La Capilla, helping substantiate the claims for a post-Chavín consolidation of power and an "elite class" (Grieder 1978, 181-182).
More recently, mapping and local survey indicate that Pashash formed part of a basin-wide settlement system over time, one that was much larger than previously known. Limited reconnaissance in 2013-2014 (Lau) and 2019 (Bongers, PIARP) revealed the presence of over 35 sites in the Cabana basin (see Figure 2). The basin is advantageous for its relatively wide, gently-sloping, and wellwatered lands (above 2500 masl) suitable for maize, quinoa, and tuber cultivation. A series of high-altitude lakes (above 4200 masl) to the east and south also nourish steppic grasslands and bofedal wetlands for grazing herds. Pashash's location thus bridges two major production zones; its position also features in local oral traditions as a place that mediates the herder groups of the high, frost-prone lands above and the farmers of the warmer valley zone (Cuba Manrique 2018). Additional survey is planned, but the evidence to date indicates that settlement placement in the coverage region almost always preferred high rocky areas, such as defensible hilltops and bluffs, with good vantage and access to productive lands and water.
Settlement in the region began at least by Late Preceramic times (Grieder et al. 1988). Recent surface collections indicate that intensive settlement in the Cabana basin began during the Early Horizon (1000-400 B.C.). The most common sites were of Recuay affiliation (A.D. 1-700). Late Intermediate Period (A.D. 1000-1450) components were also prominent, sometimes reusing older settlements. In general, there was heavy commitment to large defensible walled constructions, areas for water control/access, and intensive settlement and economic production in higher herding areas.
Pashash (Figures 3, 4) occupies a major crest of a steep ridgeline descending from the mountain, Cerro Mashgonga (ca. 4165 masl). Pashash's protected, delimited zone extends over 28 ha. The East Ridge features abundant surface architecture, both ancient and some modern, as well as Recuay style ceramics. A modern canal runs along the ridge's eastern flank and probably uses an ancient course. The ridge is heavily overgrown today but will be researched in future field seasons.
Very tall revetment walls protect either side of the narrow saddle connecting the East Ridge to La Capilla. Steep descents and cliffs defend La Capilla on all other sides. The saddle forms, effectively, a wide ramp that leads up La Capilla, whose southern edge is delimited by a long wall, parts of which consisted of very large quarried stone more than 2 m long. Any access to the main sector was therefore limited and could be well-defended.
At the base of the East Ridge are at least three large blockfaçade constructions; El Caserón (see Figure 3) is the best surviving example. In plan, these punctuate the perimeter almost like big bulwarks projecting from the slopes, although they appear to be standalone and do not connect with the major surviving perimeter walls. Very fine-quality architecture features on the Caserón building: massive blocks and spall (typical of Recuay), alternating long and short ashlars for corners (see Topic 1991), and a slight batter (full height ca. 15 m). The façade is impressively even but is deliberately set back several centimeters near the top.
Despite previous testing, the function of the blockfaçade buildings remains unclear (Grieder 1978). They may have been special funerary structures. They appear to be largely solid, and the short platforms/decks provide good vantages over the basin. The Caserón, at least, once had well-built rectangular structures on the top (Grieder 1978, 14). In the 2000s, the Ministry of Culture constructed a roofed covering (now destroyed) to protect the structure, and walls were also revealed then. It also features a small central entrance at its base, leading into a narrow ca. 4 m long shaft just large enough for a small adult to crawl through; it ends abruptly, and there is a narrower, poorly conserved duct above, perhaps a ventilation shaft or drainage canal.

Operation 10
The 2019 fieldwork conducted seven excavations in the La Capilla sector (see Figure 4). Five of these were small units, each ca. 4 × 4 m. Two larger exposures were also excavated, Operations 8 and 10. General excavation protocols were followed: trowel use, bulk sampling, and differentiation by natural levels. Screens (at least 5 mm mesh) were used for all soil.
Operation 10 examined one of the lower terraces directly below the top of La Capilla. This fairly level area crowned the northern edge of a steep drop-off and was protected by a northern, outer wall. The excavation found deep infilling sediments and stratigraphic buildup against the walls, some over 3 m in depth.
Operation 10 (ca. 20 × 5 m) ( Figure 5) was notable for its superimposed, agglutinated structures, built adjacent to a prominent exposure of bedrock (far eastern end). The surface architecture consists of ancient walls and expedient piling of fieldstones to raise their level. They are laid out as agglutinated rectangular rooms, whose entryways oriented movements into an open area (toward the interior). These expedient, enclosed areas for shelter and domestic activities were post-Recuay and made after some time of abandonment. Ceramics include local Late Intermediate Period (ca. A.D. 1000-1450) wares and fragments of Chimú-related blackware and Inca material, very often mixed with Recuay pottery.
The underlying levels are much better-preserved and correspond to the original floor and foundations of the rooms mentioned above. These are associated with a Recuay room complex of at least four differently-sized, well-made rectangular rooms. Doors connected the rooms east-west ( Figure 5A). All the floor deposits contained abundant Recuay-style ceramics ( Figure 6). A radiocarbon assay from a small hearth-like concentration of ash and charcoal on the beaten dirt floor level yielded an age of 1817 ± 19 B.P. (Table 1). The floor developed over a distinctive fill of stones, carefully arranged to fill in gaps and interdigitated together, like puzzle pieces, for strength and an even basal surface. Similar interlocking fill was found at Yayno (Lau 2010b).
The fill sealed and buried previous constructions, most notably two long and parallel walled spaces toward the south and east ( Figure 5B). The southernmost was a long passageway (Pasadizo 2). A shorter one to the north (Pasadizo 1, ca. 9 m long) ran parallel and was partitioned into at least two narrow compartments. Throughout the fill and floor deposits of the corridors and compartments, we found remains of Recuay period ceramics, many broken in situ and a few against the sides of walls. These consisted mainly of fancy serving pottery, including ring-based bowls, handled bowls, spoons, and tall-sided drinking cups. Preservation was very good, and abundant charcoal and camelid bone suggest deliberate refuse disposal, probably from high status contexts, into the parallel spaces. No major hearths were identified.
Rather than everyday domestic refuse accumulation, we suspect that the deposits comprised a special fill to mark the termination of use of the underlying rooms. The ceramics (see Figure 6) are internally consistent, suggesting that the activities of consumption and disposal occurred over a relatively short time, rather than as slow and sporadic regular accumulation. Firm interpretation must await systematic analyses and additional radiocarbon dating, but we base this supposition on preliminary observations of the context and its character: predominance of special serving wares and largely intact sherds (many refits were possible); plainwares and domestic implements are relatively infrequent; and, animal bone (many whole long bones) and other unusual finds, such as a near-complete set of panpipes and a partly-worked stone sculpture (bearing a D-shaped frame). These are unusual materials deposited before being sealed by a specially-prepared interlocked fill.
The compartments have low doorways, each roughly 1 m tall, connecting to the southern passageway. The smaller chamber contained two small interior niches. The north niche, opposite the doorway, went through the wall, allowing contact only with a bedrock outcrop (see Figure 5B), the underlying rock for this entire sector. The outcrop is higher and visible on the surface toward the east, so the chamber may have been a small, enclosed, private space to interact with the outcrop, since outcrops are often highlighted ritual features of Recuay settlements (see Lau 2011Lau , 2016. No human remains were identified in Operation 10. In the small compartment, a large portion of a young camelid, whole kaolinite handled bowls, and spoons indicate that it may have been a space also for periodic votive or offering rituals.

Operation 8
Our work on La Capilla included clearing of its thick overgrowth, mapping with total station and drone imagery, and sampling excavations. Grieder found burials along the southern part of La Capilla in rooms made out of fine stonemasonry, and he believed this to be a "burial temple" made expressly for a burial context (Grieder 1978, 41-58). Other nearby rooms were interpreted as houses. The new work indicates that the main part of La Capilla was probably not so much a temple or organic agglomeration of houses and other structures but rather a large, planned quadrangular enclosure complex ("Compound 1"), typical of major Recuay centers (Lau 2011).
Compound 1's northwestern and northeastern façades have largely collapsed; the remains of a southwestern wall can be discerned on the surface, but most of it is buried. These outer walls are thick and massive to bear the upper stories and interior walls. They survive to over 9 m in height today and would have made for a formidable edifice crowning Pashash (see Figures 3, 4). Exposed parts of the northwestern façade indicate at least one earlier wall, implying that the outermost was a renovation to widen and elevate the compound and/or shore up segments of the enclosure. The eastern wall, similarly, is a massive double wall, and portions of it measure over 2 m wide.
Like a Chinese palace or Roman villa, the interior rooms of Compound 1 feature along the sides and have doorways that opened toward an interior, enclosed courtyard. This was a common architectural form in the Andes and, by Inca times, was known as a kancha (Berquist 2021;Herrera 2008;Hyslop 1990). Today, the interior open courtyard has filled in with rubble. The southeastern façade still features a stringcourse, a ledge-like cornice of projecting stones (roughly 4 cm tall and 3-4 cm deep), which demarcates the compound's upper floor. This feature probably adorned all of Compound 1. A small segment of another nearby structure (Compound 2) also has a stringcourse. None of the extant walls today feature windows or doorways. The main entrance was probably from the east via the natural ramp of La Capilla's saddle, where traces of a ramp and stairs existed (Grieder 1978, fig . 13).
Operation 8 investigated the southern corner of Compound 1 (Figure 7). This excavation was located just west of the burial context (Cuts 10, 11, and 12) found by Grieder (1978). The upper portions had been disturbed by post-Recuay activities, but the levels below were intact, and the pit ultimately yielded the best stratigraphy known for the La Capilla sector. The excavation went down nearly 3 m and encountered two superimposed beaten dirt floors, comprising the main use and living surfaces of the Recuay complex. The floors were hard and dark-colored layers, reflecting regular burning, refuse buildup, and decomposition of organic materials.
The rooms were quadrangular, and the corner room featured a well-made doorway. These built over an earlier corner room and floor (evidently without a doorway). Both floors were broken but made some use of wide and flat flagstones. Over these were large amounts of Recuay refuse, including fancy kaolin pottery and dark ashy lenses near burned spots. Compared to Operation 10, camelid bone was very rare in Operation 8 (also Grieder 1978). The Operation 8 materials were from drinking and libation activities (but without much camelid consumption) and ultimately breaking/disposal of predominantly kaolin serving vessels.
Underneath the lowermost floor were thick deposits of soil, coarse stone blocks and chips, and spalling bedrock fragments, accompanied by much less intact and substantial refuse. These layers were rubble-fills for levelling activities that prepared the corner for the earliest construction of Compound 1. Cultural remains petered out in the lowermost strata. Diagnostic ceramics included Quinú style (local white-on-red) pottery ( Figure 8) and occasional inclusions of plastic-decorated wares associated with Chavín and pre-Recuay styles (Figure 9). Associated with the early pottery are small burned areas and a partly-walled semicircular feature with a modified bedrock hollow (ca. 70 cm across, 40 cm tall). It may have been a cist or storage space but had been partially destroyed/dismantled by the time of preparing Compound 1's foundations. A radiocarbon sample (OxA-41369) for this level yielded an age of 1816 ± 18 years B.P. The assay effectively dates the rubble associated with the foundations of Compound 1's southern corner.
North of the semicircular feature and the destroyed footings of a large wall was a compact four-sided compartment with slightly rounded corners (see Figure 7). At its base, it measured ca. 1.5 × 1.7 m. It was formed of a well-built wall to its west (part of the footings), a partition wall added to its north, and poorly preserved or irregular walls on the southern and eastern sides.
About 70 cm into the layer (mainly wallfall), the soil matrix became much darker and looser, particularly along the western and southern walls. These were the topmost deposits of a major offering context. The uppermost finds included gold lámina (hammered sheet) fragments (including small perforated discs and a repoussé butterfly with blue-green stone inlays), small camelid figurines, and several whole pottery vessels, all in Recuay style. One vessel depicted a condor-like bird, and inside it were two small camelid figurines/pendants ( Figure 10A). Around and beneath several large, flat stones were additional ceramic items, including ring-based bowls, jars with simple bowls as lids, and dozens of additional camelid pendants/figurines (Lau 2021). The excavation also uncovered several modeled kaolin figurines depicting humans ( Figure 10C-E), the first time such figurines have been found in context. Each of the standing figures is a well-attired male with fancy headgear; they hold small camelids in their left arms.
Underneath these, in the southwestern corner of the compartment, a cache of over two dozen well-preserved gilt metal pins (most 15-25 cm long) were found, bundled together with string. Some of the string was still preserved halfway up the shafts, held in place by the blue-green copper oxidation. Around the pins were two small, flat, stone plaques, light gray colored, and carefully ground to a trapezoidal form with smooth faces (ca. 3.4 × 7.1 cm) ( Figure 11). Several lenses of bright magenta-colored ground pigment, heavy and probably with metallic content, were also encountered nearby. The stones may have been used as small plinths to hold or place the items. Many camelid pendants were also part of this initial find locus.
All the pins were of gilt copper, and most were found in pairs. Such pins were worn and affixed to garments at the shoulders, with the pinheads facing forward; the pins have holes to run thread through. Because of their size and weight, the holes may have been necessary to fix the pins in place. Fragments of white cotton cloth were found nearby. One large fragment shows a paired warp and paired weft structure; the white cloth also shows substantial verdigris staining from copper oxidation.
Some of the cast pins are solid metal, while others are hollow. The solid pin heads show the heads of felines, foxes, and the Recuay crested animal motif; a pair also depict a human head with a stepped headdress. Some pins portray full figure birds and pairs of birds, perhaps macaws and condors. The birds show cropped tails, and some grasp tubular rods. There is also a multifigure composition, showing two birds atop two prone human figures. The cast hollow pins are particularly notable (Figure 12). Previous technical study of pins found in the original burial indicate they were made with at least three sequential castings ("castings-on") to fabricate and join discrete elements: shaft; truncated conical pin head, and fringe of spiky projections (Grieder 1978; Velarde and Castro de la Mata 2010). The pin head features two areas for figural elaboration: the tapering side which wraps around the head and the top surface of the pin head, usually set back into the head and ringed by the spiky fringe. The side relief typically features a band of three profile crested animals. The top is more variable, showing frontal human and feline/crested animal heads, profile felines, and standing frontal human figures. The eyes and other significant parts of the figures were often inlaid with tiny green, white, and turquoise-colored inlay stones.
Adding to their technical sophistication, at least eleven of these pins were also rattles. Nine were intact enough to yield a clinking sound when lightly shaken (Supplemental Material 1). In addition, one broken pin head revealed four small, rounded, white stones, each finely chipped to ca. 4 mm in diameter. The interior chamber also contained light chalky material, probably kaolin clay, between layers of copper metal; a previous study noted the clay was to buffer and refract heat during casting procedures (Velarde and Castro de la Mata 2010). It seems possible that the clay also helped modify the resulting rattle sound.
The clinking rattling of such pins probably complemented the louder clanking of ten copper bells (ca. 5.5 cm wide), strung together as hand or body bells, found in the doorway offering (Grieder 1978, 118-119). The sounds of ancient funerary and ancestor ceremonies are little known, but it seems likely they were crucial for the affective and public signaling of the wearer and the ritual collective, and also marking off ritual time (Arriaga 1999;Helmer 2015;Scullin 2015); rare metal items may have been particularly significant and distinctive in the ancient past (Carcedo Muro 1998, 256-262).
Flanking the pins in the offering context were several additional pots, including an effigy vessel of a camelid ( Figure 10B), depicted lying in prone position and bound in a net, probably to denote its sacrifice (Lau 2021). A charcoal sample (OxA-41441) from this locus yielded an age of 1806 ± 19 B.P.; this agrees with previous radiocarbon evidence from the burial/offering area (see Table 1). Immediately underneath this layer of items were several flagstones laid flat. These stones may have been part of a multi-part termination ritual, for they were placed over the offering objects, apparently dividing moments and contents of the offering; they covered over, but did not smash, the underlying items. Flat stones, often very thin (most under 2 cm thickness) slate or schist panel-like segments, separated offerings and demarcated special offering spaces throughout Operation 8 (also Grieder 1978); sometimes they were arranged to make small compartments.
Under the arrangement of flat stones was a much darker soil in an irregular area (ca. 60 cm across) showing high organic content, probably from decomposed cloth and wooden remains (and perhaps human tissue?) (also Grieder 1978, fig . 28); no human bones were found here. Some of this deposit was removed en bloc for future study. This comprised the main offering locus, in which a wide variety of items were found. Interspersed between what looked to be folds of cloth were distinct lenses (ca. 10-15 cm wide) of powdered pigments and hundreds of perforated beads. The powders were bright vermillion red and a lighter magenta (purplish red). The beads were of greenstone (from opaque greens to turquoise color) and tablet-shaped (most ca. 3-4 mm), and others were left in rough knobby or chipped irregular forms. There was also, surprisingly, preserved wooden material (including small pins with zoomorphic heads, carved wood items, and fragments of basketry-like material). Many of the wooden items featured small green and white stone inlays. These are rectangular, round, triangular, or diamond-shaped and often feature in sockets for eyes and ears or as decorative elements.
Hammered gold foil was also common within and around the dark organic locus in various small forms: Table 1. Radiocarbon dates mentioned in the text, including Grieder (1978) and Lau (2004Lau ( , 2011 perforated and unperforated circular discs (ca. 7-10 mm diam.), unperforated concave láminas in lozenge/ovoid forms, narrow tubes (2 mm wide; of variable length, ca. 3-5 cm), láminas of profile birds and triangular heads (viewed top down), perforated bands, and other geometric forms. We suspect the foil elements were probably sewn onto cloth and/or attached onto other items of attire. There were also miniature pins with the spiky fringe and truncated conical heads; some were cast and others were constructed out of gold-metal foil.
Along the western wall, more hammered items were found, most notably two hammered gold, wood, and mosaic earspools, with hollow balls of foil around the edges (see also Grieder 1978, 55). Unfortunately, many of the carefully fitted bluish-greenstone stone inlays have fallen off the backing (metal plate with brownish yellow, powdery material), so it is presently impossible to determine the central mosaic design. Such earspools were made and worn by Moche peoples of the northern coast (e.g., Donnan 2004, 80), so the examples in Pashash strongly indicate elite-level interactions with coastal peoples.
Lapidaries were also abundant, including several small, hollow gold beads and abundant greenstone beads of different shapes, the most impressive being in lozenge and long tube shapes. A rock crystal was also included in the cache.
Just underneath the dark deposit locus was another cache of copper pins. These were less well-preserved than the upper cache. Many featured solid cast heads, and a few were with truncated conical heads (rattle type). All were at least partly broken, disfigured, and/or heavily encrusted by copper metal corrosion. Another contrast with the first cache was that almost all of the pins had intentionally bent stems, several curving back 180°to the head. Laboratory analysis and conservation are ongoing, so fuller description will follow.
Near the second pin cache were a series of copper pendants; most of the complete examples measure about 2.5-3 cm long. Most show the same kinds of animals depicted on the pins, including birds, felines, and crested animals. Many of the pendants are found in pairs and have holes through their neck or back. Notable figures include: a feline with a human head; a seated predatorial animal (feline?) grasping another figure or head with its claws and consuming it; a feline (?) consuming a serpent or its own tail; and, a bird on the back of a human. The pendants were found together, suggesting they may have been placed or strung together or featured in/on the same object (cloth bag?) as a collection. Many of the metal pendants have string fibers still threaded through the holes, suggesting they were attached at the time of their deposition.
Other pendants were of stone, most under 2 cm in height, and in red, brown, and black stone colors. Some represented sitting/rampant felines in dark reddish purple stone; others were of lizards or perhaps caymans. Others appear to represent vessels with flaring rims; some show two or three (conjoined) globular bodies. It should be mentioned that camelids featured in over 109 small ceramic artifacts and were the most commonly depicted animal form (Lau 2021). All are of modelled and fired clay: pendants, figurines, and figural vessels. The camelid depictions are especially important because of their frequency, forming a major part of the cache's offerings, along with other accoutrements, personal items, and other markers of elite identity. They signal the growing reliance on camelids and their increasing integration into Recuay cultural production.
Operation 8 (Extensión Este) continued several meters below the offering. A 14 C point sample (OxA-41368) for the lowermost find level, underneath a layer containing Formative ceramics (see Figure 9), yielded an age of 2560 ± 18 B.P. (see Table 1). The date is the earliest date yet known for Pashash and substantiates the ceramic evidence for pre-Recuay occupation. The wider context and activities of this component remain unclear, but the broken fancy and utilitarian ceramics may indicate a light residential and/or ritual use on the hilltop.

Discussion
Recent fieldwork at Pashash thus evidences major changes in patterns of settlement, architecture, and overall cultural practices in the northern Ancash highlands during the early centuries A.D. All these mark transformations in the physical context and material culture of political centralization in Peru's North Highlands compared to the preceding period.
Pashash first saw occupation as early as the 7th century CAL B.C. and continued well into late prehistory. The principal occupation and building programs date to the site's main Recuay occupation, A.D. 200-700 (see Table 1). Monumental construction probably began on La Capilla by the later 2nd century A.D.
The area under the southern corner of Compound 1 shows the earliest evidence of occupation at Pashash. The lowermost excavation deposits revealed a range of pre-Recuay diagnostics (see Figures 8,9). A handful of fragments of polished monochrome pottery bear the stamped circleand-dots decoration typical of Chavín-related Janabarriu styles (Burger 1992). The same level revealed ceramics with plastic decoration (punctate, zoned punctate, patterned-wiping and -brushing, and incised) best known from the North Central coast and middle valley areas, such as of Santa (Wilson 1988), Nepeña (Chicoine 2011, figs. 5-6;Proulx 1985, pls.1-2, 9;Shibata 2010), and Casma (Druc 1998, figs. 12-13). Decoration features usually on the upper exterior register of bowls, and this set of material is found with large neckless ollas (globular cooking and storage vessels) (see Terada 1979, pls. 61-66, 103-106). The pottery suggests a fairly wide region of stylistic interaction, especially with peoples and regions south of the Tablachaca.
Deposits of this first occupation are overlain by a level bearing neckless ollas and white-on-red pottery, e.g., white paint on red slipped oxidized redware (see Figure 8). Grieder (1978, 63) called this material "Quinú" and suggested the possibility of two phases. The linear painting is typically on the exterior of bowls and, less commonly, jars. During the time following the cessation of Chavín-related pottery and interaction, other vessels show continuity in pottery production and use, particularly for cooking/serving activities. White-on-red can be found with styles coeval with Chavín and additional special wares with incised designs. Presently, we cannot discern any substantial architectural context associated with the early pottery.
Pottery changed dramatically a few centuries into the 1st millennium A.D., soon after the footings were established for Compound 1 (southern corner). The occupation is represented by denser and more complete pottery remains. The Recuay pottery is also very distinctive because of the emphasis on kaolin fabrics (see Figure 6), polychrome painting including resist (see also Figure 10), and the range of new designs (also Grieder 1978). Bowl forms are predominant, frequently with flanged rims, ring bases, and corniform handles. Other special forms were large basins, spoons, globular jars with open bowls for lids, and small effigy bottles. Redware necked jars replaced neckless ollas. The fancy ceramics tend to be of kaolinite, but there is also the occasional polished blackware (Figure 6M-N) and unpainted kaolinite, sometimes with incised designs ( Figure 6G). We have not yet found evidence to substantiate the three Recuay substyles sequence proposed by Grieder (1978). The work also has yet to identify material resembling Moche, Cajamarca, or Wari style pottery, nor has there been obsidian or substantial marine shell.
Subsequent re-occupation occurred on the northern terrace (Operations 10, 11, and 12), where there were clear post-Recuay components. Reuse of the terrace and older Recuay walls by Middle Horizon and Late Intermediate Period groups are associated with plainer terracotta wares, sometimes enhanced by plastic decoration. Grieder (1978) called the post-Recuay Middle Horizon component "Usú." Occasional sherds of Inka style and blackware associated with late northern coast groups were identified, as well. Previous radiocarbon evidence shows activity into colonial times (see Table 1).
Pashash was a settlement established, at least in part, for its defensive positioning. Local groups committed considerable labor for massive build-up of revetment walls, Caserón style block-façade structures, and perimeter walls across the settlement. Defense clearly also stressed very restricted access to the top of La Capilla and its key monumental buildings, each of which was protected by very high façades that fully enclosed the compounds. To date, no unequivocal or typical Recuay combat weapons, like metal maceheads or shields, have been found. Excavations have recovered stone axes and stone donut-like rings, which could have been used as clubheads. Nonetheless, iconographic evidence of warfare practice exists in the local pottery and stone sculpture from Pashash: helmeted warriors, warrior figures with shields, long scepters, and maces, trophy heads and hands, and captive-taking imagery (Grieder 1978, figs. 130, 152).
The new work updates our understanding about Pashash's layout, architecture, and chronology of construction. Compound 1 was the main edifice on La Capilla, a form of monumental construction very different to the open temple and plaza complexes of the previous Early Horizon. It was a walled enclosure with interior side rooms surrounding a courtyard. Estimated to be over 35 m on a side, it is one of the largest quadrangular compounds documented in highland Ancash, on the same scale as Structure AC-5 at Honcopampa (Tschauner 2003) and Compound C41 at Yayno (Lau 2010b;also Apolín 2009, 128). Another similarity between Compound 1 and these compounds is the very fine stonework on their interior façades. Pashash's best extant stonemasonry, including finely cut ashlars, adorned the southern interior rooms of Compound 1. This may be circumstantial evidence for a main entrance on the opposite side, the northern façade. Tschauner (2003, 209) observed that the most elaborate masonry on Honcopampa's compounds, some with the largest megaliths, adorned the interior patio walls directly opposite the main access into the courtyard. This was because local builders fashioned visually impressive displays as one entered the interior spaces.
Like these other settlements, Compound 1 was not a standalone construction; a circular compound (Compound 2) adjoins it. The discrete compounds, we believe, probably indicate multiple co-resident groups at these sites. Future work at Pashash will look to study other compound structures further. It is worth noting that Operation 8 and a previous pit (Cut 9;Grieder 1978) found everyday domestic refuse, indicating that some of the rooms were used for common habitation activities. This is also consistent with general residential use of compounds at Yayno (Lau 2010b).
Ultimately, we see Compound 1 as a "palatial" construction, and this designation is not given lightly (Couture 2004;Evans and Pillsbury 2004). Compound 1 shows tremendous labor investment, more than any other building at Pashash. It also contains very high-status indicators of social differentiation, both when compared to other parts of Pashash and when compared to other known Recuay sites. This includes architectural features (building location, double thickness walls, fine stonemasonry and stonecarving, stringcourse, and restricted access), as well as high status residential activities and assemblages resulting from unprecedented noble practices.
The key evidence for differentiation at Pashash concerns ritual practices in Compound 1. Located within the compound, the burial and offering activities repurposed part of the original complex. No doubt the flexed individual in the original burial chamber was important to the wealthy group based in the compound, probably an important matriarch or patriarch of the collective (Grieder 1978, 58, 181). The poor preservation of the skeletal remains did not allow sexual identification. Grieder (1978, 54-55) left the question open but noted that earplugs suggested an adult male, while weaving equipment indicated a female. Shawl pins are also typically associated with use by women. For this person, parts of Compound 1 were sealed off and partitioned for discrete ritual offerings and subsequent burials. An extended individual was buried later with fewer offerings, some time after the original flexed individual (Grieder 1978, 52). At least two other burial remains in the area are known: a flexed bundle found ca. 4 m to the northeast (S. Castillo, personal communication 2019) and a disturbed group of bones found outside the "burial temple" doorway, found with a cache of Usú (post-Recuay) ceramics (Grieder 1978, 39, 70).
This overall context comprises one of the best examples of early highland burial cult and gives a sense of the wealth, duration, and approximate sequence of veneration efforts. Grieder (1978) reported three offerings in and around the burial compartment; these are located a few meters north of Operation 8. The founding "burial offering" (initial and lowermost) and a "doorway offering" were formal caches. The third "fill offering," meanwhile, was "different in its casualness […] something like family members gathered around the grave, smashing vessels and tossing them in or laying down treasured objects" (Grieder 1978, 58). The recently-found cache comprises a fourth discrete component in the array of offerings made to the burial. Radiocarbon evidence (see Table 1) indicates a period of some three or four centuries in which activities of offering continued to be made to the burial area. Because the new date (OxA-41441) is the earliest (also TX-1329 and TX-1824), it may date to the time of the original burial. Additional 14 C samples from Operations 8 and 10 are currently in process.
Despite the similar pottery and artifactual patterning (careful stacking of items, lenses of organic material, and pigment offerings), it is notable that the contents of the caches differ considerably. There were stone bowls, stone spindle whorls, whistles, carved stone blocks, and many dozens of pottery vessels in the "burial" and "doorway" offerings (Grieder 1978, 51-58). The other caches also contained a range of modeled zoomorphic figures and many effigy vessels portraying zoomorphic creatures. Pins, small lapidaries, and objects with camelid imagery, meanwhile, were much more numerous in the Operation 8 offering. Also, the Moche style earspools indicate very unusual, long-distance connections.
Although their precise meanings are difficult to ascertain, it is reasonable to believe that the multiple components and methodical sequence of the offering program helped construct and venerate the deceased subject(s) over time. What's more, the burials and caches do not outwardly emphasize the priestly connections to cultic identity and shamanism seen earlier with Chavín. Rather, we see perhaps more familiar and prosaic inclusions like weaving tools, camelid figurines, and cloth. While the deceased certainly may have held ritual functions (e.g., pin-rattles and personal adornments), these seem to be associated with roles in village society, in their public celebrations and feasts, rather than the stuff of expansive cult and shamanic-based visionary practices.
The overall faunal assemblage is substantial (3990 NISP and 45.7 kg of animal bone, mainly from Operation 10). Preliminary examination suggests that camelids comprised the bulk of this sample. If patterns are similar to coeval Recuay tradition settlements (Bria 2017;Lau 2010b;Ponte 2001), domestic camelids were routinely consumed and their remains discarded on-site. Pre-Recuay deposits show very negligible faunal remains. The emphasis on camelid representations on figurines and human-camelid figures is also consistent with the growing reliance on and prestige of domestic camelids. Notwithstanding, the faunal assemblage requires systematic study, and future field seasons will enlarge the sample.
Finally, the Pashash work provides a unique window into how leaders increasingly depicted and perceived themselves. This change in their public recognition and artisanal production, we believe, comprised one of the key innovations of the first centuries A.D. The basis of Early Horizon elites, pegged to their priestly roles, gave way to a different kind of legitimation. What our research is clarifying is how, by Recuay times, the theocratic power associated with Chavínrelated centers receded in favor of elites who self-identified more as lineage heads, with command over herding, warfare, and related domains centered on the compound, or "house," as a form of organization (e.g., Lau 2010bLau , 2013Moore 2005).
Depiction of leaders was very rare before, and so we suggest that the secularization of authority was mirrored in the secularization of art imagery. Pashash was part of an increasingly widespread pattern, and newly-entrenched during Recuay times, in which leaders became integral components of new kinds of image-making, as seen in the chiefly figures found in ceramics and carved monoliths. There are certainly more elaborate examples of the multifigural compositions reported here (see Figure 10C-E). But these are crucial because they are among the earliest known in the Recuay tradition and directly link to a political center and offering ritual within a high-status compound.
Leaders began to depict themselves with certain kinds of objects as signs of distinction. Camelids, in particular, intervened as part of their wealth and status. In addition, emblems of warriorhood (e.g., weapons, helmeted headgear, and trophy hands/heads) helped constitute the other newfound source of their identity. These new domains for constituting prestige were part of the dramatic shift in cultural values between Chavín and Recuay times. Such effigies have never been reported from their original contexts before and certainly not from a known political center, where we would anticipate precisely such administrative and public ritual displays of lordly practice.
The personal ornaments and representations that pair lords and camelids are among the earliest iconographic expressions of the animal as a form of wealth, noble identity, and authority in northern Peru. Notably, they seem to present the lord as mediator between complementary opposites, an individual who brings together the masculine and warfare associations (club, shield, and flute) of the right side versus the domestic associations (camelid and offering) of the left.
Finally, it is not a coincidence that centers of political power came to occupy prominent hilltops and summits. This was not merely owed to their defensive or strategic position. Rather, we believe that leaders and ancestrality tied to landscape co-developed in a process of cosmo-political ascendancy (Lau 2016(Lau , 2018. It is worth remembering the strong equivalences made between authority and high stony places during colonial times, especially mountains, hilltops, and rocky crests seen as the patrimony, indeed the embodied and inherited remains, of sacred ancestral beings or huaca (Bray 2015;Gose 2016;Salomon 2018). As prominent settlements like Pashash and other central places became established, it seems plausible that the high places themselves came to stand for political agents and embodied the people they housed in life and in death. Attributing subjectivity and agentive power to such places was common during colonial times and remains important in oral traditions of the Cabana region today. Local folklore presents Mashgonga as the tutelary mountain (ancestral divinity) of the region, and Pashash intervenes as a sacred hilltop which mediates the peoples of the herding lands above and the agricultural lands below (Cuba Manrique 2018).

Conclusion
In this study, we have detailed recent archaeological fieldwork that illuminates the rise of native lordships and secular leadership in Peru's North Highlands during the 1st millennium A.D. The work helps reconstruct, effectively, a trajectory for the process of their uptake in north-central Peru. The new evidence from Pashash complements similar trajectories in adjacent valleys, including Nepeña, Virú, and Huamachuco, but in varying configurations and scales of centralization and associated with distinct cultural groups.
The process of organizational change coincided with new orientations very different from what characterized Chavín's height and its interaction network of temples and villages.
Four key factors featured: the formation of defensible nucleated settlements on hilltops; development of palatial compounds; burial cults within the compounds; and, intensive cultural differentiation of noble identity through markers of wealth, warfare, and authority.
None of the four, it seems, stands out as a prime mover or worked to the exclusion of the others. It also remains to be determined how these factors impacted coeval communities, especially over time. We suspect that these factors, once prevalent, continued to be decisive in shaping local trajectories across northern Peru and beyond. The current record indicates a strong synergy which helped to effect special conditions for emergent complexity and the rise of segmentary lordships (curacazgos and señorios) during the early 1st millennium A.D.
By obligation, multiple lines of evidence will be needed to assess and compare cases for the uptake of segmentary polities elsewhere, especially their timing and regional variability. Recent fieldwork at Pashash is instructive because it was possible to document reuse of parts of a major residential compound for burial cult over many generations, even as living and palatial activities continued unabated. Paradoxically, it may be the persistence of attention to the esteemed dead that tracks something of the vigor and wealth of the living, descendant group. This pattern comprises a major social innovation after Chavín, and future research will continue to shed new light on its impact for ancient Andean societies.

Notes on Contributors
George F. Lau (Ph.D. 2001, Yale University) is Professor at the Sainsbury Research Unit for the Arts of Africa, Oceania & the Americas (University of East Anglia, Norwich) and Editor of the World Art journal. He specializes in the arts and archaeology of the ancient Americas. His research and publications focus on the Central Andes, especially on the material culture, complexity, ritual practices, and diachronic change of northern Peruvian groups.
Milton Luján Dávila (Licenciatura 2006, Universidad Federico Villareal) is an independent researcher and archaeologist who has led field archaeology projects throughout Peru, including in the Chillón, Lurín, Ica, Mantaro, and Cajamarca regions. He is especially interested in burial practices, landscape and water practices, ceramic analysis, and archaeological site conservation. Jacob L. Bongers (Ph.D. 2019, University of California Los Angeles) is Senior Research Associate at the University of East Anglia and is a postdoctoral researcher on the Pashash-Cerro San Isidro "Divine Lordships" project. He specializes in GIS and digital archaeology applications, settlement patterns, and ritual and materiality of death practices of late prehispanic Central Andean groups.
David Chicoine (Ph.D. 2007, University of East Anglia) is Professor of Archaeology in the Department of Geography and Anthropology at Louisiana State University. His research and publications focus on the ancient societies, material culture, and visual arts of the Central Andes, especially coastal Peru, where he has been conducting excavations since the late 1990s.