Geoarchaeological Evidence for Ritual Closure of a Kiva at Fourmile Ruin, Arizona

Journal of Archaeological Science 40 (2013) 615e625

Contents lists available at SciVerse ScienceDirect

Journal of Archaeological Science

journal homepage: http://www.elsevier.com/locate/jas

Geoarchaeological evidence for ritual closure of a kiva at Fourmile Ruin, Arizona

Scott Van Keuren a,*, Christopher I. Roos b a Department of Anthropology, University of Vermont, Williams Hall 509, 72 University Pl, Burlington, VT 05405, USA b Department of Anthropology, Southern Methodist University, Dallas, TX 75275, USA article info abstract

Article history: Geoarchaeological analyses, including soil micromorphology and chemistry, were used to characterize Received 4 May 2012 the deposition history of a fourteenth-century Pueblo ceremonial structure (or kiva) at Fourmile Ruin, Received in revised form east-central Arizona. These subterranean structures were often ritually “closed” by Pueblo societies 11 August 2012 through the burning of roofs and the deposition of special objects. These conspicuous markers are absent Accepted 22 August 2012 in the Fourmile Ruin kiva, but the composition of constituent materials and sedimentation processes nonetheless signal complex, highly-orchestrated closure activities. The study demonstrates the efﬁcacy Keywords: of geoarchaeological analyses of stratiﬁed deposits for modeling the subtle behaviors associated with the Geoarchaeology Stratigraphy ritual closure and abandonment of sacred places. Ó Ritual 2012 Elsevier Ltd. All rights reserved. Pueblo Micromorphology Abandonment

1. Introduction articulated animals, collections of unbroken ceramic vessels, or human remains that were placed on the floor at abandonment or Ancestral Pueblo communities in the US Southwest were in post-abandonment deposits. Outside of these highly visible fundamentally reorganized during a phase of settlement aggrega- traces of singularized treatment, the recognition of ritualized tion in the early Pueblo IV period (ca. A.D. 1275/1300e1400). closure is difficult. Fortunately, singularized treatment can leave Archaeologists who work on this time period often focus on the traces in the sediments and stratigraphy of archaeological social dimensions of large village establishment and growth. In deposits. Geoarchaeological analyses, including soil micromor- general, less attention has been paid to the elusive processes by phology and chemistry, provide evidence to identify traces of which communities fission and disperse. However, one productive deposit formation history including the composition of constit- area of scholarship focuses on the formation history of post- uent materials and sedimentation processes that can identify occupational deposits in ceremonial structures, or more specifi- singularized treatment in the absence of conspicuous markers. cally, evidence for the intentional disassembly and filling of these Such analyses have been successfully employed, for instance, to structures during “closure” events. In the case of structures that are study ritual behavior in the symbolically charged construction centrally located and shared by multiple households, such activities histories of earthen monuments (Sherwood and Kidder, 2011), but can potentially reveal village-wide social processes. have not yet been routinely employed to study the closure and Singularized deposits, a phrase we borrow from Walker (1995: abandonment of sacred places. 72; see Kopytoff, 1986), are associated with ritualized closures that In this paper we report on geoarchaeological analyses of strat- involve uncommon or specialized sequences of behaviors and ified deposits in a subterranean ceremonial room (or kiva) at objects. In most cases, singularized treatment has been recognized Fourmile Ruin, Arizona (Fig. 1). The village was the largest Ancestral through obtrusive stratigraphic signatures (e.g., roof burning), or Pueblo settlement in the Silver Creek drainage, and during its on the basis of distinctive associations of artifacts, plants, or heyday included two to three large plazas surrounded by single- animals. Examples of the latter include the interment of and multi-story masonry and adobe brick rooms. Kivas near the center of each of the two larger plazas were likely locations for non- public ceremonial practices of surrounding households and supra- household social units (e.g., lineages, clans). Our excavations * Corresponding author. Tel.: þ1 802 656 8546; fax: þ1 802 656 4406. fl E-mail addresses: [email protected] (S. Van Keuren), [email protected] exposed intact sediments burying the oor of one of these (C.I. Roos). plaza kivas. Although unusual assemblages of artifacts, faunal, or

2. Research background

The close of the thirteenth century A.D. marked the end of permanent residential occupation in large portions of the Ancestral Pueblo world (Adler, 1996). Broadly speaking, this demographic event reshaped the configuration, cultural composition, and perhaps even religious practices of communities throughout the northern US Southwest (Adams and Duff, 2004; Duff, 2002; Spielmann, 1998; Van Keuren and Glowacki, 2011). By the early 1300s, local populations and migrant groups coalesced into plaza- oriented villages, many of which were organized spatially into loose settlement clusters (Adams and Duff, 2004). Fourmile Ruin was among five villages in the Silver Creek cluster, east-central Arizona (Fig. 1)(Kaldahl et al., 2004; Mills, 1998). This site and others in the cluster are perhaps best known for their production of figurative-style pottery (Fourmile Polychrome) by the 1320s (Carlson, 1970: 71), which may have been associated with new religious beliefs that spread across the Pueblo world (Adams, 1991, 1994; Hays-Gilpin and LeBlanc, 2007).

2.1. Fourmile Ruin in context

Fourmile Ruin was first recorded by archaeologists more than a century ago (Fewkes, 1904; Spier, 1919). Jesse Walter Fewkes excavated portions of the site during a two-week expedition in Fig. 1. The location of Fourmile Ruin and other Pueblo IV period villages in the Silver 1896, but his brief report and unpublished notebook provide scant Creek area and adjacent regions of east-central Arizona. details on his fieldwork and the site was only recently mapped. Like other early Pueblo IV period villages, the village is composed of botanical remains were not encountered, our chemical and long linear room blocks constructed around expansive, partially fi micromorphological analyses of sediments lling the kiva provide enclosed plazas (Fig. 2). At least two subterranean ceremonial fi evidence of singular ritual practices that de ned the closure of this structures (or kivas) were constructed in the center of two main room. The inferred closure and subsequent treatment of this kiva plazas. During its heyday, the site included 250 to 300 ground floor offers insights into the wider depopulation of this village as well as rooms. Although northern portions of the site have now been the currency of geoarchaeology for inferring complex behaviors destroyed, both Spier (1919) and Fewkes (1904) note the presence through archaeological stratigraphy. of second- and third-storied rooms at the site. We suggest

Fig. 2. Architectural plan of Fourmile Ruin with the trench excavated in Kiva 3 (in Plaza 3) indicated. Wall lines indicated in black were clearly visible at the time of mapping. The inset photograph illustrates the hearth (round), deflector and ashbox (rectangular), and the flagstone paved floor and bench exposed in Kiva 3. S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625 617 a conservative estimate of 350e400 rooms, with one small plaza We use stratigraphic deposits within the fill of a plaza kiva at and two larger plazas. Based on the presence of well-dated Fourmile Ruin to illustrate how ritual closure can be inferred using ceramic types (Carlson, 1970; Mills and Herr, 1999), the site was stratigraphic evidence and geoarchaeological observations of the first established by the end of the 1200s, and then expanded constituent sediments. Our work builds on recent studies of ritual sometime in the early- to mid-1300s. Permanent residential deposits in the Ancestral Pueblo Southwest, namely work in the occupation by Pueblo peoples ended in the area during the final Middle Colorado River Valley to the north (Fig. 1). Here, Walker and decades of the 1300s (Mills, 1998). his colleagues (Adams and LaMotta, 2006; Walker, 1995; Walker Fourmile Ruin was subject to several episodes of systematic et al., 2000) have examined the ritual closure of structures at looting prior to its transfer from a private landowner to The thirteenth- and fourteenth-century villages in the Homol’ovi Archaeological Conservancy. Although the contents of rooms were region. Adams and LaMotta (2006:60e62) discuss the nature of cleaned out, the resulting architectural remains provided a wealth of “enriched deposits” in religious structures (including kivas) at information on wall construction sequences and materials. The Homol’ovi I and II as they relate to early forms of Katsina ceremo- earliest occupants of the site built suites of masonry rooms that nialism in the area. These deposits were created through the enclosed a small plaza in the northeast area of the village. The village placement of rare objects and other “inalienable possessions” (see expanded significantly with the arrival of migrants who constructed Mills, 2004) that effectively sealed off religious spaces to future use. long, contiguous room blocks that enclosed a much larger plaza Their research builds on Walker’s earlier work. In an essay that (Plaza 2). Wall bonds and abutments reveal that a portion of the late underscores the importance of inferring ritual in the past through room blocks were “ladder” constructions that involved organized “ceremonial trash,” Walker (1995, 1996) documents the ritual efforts (Cameron, 1999: 208), with rooms delineated by shorter deposits and closure of several Homo’ovi kivas. His discussion walls placed perpendicular between long parallel walls. The cannot be fully summarized here, but underscores the complex, majority of adobe brick walls are located in these areas (Johnson, and perhaps even temporally discontinuous deposition of objects, 1992). During the final phase of aggregation, additional rooms trash, and the eventual burning of several Homol’ovi kivas (1995: were built around a third plaza (Plaza 3). The mixed use of multiple 72). architectural traditions indicates that this was a “multi-cultural” The treatment of kivas in our region during the early Pueblo IV community. The best evidence of this was the expansive use of period has not been extensively documented, but one example of adobe bricks at the site (Johnson, 1992), a construction technology ritual closure was recorded at Bailey Ruin in the southern edge of that is not found at other Silver Creek sites, but was used at the Silver Creek cluster. Here, one small room along the main plaza contemporaneous villages in Middle Little Colorado River drainage (Room 2) revealed a complex depositional sequence (Mills et al., sites to the north (Gann, 1996). Despite the paucity of professional 1999:162e172). This room was possibly once part of a larger kiva excavations at the site, the architectural record of Fourmile Ruin before partitioning walls were built. The end of Room 2’s occupa- reveals a great deal about the processes of aggregation at the site. In tion was marked by the placement of two articulated rabbits, fact, one can infer the basic scenario of population resettlement, the a piece of turquoise, and quartz crystals on the floors of these size of arriving migrant groups, and the planning and construction adjacent rooms. Fill was intentionally added to cover the fl oor in processes involved in the major fourteenth-century expansion of both rooms. Sometime thereafter, but before the upper floor the village. Despite speculation about the role of factionalism collapsed, a hawk (Buteo sp.) was buried on a bed of decorated (Kaldahl et al., 2004), we know much less about the social processes sherds in the southeast corner of the room (Room 2). Another by which Fourmile Ruin and many other Pueblo IV period villages closure event possibly occurred at nearby Pinedale Ruin. At this were vacated in the late 1300s. The stratigraphic deposits of inten- contemporaneous site, Haury and Hargrave (1931: 32) partially tionally filled rooms and extramural areas can potentially reveal the excavated a plaza kiva that was intentionally burned, but their social processes of site abandonment. report does not make note of any unusual objects or deposits in this structure. 2.2. The closure of ceremonial rooms in the Pueblo Southwest The deposition of religious items or special objects during the closure of southwestern ceremonial rooms stretches deep into the The disuse and even closure of individual rooms, along with Pueblo past (Wilshusen, 1986). The late pre-Hispanic period the processes through which entire Pueblo towns were depopu- examples we cite demonstrate that the closure of kivas involved lated, left subtle traces in the archaeological record. In rare cases, explicit ritual activities that are reflected both in the objects that catastrophic events are evidenced by the intentional burning of are deposited as well as the composition and treatment of the suites of rooms (e.g., Point of Pines Pueblo; Lindsay, 1987). A structure fill. These “singularized deposits” were not necessarily handful of ethnographic (or ethno-historic) case-studies have created through concurrent events. Rather, they appear to be documented the ways in which whole Pueblo villages fissured orchestrated during distinct episodes. The behavioral relationships during episodes of social conflict and factionalism during the between multiple stratigraphic layers can thus be complicated historic period (Cameron, 1992; Levy, 1992). In most cases, (Walker, 2002). Discontinuous closure activities may be difficult to however, the circumstances and rate of depopulation of ancestral discern on the basis of field observations alone, particularly in the villages are exceedingly difficult to detect in the archaeological absence of sacred objects, singularized deposits, or other overt record of large Pueblo villages. This is not surprising given that markers in the stratigraphic record (e.g., roof burning). Perhaps the scale of population movement and out-migration in the more important, the ritual treatment of structures may go unno- Pueblo Southwest, normally involved households and other small ticed during archaeological excavation when these features are social groups (Bernardini, 2005; Cameron, 1995; Dean, 1970; Duff, absent. 1998; Ramenofsky et al., 2009). Given the complex nature of the archaeological record of these sites, and the unlikelihood of 2.3. Traces of ritual closure in micromorphology extensive future excavations, the scrutiny of stratigraphic deposits becomes critical to modeling village histories. Because plaza kivas The archaeological analysis of ritual activities emphasizes the are associated with supra-household activities necessary for presence of, and relationships among, specific artifacts, ecofacts, community religious life, their closure can serve as a proxy of and features. The sediments and soils in which the objects and broader room block abandonments within settlements. features are observed or recovered, often referred to as the 618 S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625

“matrix,” are key dimensions of the context for traditional recovered, the presence of cross-dated ceramic types in concert archaeological subjects but are rarely treated as “artifacts” in their with construction sequence data for surrounding roomblocks own right (Goldberg and Berna, 2010; Salisbury, 2012). The study of suggest the structure was built and occupied sometime between undisturbed soils, sediments, and archaeological materials in thin the middle- to late-fourteenth century. The paved floor and asso- section (i.e., archaeological soil micromorphology) provides ciated features are similar to a kiva (designated Kiva 1) excavated by a unique avenue for inferring human behaviors that may be entirely Fewkes (1904:137e139) at the site in 1896. Based on Fewkes’ invisible in the macroscopic record of artifacts, features and macro- published diagram, features in Kiva 1 included a slab-paved floor, stratigraphy (Courty et al., 1989; Goldberg and Macphail, 2006). raised southern bench (paved), slab-lined ash pit and adjacent Microfacies analyses (Courty, 2001) of undisturbed soils through hearth, and loom holes. A larger plaza kiva (Kiva 2) was also tested petrographic observation, description, and spectroscopic in 2008, but a 1 2 m test unit revealed approximately 1.8 m of measurements have yielded extremely high-resolution evidence redeposited looter’s spoil atop two displaced paving stones that about human behaviors within strata less than 1e2 cm thick (Berna originally formed the kiva floor. et al., 2007; Goldberg et al., 2009). Micromorphological analyses Five major stratigraphic units above the paved floor in Kiva 3 excel at answering a number of questions that pertain to the were described and mapped in the field. These strata included layers singularized life histories of sacred places: What are the sediments of contrasting materials that post-date the original ceremonial use and materials that make up the matrix? How are they related? How of the kiva. Six undisturbed monoliths (25 cm 8cm 8 cm) were did they get there? And how were they modified after deposition? carved into the profiles and removed with the aid of 25-cm long Distinguishing the cultural and natural formation processes segments of polyvinyl gutter downpiping that had been cut open on responsible for the formation and alteration of the deposits (see one side. Loose “bulk” sediment samples were collected in 5-cm Schiffer, 1996) is fundamental to parsing the traces of patterned continuous intervals adjacent to the monoliths. Monoliths were behaviors that characterize ritual (Hollenback, 2010; Walker, 1995, impregnated with a polyester resin in the field and cut into 25-cm 1996). monoliths and 5.5 cm 7 cm blocks for preparation of thin The selection of sediments for cultural deposition e if this can be sections. Three monoliths and paired sets of bulk samples were corroborated based on sedimentary structures e is likely to have collected from Stratigraphic Unit IV (see Section 4.4), one across the been a purposeful decision with consequences for deposit forma- boundary of Stratigraphic Units III and IV, and two were collected tion (Sherwood and Kidder, 2011). Subsequent alteration by phys- from the top of Unit I to the contact with the paved floor. Additional ical or chemical weathering can also provide clues to the rates at undisturbed blocks were collected opportunistically from excava- which deposit formation took place (Goldberg and Macphail, tion units across the site to characterize architectural material 2006). We echo the positions of Goldberg and Berna (2010) and including adobe bricks, construction mortar, and wall plaster. others (Salisbury, 2012; Wells, 2006) that archaeological sediments are virtually untapped archives of human behavior and should be 3.2. Analytical methods and laboratory work conceptualized as “artifacts.” In the investigation of ritual actions, earth science methods and soil micromorphology in particular Loose, bulk samples were dried and ground to pass through a 2- should play a more important role than it has played to date. The mm sieve before other analyses were conducted. All bulk samples Fourmile Ruin kiva demonstrates the utility of this approach. were analyzed using loss-on-ignition to estimate organic carbon and calcium carbonate content using standard methods. Organic 3. Methods carbon was estimated gravimetrically by combusting sediments in a Skuitt programmable electric kiln for 2 h at 360 C(Storer and 3.1. Archaeological excavations in Kiva 3 Sarquis, 2005). Carbonate content was estimated by re-burning the sediments at 900 C for 2 h (Machette, 1986). Extractable The stratigraphic data we present in this paper were based on phosphorus concentrations were estimated using molybdate samples taken during excavations at Fourmile Ruin in 2008. Kiva 3 colorimetry for the fraction of sediment smaller than 125 mm using was tested in 2008 and 2009 as part of a broader but limited set of a modified version of the Mehlich-II extraction originally described excavations at Fourmile Ruin and two other contemporaneous sites by Terry et al. (2000) and described in its modified form by Roos in the Silver Creek cluster (Pinedale and Shumway ruins). The and Nolan (2012). research project focused on the recovery of ceramic data, in part to A total of 21 thin sections were made from sediments collected clarify the social context of iconographic-style pottery production during the 2008 field season. Ten thin sections from stratigraphic in the fourteenth-century. However, limited excavations were also contexts in Kiva 3 were used in this analysis. Thin sections were designed to document the settlement history and chronology of scanned on a flatbed scanner and observed using plane polarized these three large aggregated villages, all of which were located on light (PPL) and crossed polarized light (XPL) at magnifications from private land and had not been extensively documented. Excavation 7 to 400 using stereozoom and petrographic microscopes. Thin units were used to investigate room, extramural, and trash deposits sections were described using standard terminology for constituent throughout the village, including Kiva 3 and rooms in several materials, porosity, fabric, and post-depositional alterations or nearby room blocks. “pedofeatures” (Courty et al., 1989; Stoops, 2003). Kiva 3 is indicated by a shallow depression in the center of the large eastern plaza. The subterranean room was initially tested with 4. Results a1 1 m excavation unit in 2008 that extended from the modern ground surface to the paved floor of the room. The micromor- 4.1. Stratigraphic Unit I phology samples reported here were collected from the east and south excavation profiles at the close of the field season. In 2009, Stratigraphic Unit I lies conformably on the flagstone floor of the this unit was reopened and an excavation trench was extended kiva. A few, flat-lying sandstone slabs lay within these sediments from the western edge of the 1 1 m unit to the structure’s eastern less than 2 cm above the floor. In the field, Unit I was identified as wall. Both test units exposed several characteristic floor features a heterogeneous deposit with dark reddish brown (Munsell color that are found in both Ancestral and historic-period Pueblo kivas designation of 5YR 5/4 dry, 5YR 3/4 moist) silty clays and clays, and (Smith, 1972). Although no dendrochronological samples were pale light greenish gray clay loams (10Y 8/1 dry) with lenses of S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625 619

Table 1 rounded aggregates of dark red, iron rich silt and mudstones occur Summary of soil data from 5 cm continuous samples for each stratum. commonly but not abundantly within 5 cm of the slab ﬂoor. Stratum Range Mean Standard Above the ﬂoor-lying sediments, Unit I materials are a hetero- deviation geneous mix of unaccomodated, subrounded clods of contrasting Stratum IV (N ¼ 19) microfabrics. Dark red, iron-stained siltstone and mudstone e e LOI360 organic matter (%) 1.6 3.1 2.0 0.4 aggregates are locally the most abundant constituent. These e e LOI900 carbonates (%) 6.4 9.7 8.3 0.9 unaccomodated aggregates occur with complex packing voids in Extractable P (mg/kg) 46.1e72.0 63.1 6.6 Stratum III (N ¼ 5) association with multi-lithic sands, calcareous clays and rarely with

LOI360 e organic matter (%) 1.5e2.3 1.8 0.3 charcoal. Other aggregates include pale green-brown, iron- e e LOI900 carbonates (%) 7.7 11.1 9.3 1.3 depleted clays, and red-orange, limpid sandy clays. Extractable P (mg/kg) 46.4e57.5 51.8 5.1 Unit I is capped by diffuse bundles of carbonates, sands, and Stratum I (N ¼ 12) charred plant tissues, including vascular plant stems and fragments LOI360 e organic matter (%) 1.0e1.5 1.3 0.1 LOI900 e carbonates (%) 8.5e13.4 10.4 1.6 of conifer wood charcoal. Bundles of speckled, etching micritic Extractable P (mg/kg) 0.0e13.5 4.3 5.0 carbonates appear to be weathered ashes in association with the All samples (N ¼ 36) charcoal as well as with microfaunal excrements, suggestive of e e LOI360 organic matter (%) 1.0 3.1 1.7 0.5 unburned plant tissues that are no longer preserved. LOI900 e carbonates (%) 6.4e13.4 9.1 1.5 Extractable P (mg/kg) 0.0e72.0 41.9 27.9 Unit I is continuous across the entire trench covering both the floor and the bench entirely. It is noteworthy that Unit I accumulated to approximately the height of the bench over the entire cobbles and dark gray-brown sandy clays (10YR 3/2). This deposit is exposed floor but also covers the bench more or less evenly. characterized by very low organic carbon content, relatively high Imbrications of gravel near the east wall of the kiva suggest that the calcium carbonate content, and extremely low phosphorus point of origin may have been near the walls. concentrations (see Table 1). Despite the heterogeneity of constituent sediments, organic carbon and phosphorus concentrations 4.2. Stratigraphic Unit II varied little from 130 to 155 cmbd. Phosphorus concentrations were slightly higher in the 15 cm above the floor but remained Stratigraphic Unit II consists primarily of large cobbles and shaped much lower than phosphorus concentrations for all other sedi- sandstone masonry intercalated with cobble-sized clods dark gray ments analyzed from the kiva. Carbonate content reached brown sandy clays (10YR 3/2 dry). It lies conformably, although maximum values between 130 and 140 cm (>12%), but otherwise irregularly on Unit I. The contact between the two units is sharp and remained consistent with carbonate concentrations in Stratigraphic often demarcated by horizontally oriented dark gray brown sandy Unit III and IV (see Fig. 3 and Table 1). clays capped with finely laminated white layers. These laminated The heterogeneous nature of Unit I sediments are readily white layers appear identical to wall “plaster” that is often found apparent in thin section (Fig. 4). Floor contact sediments are preserved adhering to a base coat of mortar on the facing of masonry heterogeneous mix of gravel sized (>10 mm) aggregates of sandy, or adobe walls. The plaster and clays, in turn, lie atop the calcareous calcareous clays. The clays in these aggregates vary in their color in charcoal and sands at the top of Unit I. No bulk chemical data is (PPL and XPL) from pale yellow-brown to dusty orange-brown. The available for Unit II, but thin-section observations of Sample FML-001 sands are quartz dominated multi-lithic sands with feldspars, micas, indicate that the dark gray sandy clays are nearly identical to the and micritic carbonate nodules common. Coarse (ca. 1e2 mm) material used to make adobe bricks in Room 4 at the site (FML-004). rounded and subangular charcoal is present (but uncommon) There are at least eight layers of sandy, speckled micritic calcite within aggregates and between aggregates, where it can be found in that was identified as wall plaster layers in the field (adhered to association with relatively unweathered bundles of carbonate ashes fragments of construction adobe). The interior walls of kivas were (see Fig. 4, right-center), from which the slightly elevated phos- repeatedly plastered (or “whitewashed”). These were occasionally phorus concentrations at the top (near the contact with Unit II) and decorated with painted murals in the Pueblo Southwest (Smith, bottom (floor contact) of Unit I probably derive. Pebble-sized, 1952, 1972), but we encountered no evidence of the latter in this

Fig. 3. Annotated south proﬁle of the trench exposed in Kiva 3. Bulk sediment chemistry measurements for organic carbon (% LOI360), carbonates (% LOI900), and phosphorus (Mehlich-II extractable P, mg/kg) for continuous samples in Units I, III, and IV are plotted by depth adjacent to the proﬁle. The locations from which soil thin sections were collected and analyzed are plotted to the right of the chemical data. 620 S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625

Fig. 4. Annotated photograph of the lower portion of the east profile of the 1 m 1 m test unit excavated in Kiva 3 during the 2008 field season (upper right). Stratigraphic Units I, II, III, and IV are indicated, as are the locations of thin sections from Unit I (thin sections AeD). Macroscans of thin sections are all 5 cm wide with key micromorphological features highlighted. All photomicrographs include a 200 mm wide scale bar. Note that despite the “reddened” appearance of Unit I in the field, the only evidence for burning comes in the form of unweathered ashes mixed into the heterogeneous fabric of floor-contact sediments (right, middle), and the weathered ashes, fine unrounded charred plant tissues (wood fragments and reeds) visible as a <1 cm thick band below the contact between Unit I and Unit III. The common microfaunal fecal pellets within the band of charcoal and weathered ashes suggests that some of the plant material was incompletely or unsuccessfully combusted. The red color of Unit I is imparted by the abundant rounded pebble-sized aggregates of Moenkopi formation silt and mudstones mixed with other pebbles and sediments. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) kiva. Although the plaster is calcareous, it does not appear to have (10YR 5/3 dry) with rare 2e5 cm diameter pieces of wood charcoal. been produced by rehydrating quicklime (CaO). Rather, it appears The sands are well-sorted and show hints of preserved bedding. The to have been made by mixing water, wood ash, clay, and sands. The deposits have relatively high organic carbon and carbonate content adobe, by contrast, is a well-mixed and well-sorted sandy clay with but are extremely rich in phosphorus (see Table 1). In thin section, abundant fine sand-sized charcoal fragments. The clays are brown Unit III sediments are well-sorted sub-angular to rounded sand and speckled in PPL and orange-brown but weakly birefringent in grains, sand-sized aggregates of dusty clays or calcareous clays, and XPL. The plaster adheres to adobe that underlies it, indicating that subangular to rounded charred plant tissues. Sand grains often have the wall plaster faces upwards rather than towards the paved floor. thin, continuous or discontinuous coats of calcareous clays or orange As we note later, this juxtaposition between plaster and adobe is dusty clays. Aggregates and void coats of dusty, microsparitic calcite the result of human agency rather than post-occupational struc- appear to be etching or dissolving. Fragments of graded beds and tural collapse. crusts are preserved in places. Subrounded, sand-sized bone fragments are present, but uncommon (Fig. 5E and F). 4.3. Stratigraphic Unit III On the west side of the trench, a vertical discontinuity separates Unit III from a redeposited block of slumped material from Strati- Stratigraphic Unit III is a laterally fining sedimentary unit lying graphic Unit IV (see Fig. 3). On the east end of the trench, the well- conformably over Unit II, although the lower boundary of Unit III is sorted sands grade to poorly sorted gravels and cobbles toward the diffuse. In the field, this was a 10e15 cm thick unit of brown sands kiva wall. S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625 621

Fig. 5. Annotated photograph of the east profile of the 1 m 1 m test unit excavated in Kiva 3 during the 2008 field season (upper left). Stratigraphic units I, II, III, IV, and Vb are indicated, as are the locations of thin sections from Unit IV (thin sections AeD) and Unit III (thin sections E & F). Macroscans of thin sections are all 5 cm wide with key micromorphological features highlighted. Photomicrographs of highlighted areas are displayed in PPL (left) and XPL (right) with a 500 mm wide scale bar. Note the evidence for both natural water-deposited sediment (in situ crusts) and eolian (well-sorted and rounded fine-grained aggregates, sand grains, and charcoal) in Unit III. Depositional features are lacking from Unit IV thin sections. Rather, the predominant characteristics of Unit IV thin sections are pedogenic in origin, albeit in cultural sediments (macroporosity, weak granular microstructure, carbonate nodules and void coats).

4.4. Stratigraphic Unit IV content also declines over this interval from 9% near the top of Unit IV to 7.2% at 75 cmbd (cm below datum), whereas phosphorus Stratigraphic Unit IV overlies Unit III across the trench, except concentrations remain stable at approximately 60e65 mg/kg from where both Unit IV and Unit III have been cut by a vertical 35 to 90 cmbd before a slight, stepwise decrease to concentrations disconformity, indicating that both Unit III and Unit IV were in place approximately 50 mg/kg in the lowest 10 cm of Unit IV. before the disconformity was created. Unit III was identified in the In thin section Unit IV is characterized by poorly sorted pebbly, field as a moderately consolidated dark grayish brown (10YR 4.5/2 sandy clay loams with abundant ceramics. Sands are quartz- dry, 10YR 2/2 moist) pebbly mud with abundant sherds, very rare dominated but multi-lithic and include subrounded and sub- small bones and bone fragments, and common chipped stone angular sand-sized bone fragments. The micromass occurs as sand- artifacts. The cultural fill is dominated by ceramic types that date sized to pebble-sized granules of dark gray-brown (PPL), speckled between the early- to late-1300s, and Pueblo IV period diagnostic clays with a weakly expressed granostriated b-fabric. Sand-grains types that post-date the 1370s were not encountered. Unit IV is often have thin orange (XPL), dusty grain coats. thickest towards the center of the kiva depression (>70 cm) and One of the most striking features of Unit IV is the change in thins towards the margins (ca. 35 cm). The boundary between Unit macro-porosity and micromass color (PPL) with depth. Near the top III and Unit IV is gradual and indistinct and defined primarily on the of Unit IV the thin sections have a coarse, vesicular-to-vughy basis of increasing artifacts, darkening color and loss of bedding macroporosity (Fig. 5A and B) whereas macropores are less structures from the top of Unit III into the lower portion of Unit IV. common with depth. Similarly, the micromass is darker near the Organic carbon content is highest in the upper 10 cm of Unit IV top of the stratigraphic unit and lighter near the bottom. The with a gradual decline to 75 cmbd, where values stabilize at about structure of the material changes from top to bottom as well. Near 1.8% or near the mean value for the profile (see Table 1). Carbonate the top, Unit IV has a weakly expressed medium granular structure 622 S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625 with abundant meso-scale vughs, chambers, and compound event appears to have taken place in a thin layer of fuels that packing voids. Lower in Unit IV, porosity is dominated by covered the initial burial deposits. A ready source for such fine compound packing voids. material would have been the reeds or other vegetal materials that made up the fine construction materials of the roof that was 4.5. Stratigraphic Unit Va/Vb removed before burial. Combustion of this layer was probably very low intensity (no in situ thermal alteration of Unit I sediments A sharp vertical discontinuity through Stratigraphic Units III and >500 C was observed in thin section; F. Berna, personal commu- IV terminates sub-horizontally at the top of Unit II in the western nication) and some of the plant tissues may have been completely 50e60 cm of the trench. Unit Va is distinctive for its well-bedded, unburned before they were consumed by soil microfauna (Fig. 6C). unconsolidated sands interbedded with blocks of material from The deposits immediately above the charcoal layer are Unit IV. Unit Va clearly post-dates the formation of Units III and IV composed entirely of dismantled architectural debris. The nearly but did not yield any modern material culture. Unit Vb lies continuous lenses of upward facing wall plaster at the base of Unit conformably over a pebble-line at the top of Unit IV through most of II indicates that this unit probably formed by purposeful and rapid the trench but unconformably over Unit Va, wherein it fills disassembly of portions of the exposed kiva walls and their a depression identified as a “shovel-pit” in the field. The sediments deposition in a continuous layer above the burned material of Unit Vb are very loose, brown (10YR 5/3 dry) sands with dark (Fig. 6D). The removed and deposited plaster, in turn, was buried green-black plastic fragments. A “Mamba” candy wrapper that by mixed wall construction material. That this purposeful dates to the 1980s was encountered in the Unit Vb “shovel-pit” fill destruction event happened rapidly is supported by the well- above the disconformity between Unit Va and Vb. preserved, continuous lens of charcoal and ashes at the top of Unit I and the well-preserved nature of the wall plaster sheets in 5. Discussion the middle of Unit II. It is unlikely that these fragile layers would have survived undisturbed by wind and water if it they had been During the initial excavation of the deposits from the Fourmile left exposed for very long. Ruin kiva, our field observations relied on common sense and long- Exposure to wind and water action is precisely what appears to standing assumptions about the formation of stratigraphic deposits follow the destruction of the remaining architecture (Fig. 6E). The to interpret the structure’s fill (Walker, 2002). Because this was well-sorted sands (and sand-sized fine-aggregates and charcoal) a kiva, both archaeological and ethnographic analogues indicated probably formed by reworking of plaza surfaces by wind and the that it there was a high probability that the lower fill was inten- infrequently preserved graded beds and crusts indicate periodic tionally deposited as a form of closure behavior. However, our reworking of these sediments by water. Sheet middens commonly excavations failed to reveal clear evidence of the overt ritualized accumulate in plaza spaces and the enriched organic carbon and behaviors that have been noted in other kiva closure events, phosphorus signals in Unit III probably originate from the parent including roof burning or the deposit of sacred or inalienable material (sheet middens) that were reworked by natural objects. Although only a portion of the overall structure was processes. Unit III, therefore, represents a hiatus in the cultural exposed, we did not encounter any unusual objects or other activities associated with Kiva 3 and perhaps the surrounding depositional treatments that have been documented in other plazas and room blocks. It is worth noting that some rooms Pueblo IV period examples in the region (Adams and LaMotta, surrounding Plaza 3 were dismantled at some time during the 2006; Walker, 1995; Walker et al., 2000). Additionally, we occupation of the village. assumed that most if not all of the upper fill was post-occupational, Cultural activity appears to have resumed with one final burial naturally reworked material, or perhaps even deposited as event. Unit IV is artifact rich but generally lacks the quantity or a consequence of recent looting activities. In short, our field quality of animal remains characteristic of primary trash deposits at observations did not indicate that the closure of Kiva 3 was the margins of villages or those that accumulate in abandoned complexly ritualized. rooms (e.g., Mills, et al., 1999:154e156). The very high phosphorus The geoarchaeological evidence from the sediments themselves, and organic carbon concentrations are indicative of midden however, suggests a much more complicated, prolonged history for deposits but the abundance of rounded, sand-sized bone fragments the closure of Kiva 3 (Fig. 6). The sediments in Unit I included and the lack of internal bedding features are unusual. The former a variety of contrasting materials, including sandy alluvial soils (see suggests that the refuse in Unit IV was physically weathered on the also Roos, 2008 for micromorphological descriptions of young, surface before it was deposited at Kiva 3, whereas the latter alluvial soils in the area), ashes, sandstone slabs, and abundant suggests that Unit IV probably formed very quickly from dumping iron-rich (and some iron-depleted) soft sedimentary rocks. These homogenous materials. We interpret this as evidence that Kiva 3 materials are probably local and originated in the Triassic Moen- was subject to a final episode of purposeful burial, this time with kopi formation that is the underlying bedrock through much of this refuse that was redeposited from an extramural trash deposit or reach of the Silver Creek drainage. Fourmile Ruin, however, sits atop midden that had been well-trampled and weathered before a mid-Pleistocene gravel terrace, and the Moenkopi siltstones and deposition in the kiva (possibly a midden located at the northern mudstones must have been acquired off-site, transported to the edge of the village). The use of midden debris for ritual closure and terrace, and deposited up to 40 cm thick across the floor and burial is not unprecedented in the Southwest (e.g., Montgomery, covering the bench. The hints of imbrication near the walls of the 1993: 161), and the last event in the singularized life history of kiva suggest that the roof had probably been dismantled before Kiva 3 appears to be one such instance of this type of depositional burial with Moenkopi mudstones, ashes, slabs, and alluvial soils behavior (Fig. 6F). began (Fig. 6B). The last episode of burial is particularly curious because the There is evidence of burning associated with the burial of the closure of the kiva (Unit I and II) appeared to be followed by an kiva in the form of the thin (1e2 cm thick), continuous lens of interval when the remnants of the kiva and its deposits were charred plant tissues and weathered ashes capping Unit I. However, exposed to natural elements for some period of time. Without the the mixed plant assemblage (vascular plants and sand-sized use of this kiva, the ceremonial activities of the households around angular wood fragments) do not indicate combustion of the roof, Plaza 3 may have moved to a surface room or may have ceased which was probably dismantled at this point. Rather, this burning altogether. Another possibility is that the room blocks surrounding S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625 623

Fig. 6. Schematic depiction of the events in the closure, decommissioning, and burial of Kiva 3 that are inferred on the basis of the geoarchaeological data: A) The functioning kiva ceased to be used. B) The roof was probably removed before the floor and the bench were covered with dry sediment aggregates that included material collected from Moenkopi formation outcrops, from well-developed alluvial soils nearby, and some unweathered ashes. C) A thin, but continuous bed of plant material may have covered the 20e40 cm of “natural” unburned sediment. This plant material was burned, but probably incompletely. D) Shortly after the incomplete combustion event, the upper portion of the walls were dismantled and deposited in the kiva. E) Phosphate and organic rich midden material within the plaza was redeposited by wind and water, accumulating in the low-points within the decommissioned kiva depression. F) The remaining kiva depression was purposely filled with secondary refuse, including abundant but heavily trampled artifacts, bones, and charcoal. G) Until the 20th century, the re-deposited midden material was pedogenically altered with additional organic matter enrichment nearer the surface, carbonate dissolution and re-precipitation below the surface within the root zone. H) At some point in the past century, looters excavated a shovel pit in the center of the kiva down to the layer of wall debris (Unit II) before abandoning the pit. I) Naturally-deposited sediments accumulated within this pit, apparently obscuring it completely before another looter excavated a shovel pit within the filled, previous shovel pit during the 1980s, at which point redeposited looted spoil with plastic and modern trash buried the old surface.

Plaza 3 were no longer permanently occupied after this event. excavated a shovel pit that ended when dense wall debris was However, the last-phase burial of the kiva indicates that portions of encountered (Fig. 6H). Enough time passed after the first episode of Fourmile Ruin were still occupied. The ultimate burial of the kiva vandalism to allow the early shovel pit to fill in completely with with significant amounts of re-worked midden deposits that naturally deposited sediments, and the kiva was targeted again by probably originated from households living near Plaza 3 may have looters during a period of intensive and sometimes mechanized marked a final closure of this sacred space before the complete looting of Fourmile Ruin in the late 1980s. Fortunately, this more depopulation of the village. recent disturbance was limited to the margins of the earlier shovel After the deposition of Unit IV, a long interval passed wherein pit (Fig. 6I). Kiva 2 and several suites of rooms around Plaza 2 were these sediments were exposed to pedogenic processes, including systematically looted during this period. the formation of coarse channels, chambers, and vughs through biological activity and organic matter accumulation in the A 6. Conclusion horizon near the surface, and the dissolution, leaching, trans- location, and precipitation of carbonates (probably from ashes The cessation of ceremonial room use at Ancestral Pueblo within the midden materials). The strong down-profile trend in villages in the US Southwest often involved ritualized activities and organic carbon and carbonates is indicative of pedogenic alteration the deposition of sacred objects or “ceremonial trash” (Walker, and the microfabric changes support this interpretation as well 1995). At times, however, it is impossible to infer the complex- (Fig. 6G). In more recent decades, or perhaps earlier in the twen- ities of these archaeological deposits based on field observations tieth century, the shallow surface depression attracted a looter who alone. Our analysis of a plaza kiva at Fourmile Ruin demonstrates 624 S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625 the efficacy of geoarchaeological analyses for modeling the social Adams, E.C., Duff, A.I., 2004. Settlement clusters and the Pueblo IV period. In: e processes that created these complex deposits. This kiva lacks overt Adams, E.C., Duff, A.I. (Eds.), The Protohistoric Pueblo World, A.D. 1275 1600. The University of Arizona Press, Tucson, pp. 3e16. features of ritual closure (burned roofing materials or sacred Adams, E.C., LaMotta, V.M., 2006. New perspectives on an ancient religion: Katsina objects on the floor). Nonetheless, pedogenic, sedimentary, and ritual and the archaeological record. In: VanPool, C.S., VanPool, T.L., chemical characteristics demonstrate that the structure was closed Phillips Jr., D.A. (Eds.), Religion in the Prehispanic Southwest. AltaMira Press, Lanham, Maryland, pp. 53e66. through a series of orchestrated ritual activities. The deposition of Adler, M.A., 1996. “The Great Period”: the Pueblo world during the Pueblo III period, off-site mudstones and the intentional placement and covering of A.D. 1150 to 1350. In: Adler, M.A. (Ed.), The Prehistoric Pueblo World, A.D. 1150e wall fragments (with intact layers of plaster) are particularly 1350. The University of Arizona Press, Tucson, pp. 1e10. Berna, F., Behar, A., Shahack-Gross, R., Berg, J., Boaretto, E., Gilboa, A., Sharon, I., intriguing behaviors that to our knowledge have not been docu- Shalev, S., Shilstein, S., Yahalom-Mack, N., 2007. Sediments exposed to high mented elsewhere. Perhaps most interesting, the evidence pre- temperatures: reconstructing pyrotechnological processes in Late Bronze and sented here indicates that this treatment likely occurred in Iron Age Strata at Tel Dor (Israel). Journal of Archaeological Science 34, 358e373. numerous episodes that may well have stretched over several years Bernardini, W., 2005. Hopi Oral Tradition and the Archaeology of Identity. The (or decades). Both the deposits in Kiva 3 and evidence for the University of Arizona Press, Tucson. deconstruction of nearby rooms (around Plaza 3) indicate that this Cameron, C.M., 1992. An analysis of residential patterns and the Oraibi split. Journal e portion of the village was depopulated sometime in mid- to late- of Anthropological Archaeology 11, 173 186. Cameron, C.M., 1995. Migration and the movement of southwestern peoples. fourteenth century. It is worth noting that this kiva, plaza, and Journal of Anthropological Archaeology 14, 104e124. surrounding room blocks were built during the last phase of village Cameron, C.M., 1999. Room size, organization of construction, and archaeological expansion, but were quite possibly the first to be abandoned. interpretation in the Puebloan southwest. Journal of Anthropological Archae- ology 18, 201e239. Rather than thinking of ritualized closures as discrete episodes, we Carlson, R.L., 1970. White Mountain Redware: a Pottery Tradition of East-Central echo the statements of Walker and others that such activities are Arizona and Western New Mexico. In: Anthropological Papers 19. University frequently part of drawn-out sequences of ritual behaviors. We also of Arizona Press, Tucson. Courty, M.-A., 2001. Microfacies analysis assisting archaeological stratigraphy. In: concur with others who point out that abandonment does not Goldberg, P., Holliday, V.T., Ferring, C.R. (Eds.), Earth Sciences and Archaeology. mean an end to cultural use or visitation (Nelson and Schachner, Kluwer, New York, pp. 205e239. 2002: 194; Zedeño, 1997: 94), and some of the depositional activ- Courty, M.-A., Goldberg, P., Macphail, R.I., 1989. Soils and Micromorphology in Archaeology. Cambridge University Press, Cambridge. ities we detected may well have occurred after residential occu- Dean, J.S., 1970. Aspects of Tsegi phase social organization: a trial reconstruction. In: pation of the entire village had ended. Although it is impossible to Longacre, W.A. (Ed.), Reconstructing Prehistoric Pueblo Societies. The University infer the exact reasons why Fourmile Ruin and the Silver Creek area of New Mexico Press, Albuquerque, pp. 140e174. Duff, A.I., 1998. The process of migration in the late prehistoric southwest. In: was depopulated by Pueblo groups, modeling the closure of this Spielmann, K.A. (Ed.), Migration and Reorganization: The Pueblo IV Period in kiva does provide insight into the ways that these demographic the American Southwest. Anthropological Research Papers 51. Arizona State events occurred. Our case-study thus illuminates some of the social University, Tempe, pp. 31e52. processes that encompassed broader depopulation events in the Duff, A.I., 2002. Western Pueblo Identities: Regional Interaction, Migration, and Transformation. The University of Arizona Press, Tucson. Southwest (see Nelson and Schachner, 2002). Fewkes, J.W., 1904. Two Summers’ Work in Pueblo Ruins, Twenty-second Annual In addition to the obtrusive evidence for burning or the burial of Report of the Bureau of American Ethnology, Part 1. Government Printing Office, e inalienable goods or rare biota, we suggest that understanding the Washington, D. C., pp. 1 196. Gann, D.W., 1996. The use of adobe architecture in the Homol’ovi region. In: origins of the sediments may play an important role in identifying Adams, E.C. (Ed.), River of Change: Prehistory of the Middle Little Colorado River the ritualized life histories of sacred places in the Ancestral Pueb- Valley. Archaeological Series 185. Arizona State Museum, Tucson, pp. 93e105. loan world and elsewhere. Our study joins those of other Goldberg, P., Berna, F., 2010. Micromorphology and context. Quaternary Interna- tional 214, 56e62. researchers who have renewed attention to the behavioral Goldberg, P., Macphail, R.I., 2006. Practical and Theoretical Geoarchaeology. Black- complexities of cultural beliefs about soils (Salisbury, 2012; Wells, well Science, Oxford. 2006) and the complexity of stratigraphic layers in archaeological Goldberg, P., Miller, C.E., Schiegl, S., Ligouis, B., Berna, F., Conard, N.J., Wadley, L., 2009. Bedding, hearths, and site maintenance in the middle stone age of Sibudu contexts (McAnany and Hodder, 2009; Sherwood and Kidder, cave, KwaZulu-Natal, South Africa. Archaeological and Anthropological Sciences 2011). 1, 95e122. Haury, E.W., Hargrave, L.L., 1931. Recently Dated Pueblo Ruins in Arizona. In: Smithsonian Miscellaneous Collections, vol. 82 (11). Government Printing Acknowledgments Of fice, Washington, D. C. Hays-Gilpin, K., LeBlanc, S.A., 2007. Sikyatki style in regional context. In: Schaafsma, P. (Ed.), New Perspectives on Pottery Mound Pueblo. University of We thank two anonymous reviewers for thoughtful comments New Mexico Press, Albuquerque, pp. 109e136. on an earlier draft of this manuscript. The Archaeological Conser- Hollenback, K.L., 2010. Ritual and religion. In: Schiffer, M.B. (Ed.), Behavioral Archaeology: Principles and Practice. Equinox Publishing, Ltd., London, vancy granted permission to conduct limited excavations at Four- pp. 156e163. mile Ruin in 2008 and 2009. Portions of those excavations and the Johnson, D.A., 1992. Adobe brick architecture and Salado ceramics at Fourmile Ruin. subsequent laboratory analyses for this paper were funded by In: Lange, R.C., Germick, S. (Eds.), Proceedings of the Second Salado Conference. fi Arizona Archaeology Society, Phoenix, pp. 131e138. a grant awarded to the rst author by the National Science Foun- Kaldahl, E.J., Van Keuren, S., Mills, B.J., 2004. Migration, factionalism, and the dation (BCS 0753156). Additional funding was provided by the trajectories of Pueblo IV period clusters in the Mogollon Rim region. In: Duff, A., University of Vermont and the Downey Family Fund for Faculty Adams, E.C. (Eds.), The Protohistoric Pueblo World: A.D. 1275e1600. The e Excellence from the Institute for the Study of Earth and Man at University of Arizona Press, Tucson, pp. 85 94. Kopytoff, I., 1986. The cultural biography of things. In: Appadurai, A. (Ed.), The Social Southern Methodist University. We thank Alejandro Figueroa for Life of Things: Commodities in Cultural Perspective. Cambridge University his work in the lab to generate LOI and P data while the second Press, Cambridge, pp. 64e91. fi “ ” author was at the University of South Florida. Levy, J., 1992. Orayvi Revisited: Social Strati cation in an Egalitarian Society. School of American Research Press, Sante Fe. Lindsay, A.J., 1987. Anasazi population movements to Southeastern Arizona. American Archeology 6, 190e198. References Machette, M., 1986. Calcium and magnesium carbonates. In: Singer, M.J., Janitsky, P. (Eds.), Field and Laboratory Procedures Used in a Soil Chronosequence Study. Adams, E.C., 1991. The Origin and Development of the Pueblo Katsina Cult. United States Government Printing Office, Washington D.C, pp. 30e33. University of Arizona Press, Tucson. McAnany, P.A., Hodder, I., 2009. Thinking about stratigraphic sequence in social Adams, E.C., 1994. The Katsina Cult: a western Pueblo perspective. In: Schaafsma, P. terms. Archaeological Dialogues 16, 1e22. (Ed.), Kachinas in the Pueblo World. University of New Mexico Press, Albu- Mills, B.J., 1998. Migration and Pueblo IV community reorganization in the Silver querque, pp. 35e46. Creek area, east-central Arizona. In: Spielmann, K.A. (Ed.), Migration and S. Van Keuren, C.I. Roos / Journal of Archaeological Science 40 (2013) 615e625 625

Reorganization: the Pueblo IV Period in the American Southwest. Anthropo- Smith, W., 1972. Prehistoric Kivas of Antelope Mesa, Northeastern Arizona: Reports logical Research Papers 51. Arizona State University, Tempe, pp. 65e80. of the Awatovi, Expedition, No. 9. In: Papers of the Peabody Museum of Mills, B.J., 2004. The establishment and defeat of hierarchy: inalienable possessions Archaeology and Ethnology, vol. 39 (1). Harvard University, Cambridge, and the history of collective prestige structures in the Pueblo Southwest. Massachusetts. American Anthropologist 106, 238e251. Spielmann, K.A. (Ed.), 1998. Migration and Reorganization: The Pueblo IV Period in Mills, B.J., Herr, S.H., 1999. Chronology of the Mogollon Rim region. In: the American Southwest. Anthropological Research Papers 51. Arizona State Mills, B.J., Herr, S.A., Van Keuren, S. (Eds.), Living on the Edge of the Rim: University, Tempe. Excavations and Analysis of the Silver Creek Archaeological Research Spier, L., 1919. Notes on Some Little Colorado Ruins. In: Anthropological Papers, vol. Project, 1993e1998. Archaeological Series 192. Arizona State Museum, 18, pt. 5. American Museum of Natural History, New York. Tucson, pp. 269e293. Stoops, G., 2003. Guidelines for Analysis and Description of Soil and Regolith Thin Mills, B.J., Van Keuren, S., Stinson, S.L., Graves III, W.M., Kaldahl, E.J., Newcomb, J.M., Sections. Soil Science Society of America, Madison. 1999. Excavations at Bailey Ruin. In: Mills, B.J., Herr, S.A., Van Keuren, S. (Eds.), Storer, D., Sarquis, M., 2005. The Chemistry of Soil Analysis. Terriﬁc Science Press, Living on the Edge of the Rim: Excavations and Analysis of the Silver Creek Middletown. Archaeological Research Project, 1993e1998. Archaeological Series 192. Arizona Terry, R.E., Hardin, P.J., Houston, S.D., Nelson, S.D., Jackson, M.W., Carr, J., Parnell, J., State Museum, Tucson, pp. 149e242. 2000. Quantitative phosphorus measurement: a ﬁeld test procedure for Montgomery, B.K., 1993. Ceramic analysis as a tool for discovering processes of archaeological site analysis at Piedras Negras, Guatemala. Geoarchaeology 15, Pueblo abandonment. In: Cameron, C.M., Tomka, S.A. (Eds.), Abandonment 151e166. of Settlements and Regions. Cambridge University Press, Cambridge, Van Keuren, S., Glowacki, D.M., 2011. Studying Ancestral Pueblo religion. In: pp. 157e164. Glowacki, D.M., Van Keuren, S. (Eds.), Religious Transformation in the Late Pre- Nelson, M.C., Schachner, G., 2002. Understanding abandonments in the North Hispanic Pueblo World. The University of Arizona Press, Tucson, pp. 1e22. American Southwest. Journal of Archaeological Research 10, 167e206. Walker, W.H., 1995. Ceremonial trash? In: Skibo, J.M., Walker, W.H., Nielsen, A.E. Ramenofsky, A.F., Neiman, F.D., Pierce, C.D., 2009. Measuring time, population, and (Eds.), Expanding Archaeology. University of Utah Press, Salt Lake City, pp. 67e residential mobility from the surface at San Marcos Pueblo, north central New 79. Mexico. American Antiquity 74, 505e530. Walker, W.H., 1996. Ritual deposits: another perspective. In: Adams, E.C. (Ed.), River Roos, C.I., 2008. Fire, climate, and socialeecological systems in the ancient South- of Change: Prehistory of the Middle Little Colorado River Valley. Archaeological west: alluvial geoarchaeology and applied historical ecology. PhD dissertation, Series 185. Arizona State Museum, Tucson, pp. 75e91. Department of Anthropology, University of Arizona, Tucson. Walker, W.H., 2002. Stratigraphy and practical reason. American Anthropologist Roos, C.I., Nolan, K.C., 2012. Phosphates, plowzones, and plazas: a minimally inva- 104, 159e 177. sive approach to settlement structure of plowed village sites. Journal of Walker, W.H., LaMotta, V.M., Adams, E.C., 2000. Katsinas and kiva abandonment at Archaeological Science 39, 23e32. Homol’ovi. In: Hegmon, M. (Ed.), The Archaeology of Regional Interaction: Salisbury, R.B., 2012. Engaging with soil, past and present. Journal of Material Religion, Warfare, and Exchange across the American Southwest and Beyond. Culture 17, 23e41. University Press of Colorado, Boulder, pp. 341e360. Schiffer, M.B., 1996. Formation Processes of the Archaeological Record. University of Wells, E.C., 2006. Cultural soilscapes. In: Frossard, E., Blum, W.E.H., Warkentin, B.P. Utah Press, Salt Lake City. (Eds.), Function of Soils for Human Societies and the Environment. Geological Sherwood, S.C., Kidder, T.R., 2011. The DaVincies of dirt: geoarchaeological Society, London, pp. 125e132. perspectives on American mound building in the Mississippi River Basin. Wilshusen, R.H., 1986. The relationship between abandonment mode and ritual use Journal of Anthropological Archaeology 30, 69e87. in Pueblo I Anasazi protokivas. Journal of Field Archaeology 13, 245e254. Smith, W., 1952. Kiva Mural Decorations at Awatovi and Kawaika-a, with a Survey of Zedeño, M.N., 1997. Landscapes, land use, and the history of territory formation: an Other Walls Paintings in the Pueblo Southwest. Harvard University, Cambridge, example from the Puebloan Southwest. Journal of Archaeological Method and Massachusetts. Theory 4, 67e103.