Shumway Ruin and the Late Pre-Hispanic Period in East-Central Arizona Scott Van Keuren^ ABSTRACT. Shumway Ruin is a large Ancestral Pueblo settlement in east-central Arizona occupied during the fourteenth century a.d. The settlement was one of six villages in the Silver Creek drainage established duringa phase ofcommunity reorganizationthroughout the North American Southwest. A new architectural map shows that the village was composed of several contiguous room blocks built around two major plazas and a smaller plaza or possible great kiva. Excavations in one of the large plazas and two nearby rooms demonstrate that Shumway Ruin was one ofthe latest occupied Pueblo villages in the region and was a possible producer oficonographic-style red ware pottery. At the close of the thirteenth century a.d., RESEARCH BACKGROUND Ancestral Pueblo communities in the Silver Creek drainage, Arizona, underwent an importantphase Shumway Ruin is located along the northern ofsettlementreorganization. Earlier smallvillages Silver Creek, a tributary of the Little Colorado were abandoned and populations aggregated at River in east-central Arizona north of the six large towns (Figure 1). These sites set a new Mogollon Rim (Figure 1). The village lies along threshold in community size, the scale of extra- the MogollonSlope, a naturalgradientatthe edge mural ceremonial spaces (or plazas), and perhaps of the Colorado Plateau in which drainages flow commitment to intensive agriculture in the drain- from the edge of the Mogollon Rim north to the age (Kaldahl et ah, 2004). One or more of these Little Colorado River. Today, private landhold- villages also began to produce a new iconograph- ings account for much of the northern Silver ic-style red ware pottery type (Fourmile Poly- Creekdrainage,whilemostofthe southernhalfof chrome) depicting birds, zoomorphic figures, and the drainage lieswithintheApache and Sitgreaves other symbols (Table 1; Carlson, 1970). The National Forests. In this contribution, I outline appearance of this ceramic type presumably the late pre-Hispanic cultural sequence in the signals the emergence of new religious beliefs at northern Silver Creek drainage and review the a time when ceremonial spaces expand in size history of research at Shumway Ruin and nearby (Adams, 1991). These changes in the Silver Creek Pueblo IV period sites before presenting the area reflect broader cultural transitions that results of recent fieldwork. marked the beginning of the Pueblo IV period (a.d. 1275-1540) in the greater Southwest. This THE LATE PRE-HISPANIC PERIOD IN THE paper documents new fieldwork at Shumway SILVER CREEK DRAINAGE Ruin (AZ P:12:127 ASU), one of the largest of the six latest Silver Creek Pueblo villages and one Archaeological sites in the Silver Creek area span ofthe lastintactarchaeological sitesfromthe area a broad time range, beginning with evidence for that was occupied during the period in which Paleo-Indian and Archaic period occupations and changes in community organization, craft econo- extending through the arrival of Apache groups mies, andthecompositionandscale ofceremonial during the proto- and early-Historic periods (see spaces occur. Other Pueblo IV period sites have Lightfoot, 1984; Newcomb, 1999). Shumway beenlargelydestroyed by lootingortheexpansion ofmodernsettlements. Inthis report, I summarize iRnuinthies oSnielvoefraChraenedkfularoefaAnceessttabrlailshPeudeblaot sitthees excavations conducted in 2002 and provide the beginning of the Pueblo IV period (1275-1540; first accurate architectural plan view of the village. This work sheds new light on the timing see Table 1). This period of aggregation and and occupational history ofthis important area of appearance of large towns marked the last phase ofpermanent Pueblo habitation in this part ofthe the Ancestral Pueblo world and represents the upper Little Colorado Rivervalley (Kaldahl et ah, first step in a long-term research project in the 2004). Shumway Ruin and other late pre-Hispan- area. ic population centers discussed in this report 1. Department of Anthropology, Natural History belong to the broader Western Pueblo region Museum of Los Angeles County, 900 Exposition (Reed, 1948); the people who occupied these sites Boulevard, Los Angeles, California 90007, USA. Email: were ancestors of the Hopi, Zuni, and other [email protected] modern Native peoples. Contributions in Science, Number 508, pp. 1-19 Natural History Museum ofLos Angeles County, 2006 2 Contributions in Science, Number 508 Van Keuren: Shumway Ruin Pueblo peoples in the Silver Creek drainage thirteenth-century villages in the northern Silver lived at a series of small, dispersed villages near Creekdrainage recorded byprevious surveys (e.g., the close of the thirteenth century. Most of these Lightfoot, 1984; Neily, 1988) resemble contem- sites are contemporaneous with Linden Phase poraneous villages in the Hay Hollow Valley to villages to the south and were occupied prior to the east (see Martin et ah, 1964, 1975). In the the appearance of Pinedale-style ceramics (see northern part ofthe drainage, these villages range Mills and Herr, 1999). Aside from the University between 20 and 50 rooms with contiguous room of Arizona’s recent work at sites in the southern blocks built around small plazas or possibly great Silver Creek drainage (Mills et ah, 1999b), few of kivas. In 2002, my crew relocated and mapped these thirteenth century villages have been exca- Flake Ruin and AZ P;12:87 [ASUl, two northern vated or documented. Many of the mid- to late- Silver Creek villages dating to this period. Both Table 1 Chronological sequence ofthe Silver Creek area Dates (a.d.) Period Phase Diagnosticpottery style Major sites 1200-1275 Pueblo III Linden Snowflake, Reserve/ Flake Ruin; AZ P:12:87 ASU; AZ Tularosa, and P:12:99 ASU Tularosa styles 1275-1325 Pueblo IV Pinedale Pinedale style Bailey, Pinedale, Showlow, Fourmile, Shumway, and Tundastusa ruins 1325-1390 Pueblo IV Canyon Creek Fourmile and Gila Pinedale, Showlow, Fourmile, Shumway, styles and Tundastusa ruins Contributions in Science, Number 508 Van Keuren: Shumway Ruin 3 share the same basic architectural layout as ably, its occupants either joined one of the five Broken K Pueblo (Hill, 1970), with contiguous other villages in the drainage or moved out ofthe room blocks enclosing a small central plaza. The area. It is also at this time that there is clear ceramic assemblages of these villages include evidence for the arrival of migrant groups at one Reserve/Tularosa-, Tularosa-, and Snowflake- of the remaining villages, Fourmile Ruin (John- style Cibola White Ware, along with minor son, 1992). Shumway Ruin was part of this amounts of both Showlow Black-on-red (Puerco cluster of large Pueblo towns that were first Red Ware) and early Pinto Black-on-red (Roose- established in the late 1200s and then expanded velt Red Ware). These decorated assemblages by the early to mid-1300s. The second phase of indicate that the occupation ended at or around aggregation in the area was short lived, and there the appearance of early Pinedale-style pottery, is no evidence of residential Pueblo occupation perhaps by the 1270s or 1280s based on the after 1400. Indeed, the latesttree-ringcuttingdate dating of Pinedale-style Roosevelt Red Ware of 1384 at Showlow Ruin suggests that the elsewhere (Montgomery and Reid, 1990). Un- decline of these villages began in the last quarter fortunately, three to four other Pueblo III period of the fourteenth century (Mills and Herr, 1999). sites in the northern Silver Creek drainage were bulldozed by looters prior to the 1990s, and it is PREVIOUS RESEARCH now impossible to characterize the layout or artifact assemblages of most of the early plaza- Archaeological work in the northern Silver Creek oriented sites that immediately pre-date the drainage began a century ago with Adolph aggregation ofShumway Ruin. It is probable that Bandelier’s (1892) reconnaissance in the area, many of the fourteenth-century villages began as followed by Jesse Walter Fewkes’s (1904) collec- smaller thirteenth-century pueblos similar to AZ tions-motivated excavations at Fourmile and P:12:87 lASUl and Flake Ruin. These earlier Pinedale ruins in 1897 and Walter Hough’s brief occupations are presumably obscured by later visit to Shumway and Tundastusa Ruins in 1901 construction. (Hough, 1903:302). Leslie Spier (1919) later The production ofPinedale-style pottery in the visited the area and published the earliest sketch Silver Creek drainage is associated with the first map of Fourmile Ruin. A few decades later, Emil phase of population resettlement into large W. Hauryand Lyndon L. Hargrave conductedthe enclosed-plaza towns (Mills and Herr, 1999), first professional excavations at Pinedale Ruin local changes that mirror settlement shifts across and other fourteenth-century villages in the Silver the southern Colorado Plateau by 1300 (Adams, Creek drainage as part of the Third Beam 1989; Adams and Duff, 2004; Bernardini, 1998; Expedition (Haury and Hargrave, 1931). With Cordell et ah, 1994). In the southern Silver Creek the exception of Haury and Hargrave’s work, area, large pueblos include Showlow Ruin, Bailey most early excursions were brief and poorly Ruin, Pinedale Ruin, and Tundastusa Ruin (the documented, geared toward the recovery of latter site is just south of the Mogollon Rim in painted pottery to populate the culture area the Forestdale Valley). Shumway Ruin and exhibits of major museums. Fourmile Ruin were established in the northern Following up on Spier’s (1919) early reconnais- part of the drainage. By the first decade of the sance in the region, Kent Lightfoot assembled the fourteenth century, most Pueblo peoples in the most comprehensive settlement survey data in the area were living at one of these six towns. All northern part of the drainage in the 1970s range in size between 250 and 500 rooms (Lightfoot, 1981, 1984). In one of his three large (Kaldahl et ah, 2004; Mills, 1998), with contig- survey blocks, the Snowflake Study Area, Light- uous room blocks that fully or partially enclose foot recorded cultural resources in the vicinity of one or more plazas. There is evidence that many Shumway Ruin in the northern Silver Creek of these Silver Creek settlements welcomed drainage. Despite early debates on the contempo- immigrants from adjacent parts of the northern raneity of settlements in the region (see Lightfoot Southwest during the early fourteenth century and Most, 1989), his survey data demonstrate (Johnson, 1992; Mills, 1998). that the shift from small to large villages occurred A second phase of aggregation began in early at the close of the thirteenth century. Although fourteenth century with the expansion of existing fourteenth-century groups probably continued to villages (Table 1). This period of village expan- utilize field houses and limited activity sites, there sion roughly coincides with the appearance of is no substantive evidence for the habitation of Fourmile-style White Mountain Red Ware during villages outside of the handful of towns men- the 1320s (Carlson, 1970). Thisnewiconographic tioned earlier. Because several key thirteenth- style marks the cessation ofPinedale-style pottery century villages have been entirely bulldozed by production, and it was around this time that very looters since Lightfoot’s (1984) survey 25 years large plazas were constructed at Fourmile and ago, his settlement data and surface collections Shumway ruins and possibly at Pinedale and remain crucial to understanding settlement Showlow ruins. In the southwestern part of the changes in the region during the early Pueblo IV drainage, Bailey Ruin was abandoned. Presum- period. 4 Contributions in Science, Number 508 Van Keuren: Shumway Ruin Despite the paucity of work in the northern RESULTS OE THE 2002 EIELD SEASON Silver Creek area, recent research in the southern AT SHUMWAY RUIN halfofthe drainage by the University ofArizona’s Silver Creek Archaeological Research Project, ShumwayRuin sits atop a sandstone outcrop near particularly at Bailey Ruin, has begun to answer theconfluence ofthe ShowLow and Silvercreeks, important questions about the social and eco- at an elevation of 1738 m (5701 feet). Today, the nomic organization of Pueblo settlements in the surrounding area is pihon-juniper woodland Silver Creek area prior to 1330. A number of mixed with short grassland communities ranging published and unpublished surveys conducted in in elevation between 1700 and 1900 m. This arid the Sitgreaves National Forest and adjacent areas environ receives less than 14 inches of rainfall provide additional settlement data (Neily, 1988; annually (Kaldahl and Dean, 1999); the bulk of Newcomb, 1999). this moisture is delivered by late summer storms. The site is roughly bisected by a fence line that SHUMWAY RUIN AND THE PREHISTORY delineates two separately owned parcels of land near the small town of Shumway, Arizona. OE THE SILVER CREEK DRAINAGE Shumway Ruin is situated adjacent to one of the largest and most well-watered expanses of agri- Despite recent archaeological work, we still know cultural land in the drainage. Although a modern very little about the early Pueblo IV period in the canal cuts through the southeastern edge of the Silver Creek area, particularly during the phase of site and a portion of the northern edge of the site siectotnloegmreanpthiacg-gsrteyglaetiWohniatnedvMiolulangteaeixnpanRseidonWaaftreer was bulldozed in the 1980s, the village is well preserved. and large plazas appear. Because of rampant pot Mapping and excavation at Shumway Ruin hunting and vandalism and the location of key were conducted between April and June 2002. sites on private land, this cultural landscape Our first effort focused on the collection of remains largely inaccessible to archaeologists. architectural data for use in drafting a plan view Many of the Pueblo IV period towns either are and reconstructing the settlement history and nowentirelydestroyed (suchas Showlow Ruin) or growth of the village. The second goal of the were greatly disturbed by heavy machinery in the field season was to excavate intact and un- past few decades (Fourmile and Pinedale ruins). disturbed contexts. This report provides a pre- Shumway Ruin is mostly intact, and archaeolog- liminary tabulation of all recovered ceramics, ical work at this site is essential to answering which clarifies the timing of occupation and the questions aboutthe timing, structure, andcultural cultural identity of the village’s occupants. composition of late pre-Hispanic communities in Future analyses of macrobotanical, faunal re- this region. mains, and pollen samples recovered during this Preliminary evidence suggests that the settle- field season will shed light on the subsistence ment system, architectural layout, and possibly base and economic organization of the settle- even cultural affiliation varied greatly within the ment. drainage (Lightfoot, 1984; Van Keuren, 2001), Architectural data were recorded with a Nikon perhaps because of historical circumstances of DTS-330 Total Station and referenced to a prima- migration and trading ties between specific Silver ry site datum. The village includes two major Creek villages to contemporaneous towns outside room blocks referred to here as the southern and the drainage. The fieldwork at Shumway Ruin northern room blocks (Figure 2). Two rooms and thus attempted to answer several key questions: one extramural space were tested. Room 1 is How large is the village, and what public located in the southeastern section of the large architecture is present? Does the layout of the southern room block. The room lies within village resemble nearby Fourmile Ruin and a double row of rooms that defines the south- other late Silver Creek villages? When was the eastern boundary of the large Plaza 2. Room 3 is site occupied based on decorated pottery? Is located in the eastern edge of the northern room White Mountain Red Ware the predominant block, among rooms that lie along the southeast- painted pottery at the site, and how abundant ern edge of Plaza 1. A third room designated as is Roosevelt Red Ware and other decorated Room2 duringthe field seasonwas notexcavated wares (e.g., Jeddito Yellow Ware)? Is there beyond initial wall tracing. An additional 1- by 2- architectural or ceramic evidence for the settle- m unit was excavated in the southwestern half of ment of migrant groups at Shumway Ruin (as Plaza 1. there clearly is at Fourmile Ruin)? And finally, To ensure that data would be comparable with what is the basic occupational history of the recent work in the region, I adopted with slight village, based on the configuration and layout of modifications the field methods and recording room blocks and extramural spaces? In the systems utilized by the Silver Creek Archaeolog- sections that follow, I detail the excavation and ical Research Project (Van Dyke et ah, 1996). architectural evidence collected in the 2002 field Rooms were tested by excavating a single quad- season. rant and removing fill from discrete vertical units. Contributions in Science, Number 508 Van Keuren: Shumway Ruin 5 Figure 2 Architectural plan view ofShumway Ruin either full strata (e.g., upper room fill composed disturbed areas of the site. Fortunately, the of structural collapse) or 10-cm arbitrary levels visibility of room alignments, masonry wall (e.g., lower room fill just above floor). The plaza abutment and bonding patterns, and other archi- test unit was excavated in 10-cm arbitrary levels tectural evidence allow for both the mapping of down to sterile natural bedrock. All excavated fill architecture from the modern ground surface and was screened through one-quarter-inch mesh, and the analysis ofconstruction sequences. Inhis early artifacts were bagged according to material type. sketch map of the entire village. Hough (1903) Soil flotation samples were collected from all illustrated approximately 50 rooms surrounding excavated strata, and pollen samples were col- a small plaza (Figure 3). At its peak of occupa- lected from floors exposed in the room test units. tion, however, the village may have included as All test units were mapped, documented, photo- many as 350 to 400 masonry rooms, including graphed, and completely backfilled at the close of two-story rooms in the northern room block. The the season. destruction of a portion of the northern room block makes an exact room count estimate VILLAGE ARCHITECTURE AND LAYOUT impossible. Rooms throughout the site are gener- ally rectangular and range in area from 8 to 14 Although early explorers in the region reported square meters. Walls were constructed ofcareful- seeing partially standing walls at ancient ruins lyfaced and shaped sandstone blocks, presumably (e.g., Bandelier, 1892), most archaeological sites wet-laid with mortar and chinking stones. Al- have long since been reducedto rubble mounds of though adobe brick construction is prevalent at collapsed masonry, structural fill, and cultural nearby Fourmile Ruin (Johnson, 1992), there is refuse. At Shumway Ruin, there are no standing no evidence of this type of construction material or exposed walls outside of those in recently at Shumway Ruin. 6 Contributions in Science, Number 508 Van Keuren: Shumway Ruin later expanded with additional rows of rooms. The original plaza surface was likely the sand- stone bedrock that is partially exposed today. Plaza 2 was originally constructed around a nat- ural spring from which water flowed northwest- erly from the center ofthe plaza to the edge ofthe sandstone outcropping on which the village sits. Sometime in the past century, the channel was deepened and extended for use as an agricultural canal. In contrast to the slow accretion of rooms indicated in other parts of the pueblo, the presence of long contiguous room blocks here suggests that groups of individuals and house- holds from other parts ofthe Silver Creek area or adjacent regions arrived and joined an already established village. Although the temporal se- quence of room block expansion remains un- resolved, I presume that Plaza 2 was a late addition to the site. Future work is needed to clarify how many episodes of construction are reflected here and if the room block represents a late-constructed addition to the village. TEST EXCAVATIONS Figure 3 Hough’s 1901 sketch of Shumway Ruin (from Room 1 Hough, 1903:pl. 22; no scale in original) Room 1 is located on the eastern edge of the village along the edge ofPlaza 2, near the eastern opening (or gap) into the plaza (Figure 4). The Figure 2 shows that the village is composed of roomwas selected fortesting based onits location room blocks that delineate a small rectangular in what is presumed to be a late-constructed suite plaza (Plaza 1), a large irregular shaped plaza ofrooms that define Plaza 2. The modern ground (Plaza 2), and a possible great kiva (or small surface in this area ofthe village is largely devoid plaza). The latter ceremonial structure is evi- of vegetation because of grazing (Figure 5). With denced by a large rectangular depression, the size the exception of the buried or collapsed northern of which is consistent with great kivas in the and eastern walls, all other walls of this room region (Riggs, 2002:107). The sequence ofoverall were easy to discern from the modern ground village construction is similar to that of Fourmile surface. The room is approximately 3.4 m long Ruin. The original, or core, room blocks at and 2.7 m wide. The room was divided into Shumway Ruin appear to lie at the northern edge roughly equal quadrants. The southwestern quad- of the site, adjacent to the cliff edge that drops rant was excavated down to the original floor of some 10 m to the valley floor. These early rooms the room; the three remaining quadrantswere not delineated the possible great kiva and Plaza 1. A excavated. Strata 1 and 2 were wholly excavated room excavated in this area of the pueblo by the as discrete, separate vertical units; Stratum 3 was currentlandownerscontained one Pinedale Black- excavated in 10-cm arbitraryunits (conformingto on-white olla and three Pinedale Polychrome the upper and lower boundaries of the stratum). bowls, all possibly sitting on the lower floor of Excavations were referenced horizontally to three a two-story room. These vessel types date mappingpoints andverticallyto a single elevation between 1275 and 1325 (Mills, 1999) and datum (vertical datum 2002A, 96.87 m). Al- indicate that the northern room block was though surrounding areas of the village show perhaps occupied as early as the end of the evidence of recent pot-hunting activities, there is thirteenth century. The room also confirms that little or no disturbance within Room 1. the village was multitiered with at least two-story ROOM EILL. Three postoccupational strata rooms in this portion of the village. were defined in the fill of Room 1 (Figure 6). Sometime later in the history of the village, Here, I use “postoccupational” to denote depos- a larger plaza (Plaza 2) was defined with the its formed after the structure fell into disuse but addition of one or more contiguous room blocks before the abandonment of the entire settlement; at the southwestern and southeastern edges ofthe “postabandonment” refers to deposits formed village. Thisplazawas originallydefined byacore after the entire site was abandoned. Stratum 1, room block (withtwotothreerowsofrooms) and the uppermost stratum, was a thin (7 to 15 cm) Contributions in Science, Number 508 Van Keuren: Shumway Ruin 7 Figure 4 Plan view ofRoom 1 8 Contributions in Science, Number 508 Van Keuren: Shumway Ruin Figure 5 Photograph of Room 1 following excavation deposit of wind- or water-deposited alluvium a thin (1 to 2 mm) layer ofgray clay. The layer is mixed with minor amounts of organic material best preserved along the southern and eastern (roots). The layer was largely devoid of cultural walls and poorly preserved near the center of the material. Stratum 2 was a thicker layer (28 to room. A Fourmile Polychrome bowl rim sherd 46 cm) of collapsed masonry wall fall, which was recovered on the floor along the northern included large shaped sandstone blocks, smaller edge of the unit. Aside from one additional sherd chinking material, and red-brown clayish mortar. and four lithic flakes, no other artifacts were The top of this layer sloped downward from the present at floor contact. No floor features were southwestern to northeastern corners of the uncovered in the southwestern quadrant. quadrant and was thickest adjacent to the The excavation of the southwestern quadrant exposed southern and western walls. The layer confirmedwhatwas evident on the surface; that is, contained very little cultural material. Stratum 3 both the southern and the western walls faced was a thinner layer (10 to 15 cm) of structural a courtyard area prior to the construction of this collapse and contained fewer wall stones and room. Both are fully coursed, well-prepared exte- chinking material deposited above and directly rior facing masonry walls; the southern wall abuts on the cultural surface. The clay content and the westernwall. The existing heightofthese walls matrix of this layer suggest that it was deposited rangesbetween .35 and .40 m,andtheshallowness with the erosion of wall plaster and other ofroom fill and structural collapse suggestthat the construction materials of the surrounding walls. walls of Room 1 and surrounding rooms were not There was no clear evidence of roofing material full height at abandonment. Alternatively, wall in this stratum, nor was there evidence that the stones may have been removed from the wall for room was burned. The boundary between Stra- construction in other portions of the pueblo. The tum 3 and the cultural surface was clearly western and southern walls include three to four delineated in most of the quadrant, suggesting intact courses of masonry, with perhaps two to that sediment was deposited fairly quickly, pre- threecollapsedintotheroom.Thewallheightprior serving the original cultural surface. Like the to collapse was less than 1 m. The northern and stratum above it. Stratum 3 contained little eastern walls of the room are difficult to discern cultural material. from the modern ground surface. Based on CULTURAL SURFACES AND ARCHITEC- architectural evidence in Room 1, coupled with TURE. Surface 1 was a leveled floor capped with the bondand abutmentofthewallsofsurrounding Contributions in Science, Number 508 Van Keuren: Shumway Ruin 9 cm Figure 6 East excavation profile, southwestern quadrant, Room 1 rooms, it is clear that a single row of rooms and 2.6 m wide. The room was divided into originallydefinedthisedgeofPlaza2. Oneormore roughly equal quadrants, and the southeastern roomswerethenaddedinanotherrowastheroom quadrant was excavated down to the cultural block expanded (to the southeast). At some later surface. Strata 1 to 2 were wholly excavated as time, Room 1 andperhaps oneortwomorerooms discrete, separate vertical units; Strata 3 and 4 were added to this outer row. were excavated in arbitrary units conforming to SUMMARY. Room 1 was builtwithinarowof strata boundaries, each level not exceeding rooms that define the southeastern edge of Plaza 10 cm. Excavations were referenced horizontally 2, near the open corridor that leads from this to three mapping points and vertically to a single plaza to the exterior of the village. There is no elevation datum (vertical datum 2002C, evidence that the room was ever finished or that 97.20 m). Although Room 3 is not impacted by the existing structural collapse was associated pot-hunting activities, several adjacent rooms are witha full-heightmasonrywall and intactroof. In heavily disturbed. The modern ground surface in fact, the only indication that Room 1 was ever this area is largely devoid ofvegetation because of completed and occupied is a well-prepared and recent grazing (Figure 8). ROOM thinly plastered floor. It is also possible, however, FILL. Three postoccupational strata that the roofs and walls of Room 1 and and one mixed stratum were defined in the fill of surrounding rooms were partially disassembled Room 3 (Figure 9). Stratum 1, the uppermost at abandonment. The paucity ofcultural material stratum, was a thin (7 to 13 cm) deposit ofwind- in all fill was remarkable and suggests that all or water-deposited alluvium mixed with minor room fill was postabandonment collapse or amounts of organic material (roots), cultural deposited after the village was abandoned. There material, and a few large wall stones. Stratum 2 is no indication that ongoing cultural activities at was a thicker layer (49 to 56 cm) of collapsed the site influenced any ofthe deposition of Room masonry wall fall, composed of large shaped 1 fill. Further excavation is needed in the re- sandstone blocks, smaller chinking material, and mainder of the room to locate floor features and red-brown clayish mortar. The top of this layer establish room function. sloped downward from the southwestern to the northeastern corners of the quadrant and was Room 3 thickest adjacent to the exposed southern and eastern walls. Two segments of intact wall Room 3 is in the southeastern edge of the collapse were noted during the excavation of this northern room block adjacent to Plaza 1. It is unit (see Figure 9). The upper layer included partofa suite ofrooms built onthe southernedge a segment of the eastern wall with four to five of the plaza, all with deposits that appear much courses; below this, two to four courses of fallen more substantial and deeper than those in the stones from the southern wall were encountered. vicinity of Room 1. The room was selected These layers within Stratum 2 presumably repre- because of its location in the earlier-constructed sent two separate episodes of wall collapse. The or -occupied portion ofthe village. stratum was largely devoid of cultural material. The walls ofRoom 3 were easyto discern from Stratum 3 was a thinner layer (16 to 22 cm) of the surface (Figure 7). The room is 2.7 m long mixedstructuralcollapse. The layerincludedfewer 10 Contributions in Science, Number 508 Van Keuren: Shumway Ruin SHUMWAY RUIN D KEY: AZP:12:127 ASU wallstone ROOM 3 s sherd SOUTHEAST QUADRANT F lithicflake • mappingpoint 50 100 A elevationdatum Figure 7 Plan view ofRoom 3 wall stones and chinking material, mixed with roofing material. There was a higher density of a loosely consolidated clayish matrix. Unlike cultural fill in this layer with a distinct increase in Stratum 2, this layer included pockets of clay, trash at the transition to the nextstratum. Stratum charcoalflecks,andalooselyconsolidated,decayed 4 was a thin (6 to 10 cm) deposit composed of organic matrix. Although we did not encounter artifacts mixed with collapsed structural fill, all direct evidence of wooden beams or slats, the deposited directly on the prepared and level room overall fill isconsistentwithcollapsedand decayed floor. The layer contained a moderate density of