A multi-regional obsidian database for the Eastern Plains

Plains Anthropologist ISSN: 0032-0447 (Print) 2052-546X (Online) Journal homepage: http://www.tandfonline.com/loi/ypan20 A multi-regional obsidian database for the Eastern Plains Travis W. Jones, Robert J. Speakman, William T. Billeck & Robert J. Hoard To cite this article: Travis W. Jones, Robert J. Speakman, William T. Billeck & Robert J. Hoard (2018): A multi-regional obsidian database for the Eastern Plains, Plains Anthropologist, DOI: 10.1080/00320447.2018.1480860 To link to this article: https://doi.org/10.1080/00320447.2018.1480860 Published online: 12 Jul 2018. Submit your article to this journal Article views: 116 View Crossmark data Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=ypan20 plains anthropologist, 2018, 1–20 REPORT A multi-regional obsidian database for the Eastern Plains Travis W. Jones University of Georgia Robert J. Speakman University of Georgia William T. Billeck Smithsonian Institution Robert J. Hoard Kansas Historical Society Northern and Central Plains obsidian artifacts curated by the Smithsonian Institution’s National Museum of Natural History have received little attention by researchers working to understand the nature of long-distance trade, exchange, and interaction. We present the results of a chemical analysis of obsidian stone tools and debitage from these collections. Significant differences in patterns of obsidian use exist between the Northern and Central Plains. Shifts in obsidian use through time within the Central Plains may indicate larger socioeconomic shifts, while obsidian from Northern Plains assemblages suggests an antiquity to interaction networks at least as old as the first Plains Village sites in the region. By creating the first multi-regional obsidian database encompassing parts of the Northern and Central Plains, we expect that the data and our interpretations enhance discussions at the intersection of trade, exchange, and inter-group interaction in the Northern as well as Central Plains. keywords Central Plains, Northern Plains, obsidian sourcing, regional trade, exchange, and interaction © 2018 Plains Anthropological Society DOI 10.1080/00320447.2018.1480860 2 JONES ET AL. The Smithsonian Institution’s National Museum of Natural History (NMNH) currently curates substantial archaeological collections from both the Northern and Central Plains. A small percentage of these collections, obsidian stone tools, and debitage have received little attention from past researchers. Most obsidian artifacts found in Northern and Central Plains are particularly difficult to study because macroscopic sourcing can be problematic and, as an exotic material, obsidian typically comprises only a small portion of lithic assemblages in these regions (Ahler 1977:134; Hoard et al. 2008:219). Instead, many archaeologists choose to focus on more abundant local lithic materials such as Knife River flint, Tongue River silicified sediment, and others (e.g., Ahler 1977, 1975, 1979; Johnson 1984). But chemical sourcing technologies have the potential to give these collections new significance. Considering the lack of naturally occurring tool-grade obsidian in either the Central Plains or the eastern portions of the Northern Plains (i.e., Kansas, Nebraska, and North and South Dakota), sourcing obsidian artifacts to their geologic origins is useful for investigating past human behaviors related to long-distance interactions. Large-scale regional and multi-regional obsidian sourcing projects potentially inform specific archaeological questions at both the local and regional scales (e.g., Craig et al. 2007; Fitzhugh et al. 2011; Golitko et al. 2012; Golitko and Feinman 2015; Hoard et al. 2008; Ibáñez et al. 2015; Speakman et al. 2007). The utility of regional and multi-regional obsidian studies is their ability to uncover large-scale spatiotemporal trends related to inter-group economic and social practices, a perspective that enables researchers to produce valuable information about trade, exchange, and general interaction. For instance, the timing and temporality of variability within regional and multi-regional patterns of obsidian utilization can be one indication of broader changes in societal practices related to socioeconomic or political networks of interaction. Such shifts in network interactions may, in turn, be indicative of underlying and systemic changes at the societal level related to evolving social identities (e.g., Bauer and Agde-Davies 2010; Carter and Milić 2013; Donnelly 2015; Jenkins 2008; Keyser and Mitchell 2001; Mac Sweeney 2011:56; Spielmann 2004). To effectively address multi-scalar changes in networks of trade, exchange, and general inter-group interaction across the Plains, researchers must first establish macroregional databases through large-scale provenance studies incorporating source information from across entire regions (e.g., Ferguson and Skinner 2003; Hoard et al. 2008; Jones et al. 2016). Our study contributes over 100 sourced stone tools and debitage to a preexisting database, referred to here as the Central Plains Obsidian (CPO) dataset, created by Hoard and colleagues (Hoard et al. 2008:Table 1). It also expands the CPO dataset into the first multi-regional obsidian database for the Plains by incorporating artifacts from the Middle Missouri Trench subregion of the Northern Plains (Lehmer 1971). Because of a lack of discernable macroscopic variation in most obsidian artifacts recovered from both Central and Northern Plains contexts, elemental analysis is often the only way to identify and assign geologic sources. Elemental methods such as X-ray fluorescence spectrometry, inductively coupled plasma mass spectrometry (ICP-MS), and instrumental neutron activation analysis (INAA) go PLAINS ANTHROPOLOGIST 3 beyond macroscopic properties to define the chemical fingerprint (i.e., chemical composition) of individual artifacts. Comparing chemical fingerprints to known geologic units yields the identity of the material’s geologic origin and thus the location of a material’s original location on the landscape. Below, we present the results of a study utilizing portable X-ray fluorescence (pXRF) to source obsidian stone tools and debitage curated at NMNH from archaeological sites in the Northern and Central Plains (Supplemental Table 1). Materials and methods Using established obsidian analysis protocols (Shackley 2011; Speakman 2012; Speakman and Shackley 2013), 138 obsidian artifacts from 25 sites were analyzed in the NMNH dataset using a Bruker Tracer series III portable X-ray spectrometer (pXRF).1 One hundred and three samples are from 11 Kansas sites, 22 samples are from 9 South Dakota sites, 8 samples are from 4 North Dakota sites, and 5 samples originate from 1 site in Nebraska (Figure 1, Supplemental Tables 1 and 2). Other figure 1 Map of sites and geological obsidian sources discussed in this study. 1. Elliot (14CO2); 2. Larcom Haggard (14CO1); 3. (14MT502); 4. Pratt (14PT1); 5. Kermit Hays (14RC13) 6. Tobias (14RC8); 7. Malone (14RC5); 8. Thompson (14RC9); 9. Paul Thompson (14RC12); 10. Risston (14SC4); 11. Scott County pueblo (14SC1); 12. Signal Butte (25SF1); 13. 39FA83; 14. 39HU231; 15. Breeden (39ST15); 16. Dodd (39ST30); 17. Black Widow (39ST3); 18. Frank Risen (39PO8); 19. Hosterman (39PO7); 20. Molstad (39DW234); 21. Leavenworth (39CO9); 22. Boundary Mound Village (32SI1); 23. Hintz (32SN3); 24. Rock Village (32ME15); 25. 32MN1. 4 JONES ET AL. than a small number of artifacts lacking adequate provenience information, the NMNH dataset discussed here and in the Results section represents primarily Late Prehistoric period sites in the Northern and Central Plains (approximately 1000–150 BP). We chose pXRF to analyze for multiple reasons. First, pXRF benefits researchers and curation facilities as analysis times are relatively short and costs per sample are inexpensive compared to other methods such as INAA or ICP-MS. This allows for a greater volume of analyses within a shorter time-frame. Second, unlike ICP-MS and INAA, pXRF is non-destructive/non-invasive. By maintaining the sample’s physical integrity, the aesthetic, cultural, and scientific value for each artifact is completely preserved. Last, the high portability of pXRF devices easily accommodates on-site analysis, reducing the risk of lost or damaged artifacts by eliminating the need for institutional loans and transportation. Central Plains samples The Central Plains sample-set consists of 108 samples from 12 sites located in Kansas (n = 11) and Nebraska (n = 1). Most Central Plains samples (84%, n = 91) originate from seven Great Bend aspect (ancestral Wichita, about AD 1350–1700) sites in Kansas. After the Great Bend aspect, Dismal River sites make up 8% (n = 9) of the total (Supplemental Table 1). Four of the five artifacts from the Signal Butte site (25SF1) in Nebraska are associated with the site’s third cultural layer, containing both Upper Republican and Dismal River ceramics. Only one artifact, a projectile point, in the Central Plains NMNH sample-set predates the above-mentioned components. The projectile point (Figure 2) is from a Middle Ceramic period (AD 1000–1500) Pratt site (14PT1) site in Central Kansas (Hoard and Banks 2006:6). The remaining samples are from sites lacking adequate provenience information to securely place them into a single component. Northern plains samples The Northern Plains sample-set consists of 30 artifacts from one subregion of the Northern Plains – the Middle Missouri Trench located in North and South Dakota (Lehmer 1971). Of the 13 sites represented, 10 are Late Prehistoric period figure 2 Photograph of an obsidian projectile point from the Pratt site, 14PT1. PLAINS ANTHROPOLOGIST 5 Plains Village tradition sites (AD 900–1450) and one, Leavenworth (39CO9), is a historic Arikara village (Johnson 2007; Krause 1972) (Supplemental Table 1). Among the 10 Plains Village sites, Black Widow (39ST3), Breeden (39ST15), and Dodd (39ST30) exhibit multiple occupations. The obsidian tools and debitage from these three multicomponent sites lack adequate provenience information for assignment to specific components, hence we refer to them as only Plains Village tradition sites as opposed to the more specific Middle Missouri and Coalescent traditions found in this region. The remaining two Northern Plains sites in our study, 32MN11 and 39FA83, lack sufficient information to place them into any period more specific than Prehistoric. Results Each of the obsidian stone tools and debitage in the NMNH dataset matches one of seven unique chemical signatures based on their relative concentrations of the elements yttrium (Y) and zirconium (Zr) (see Figure 3). All seven chemical groups are from known geologic sources ranging from the northern Rocky Mountains to the Great Basin and Southwest. The northern Rocky Mountains (Bear Gulch and Obsidian Cliff) and Great Basin sources (Malad and Browns Bench) occur in present-day Idaho, Wyoming, and Montana (Hughes 1984:Table 3; Hughes and Nelson 1987:Table 1; Skinner and Thatcher 2009:Figure 2). The Southwestern figure 3 Biplot of yttrium versus zirconium for all samples analyzed in the NMNH dataset. 6 JONES ET AL. sources (Cerro Toledo, Valle Grande, and El Rechuelos rhyolite) occur within the Jemez Mountains source locality, a series of closely occurring but chemically distinct obsidian outcrops located in north-central New Mexico (Shackley 2005a:64–74, Shackley 2013). Artifacts chemically matching all but one of the known sources (Browns Bench) commonly occur in archaeological assemblages throughout the Plains. Central Plains Of 108 Central Plains artifacts, the majority (n = 103) chemically match sources within the Jemez Mountains locality (Figure 4). Approximately 66% (n = 72) of the stone tools and debitage match the Valle Grande source, making it the most common obsidian source-type found in the Central Plains portion of the NMNH dataset. Cerro Toledo is the next most commonly identified source-type with approximately 26% (n = 28) of all Central Plains artifacts chemically matching this source. The El Rechuelos source is the least common Jemez source-type with approximately 3% (n = 3) of the sample-set matching this source. Only five Central Plains NMNH artifacts are chemically identical to northern Rocky Mountains and Great Basin source localities (i.e., Obsidian Cliff and Malad). The oldest obsidian artifact in the study comes from the Pratt site (14PT1), a Middle-Ceramic period (AD 1000–1500) site located in eastern Kansas (Supplemental Table 1 and Figures 1 and 2). This artifact is a single projectile point made from Cerro Toledo obsidian. Previous obsidian sourcing studies also report a trend figure 4 Percentages of obsidian sources identified per site in the NMNH dataset. The thick gray line delineates the Central Plains from the Northern Plains. PLAINS ANTHROPOLOGIST 7 toward higher percentages of Cerro Toledo obsidian within assemblages dating to the Middle Ceramic period (e.g., Shackley 2005a, 2005b, 2006; see also Hoard et al. 2008: Table 1). The numerically largest group of tools and debitage in our study occur with Great Bend aspect sites (AD 1300–1670s) in central and southern Kansas (Hoard 2012). In contrast to the prevalence of Cerro Toledo obsidian among Middle Ceramic sites, artifacts originating from the Valle Grande source predominate (70%, n = 64) among Great Bend aspect sites (Figures 4 and 5). Considering site-level obsidian distributions, we observe trends demonstrating this regional pattern for the Tobias (14RC8), Paul Thompson (14RC12), and Malone (14RC5) sites. Although Valle Grande and Cerro Toledo obsidians are well-represented, Valle Grande predominates at all three sites. Only two artifacts, Tobias and Paul Thompson, match the El Rechuelos source. The latest Central Plains site represented in the NMNH sample-set is Scott County Pueblo (14SC1), a protohistoric site in east-central Kansas with evidence for a mixed Dismal River/Puebloan occupation from the mid-to-late AD 1600s (see Beck and Trabert 2014). All three Jemez sources are represented in Scott County Pueblo assemblage. When studying a separate collection of obsidian from the site, Hoard and colleagues (2008:225) identify a clear Southwestern connection in their analysis. Other work at the site corroborates such connection through the presence of figure 5 Pie charts depicting the percentage, by geologic source, of obsidian artifacts in the NMNH dataset. 8 JONES ET AL. diagnostic Southwestern material culture like pottery and architecture (for a detailed discussion see Scheiber 2006:144–146). Our analysis found similar results. However, whereas Hoard and colleagues report a strong preference for Cerro Toledo (56%) over Valle Grande (44%), the NMNH data demonstrate the opposite. Fifty-six percent of the NMNH Scott County Pueblo sample-set matches the Valle Grande source and 33% matches Cerro Toledo (Figure 4). When combining the NMNH data with Shackley’s (2005b) findings from another Dismal River site in Nebraska (Lovitt, 25CH1), the trend toward Valle Grande obsidian persists. Hoard and colleagues also note this possibility (2008:225). Five artifacts from the NMNH Central Plains collections match non-Jemez sources. These stone tools and debitage chemically match northern Rocky Mountains and Great Basin sources. A flake from Paul Thompson – a Great-Bend aspect site in Central Kansas – matches the Malad source in southeast Idaho. This is the only Central Plains NMNH artifact sourced to Malad. Although a handful of obsidian from Malad occurs in both Nebraska and Kansas assemblages from Great Bend aspect sites (e.g., Hoard et al. 2008; Hughes and Roper 1999), the existence of a relatively rare type of obsidian for the Central Plains is notable. The other five non-Jemez artifacts have chemical fingerprints matching the Obsidian Cliff geologic source. These samples are from two site assemblages: Signal Butte (25SF1) in west-central Nebraska and 14MT501 located in the southwestern corner of Kansas. Unfortunately, the provenience information attached to the single flake from 14MT501 is labeled “N/A (blowout)” so its temporal context could not be determined below the site level. Signal Butte is noteworthy as the only Nebraska site in this study. It is also the only site among both the Central and Northern Plains sample-sets to have near-equal amounts of Southwestern and Rocky Mountains/Great Basin obsidians (Figure 4). Four of the five artifacts analyzed from the Signal Butte site (a flake and three non-diagnostic point fragments) are associated with the site’s most recent cultural layer. This occupational layer, layer-III (SB III), consists of Upper Republican (AD 900–1400) and Dismal River (AD 1675–1750) ceramics as well as glass beads and copper (Forbis 1985; Strong 1933, 1935, 1940). Of these four artifacts, three chemically match the Obsidian Cliff source and one matches Valle Grande. Excavation records are unclear whether the fifth artifact, a flake fragment of Valle Grande obsidian, is associated with level II (SBII), an archaic occupational level, or level III (Strong 1935:228–234). Northern Plains Less work has been done to source obsidian artifacts from the Middle Missouri subregion of the Northern Plains (located within modern day North and South Dakota). Without evidence to the contrary, an assumption persists that obsidian found in North and South Dakota originates from only a few sources in the northern Rocky Mountains (e.g., Ahler and Toom 1993:228; Vehik and Baugh 1994). Though a handful of previously sourced artifacts support such an assumption (e.g., Ahler and Haas 1993:150–151), an adequate sample size was never utilized to validate this hypothesis. PLAINS ANTHROPOLOGIST 9 Contrary to the Central Plains samples, northern Rocky Mountains obsidians predominate the Northern Plains sample-set. Of the 30 samples analyzed from the Northern Plains, one piece of Southwestern obsidian is present. Most samples match Obsidian Cliff, Wyoming (63%, n = 19) and Bear Gulch, Idaho (23%, n = 7) source localities (Figure 4). The remaining three samples are from two Great Basin sources in southern Idaho (Malad and Browns Bench) and a Jemez Mountains source (Valle Grande). The Dodd (39ST30) and 39FA83 assemblages contain a projectile point and a scraper from the Malad source, Breeden (39ST15) contains a flake tentatively sourced to Browns Bench, and the Black Widow (39ST3) assemblage contains a core made of Valle Grande obsidian. Of the 13 sites represented among the Northern Plains sample-set, artifacts from five sites have sufficient provenience information to designate specific time periods. The earliest of these are two Plains Village tradition sites along the Missouri River in South Dakota. Molstad (39DW234) is an Extended Coalescent village site located in north-central South Dakota. Radiocarbon dates and a ceramic ordination reported in Johnson (2007:178–185) place this site between AD 1400 and 1500. The Molstad samples consist of three scrapers and a single point matching the Obsidian Cliff and Bear Gulch sources, respectively. The Hosterman site (39PO7) – a later Extended Coalescent village dated to between AD 1500 and 1550 (Johnson 2007:185–188) – shares a similar pattern of source use. Hosterman contains a projectile point and a scraper matching the Obsidian Cliff source and one point matching Bear Gulch. The later, post-contact sites of Leavenworth (39CO9), Rock Village (32ME15), and Hintz (32SN3) (see Johnson 2007:199–202; Wheeler 1963:167– 233) contain only obsidian sourced to the Obsidian Cliff locality. Obsidian artifacts associated with the remaining eight Northern Plains sites lack the necessary provenience or typological information to place them within specific phases. Of these remaining sites, six are associated with the Plains Village tradition. Artifacts from Black Widow, Dodd, and Breeden can be placed into one of two time periods. Black Widow is a multi-component site with village occupations associated with Extended Coalescent and post-contact components, occurring in the AD 1400s and 1700s, respectively (Billeck 2000; Johnson 2007:152–153). The diversity of obsidian source types at this site is higher than any other in the Northern Plains NMNH sample-set, with artifacts matching Obsidian Cliff, Malad, and Valle Grande. The Dodd site consists of at least two occupations (Johnson 2007:172– 193). The first, an Initial Middle Missouri component occupation occurs between AD 1100 and 1200 and the second is a late Extended Coalescent or Post-Contact horizon between AD 1650 and 1700. Only one artifact, a modified flake sourced to Malad, is associated with the site. Breeden’s occupational history, like Dodd’s, has a 700-year gap, with Initial Middle Missouri (AD 1000–1100) and later Talking Crow components (AD 1700–1750) (Johnson 2007:168, 197). The only Browns Bench obsidian identified in the study is a single flake from the Breeden site. Although studies have determined general periods of occupation at each of these sites, NMNH excavation records associated with each artifact analyzed here lack the provenience information necessary to place them into specific occupational components. 10 JONES ET AL. The remaining three Plains Village sites, 39HU231, Boundary Mound Village (32SI1), and Frank Risen (39PO8), lack both adequate excavation records and intra-site chronologies necessary to identify them as anything more than stone tools and debitage associated with the Plains Village tradition (Supplemental Table 1). Obsidian from all three sites correspond to the Obsidian Cliff and Bear Gulch source localities. Site 39HU231 contains a modified flake sourced to Obsidian Cliff and both Boundary Mound and Frank Risen contain a modified flake and point, respectively, from the Bear Gulch source locality. Finally, we identify three flakes from 32MN11 and one scraper from 39FA83 as general Prehistoric according to NMNH records (Supplemental Table 1). These artifacts, like those from the six village sites above, lack either adequate internal chronologies or excavation records to place them into specific occupational periods. Gunnerson (1960:257) suggests that 39FA83 may be a Dismal River occupation site (AD 1640-1750). The assemblage from 39FA83 contains the only other Northern Plains artifact in our study that matches the Malad source. The three flakes from 32MN11 chemically match the Bear Gulch source locality. Discussion Considering both the new NMNH dataset (reported above) and the preexisting CPO dataset compiled by Hoard and colleges (2008:Table 1), several discrete patterns are discernable regarding both inter- and intra-group interactions in the Northern and Central Plains. We adopt the general term interaction as opposed to trade or exchange as the socioeconomic mechanism facilitating the movement of obsidian. The terms trade and exchange can have specific anthropological meanings in the literature that are beyond the resolution and scope of our study (e.g., Baugh and Ericson 1994; Chartkoff 1989; Earle 1982; Fredrickson 2003; Hodder 1982). Conveyance is another term associated with trade and exchange, especially in reference to the movement of obsidian (e.g., Bamforth 2009; Hughes 2011; Jones et al. 2012, 2003). However, we do not employ the term here as the processes addressed by conveyance are better suited to discussing the interactions of mobile foraging populations. Conveyance may help to understand processes that moved obsidian to villages in the Middle Missouri Trench, but the term is too restrictive for discussing the interactions experienced by village societies and the mobile groups that most likely brought obsidian to them. By collapsing trade, exchange, and other possible mechanisms for the movement of obsidian within a general term like interaction, we can more thoroughly cover the entire spectrum of possible human behaviors associated with both the long-distance movement of materials by mobile groups and its consumption by village populations. We do not deny the importance of specifically addressing trade and exchange as they are important to any social, political, and/or economic system (e.g., Baugh and Ericson 1994; Dillian and White 2010; Fry 1980; Hughes 2011; Renfrew 1977; Vehik 2002). However, we do not distinguish between trade and exchange mechanisms and others like direct procurement or even theft and trophytaking, as the current data do not allow for such fine-grained differentiation. PLAINS ANTHROPOLOGIST 11 Central Plains The NMNH dataset presents a clear association between Southwestern obsidians and specific groups living in central and western Kansas. This pattern of Southwestern obsidian usage persists from the Middle Ceramic to the Historic period. However, each locality and period within the region displays unique trends in the distribution and frequency of obsidian source-types. Many of the results and observations presented here also parallel those noted in the CPO dataset (Hoard et al. 2008; Hoard and Ferguson 2011). Among the trends we identify in both the NMNH and CPO datasets, the most evident are the predominance of Jemez obsidian usage in central and western Kansas beginning as early as the Middle Ceramic period and the subsequent shifts in the distribution and frequency of specific Jemez obsidians through time. During the Middle Ceramic period, obsidian procurement shifts among groups in the central and southwestern areas of Kansas. Whereas obsidian tools from virtually all pre-Middle Ceramic contexts originate from northern Rocky Mountains and Great Basin sources, obsidian occurring during and after the Middle Ceramic in the western and central portions of the region tend to originate from Jemez sources (Supplemental Table 1 and Hoard et al. 2008:Table 1). This trend toward Jemez obsidian usage contrasts with the eastern and northern expanses of the region where obsidian distribution patterns remain similar to previous periods. Although the NMNH dataset contains only one artifact from a Pre-Great Bend aspect assemblage (a projectile point made of Cerro Toledo obsidian from the Pratt site, Supplemental Table 1), its presence along with other pre-Great Bend aspect obsidian in the CPO dataset corroborates Hoard and colleagues’ assertion that a shift toward increased interactions between western Central Plains peoples and groups to the southwest begins at least as early as the Middle Ceramic period (Hoard et al. 2008:226). Such a shift in obsidian usage could indicate the migration of groups from the Southwest but this probably is not the case (see Vehik 1976). Instead, we suggest this pattern indicates the divergence of interaction networks between eastern and western groups within Kansas as western groups began to reorient interactions in a new direction and develop stronger ties between pre-Great Bend aspect groups and groups to the south and west. Specifically, which groups pre-Great Bend peoples interacted with in the south and west remains unclear. It could not be determined if artifacts associated with pre-Great Bend aspect sites were procured directly from the Jemez Mountains or farther south as both Cerro Toledo and El Rechuelos sources erode into the Rio Grande and can be carried as far downriver as Mexico (Shackley 2005c, 2013). Though the general trend in Jemez obsidian use among central and western groups in Kansas continues into later periods, we identify a clear shift toward a specific Jemez source, Valle Grande obsidian, occurring within Late Ceramic assemblages. Considering the geologically restricted nature of Valle Grande obsidians – the Rio Grande does not carry Valle Grande obsidian down river (Shackley 2005c, 2013), the shift toward Valle Grande obsidian is interpreted by others (Hoard et al. 2008:226) as evidence for early development of interactions between some Central Plains villagers and Southwestern Puebloan groups that 12 JONES ET AL. eventually become the formal trade and exchange relationships identified at later Dismal River sites (Gunnerson 1968; Scheiber 2006:144–146). Such early interactions may have laid the ground work for latter migrations by Puebloan groups. As indicated in both the NMNH and CPO datasets, the trend toward Valle Grande obsidian continues into the Historic period at the Dismal River site, Scott County Pueblo in west-central Kansas. This shift in obsidian source-types could also indicate a narrowing of interaction with other groups to the south and west and a greater focus on groups with access the Valle Grande obsidian. Northern Plains Due to a lack of obsidian sourcing studies within the Middle Missouri and adjacent subregions of the Northern Plains, fewer obsidian artifacts are present in the literature (but see Baugh and Nelson 1988). Still, we can forward some general interpretations. For example, prior to our study, some researchers assumed that obsidian found within village contexts in the region originated exclusively from northern Rocky Mountains sources (e.g., Vehik and Baugh 1994:264). Our data show that this is not always the case. In some instances, people accessed obsidian from sources as far afield as Browns Bench in the Great Basin and Valle Grande in the Southwest. Another general observation we make using the current data is the predominance of northern Rocky Mountains sources across all periods represented in the NMNH dataset (Figure 4). Though the distribution and frequency of source-types shift over time, obsidian artifacts from Middle Missouri contexts are almost exclusively from Obsidian Cliff and Bear Gulch. The presence and persistence of Obsidian Cliff and Bear Gulch source-types suggests that long-distance interaction between Plains Village and nomadic groups directly west of the Middle Missouri Trench persisted for at least a millennium. Long-term, west-to-east trade and exchange are not a new focus among researchers in the Middle Missouri subregion. Multiple studies discuss the possibility of Late Prehistoric period interaction networks between Plains Villages and western nomadic groups in Montana and Wyoming (e.g., Johnson 1984; Spielmann 2014:889–893; Vehik and Baugh 1994). A long history of obsidian consumption from Rocky Mountains sources, specifically Obsidian Cliff and Bear Gulch, reaches as far east as Illinois and Ohio (Griffin et al. 1969). However, the distribution of specific source-types is not uniform across the region. The use of Browns Bench and Malad obsidian at the Breeden and Dodd sites, respectively, is evidence for a lack of uniformity. We cannot currently place artifacts from the two sites into a single occupation; however, we can still use them to form a set of hypotheses for future work to test. If artifacts Breeden and Dodd are associated with earlier Initial Middle Missouri occupations (AD 1000–1300) (Johnson 2007:168–175) they may evidence differential sociopolitical networks between early Initial Middle Missouri and later Initial Coalescent and Extended Coalescent villagers. This would certainly support other non-obsidian lithic studies in the region that suggest Middle Missouri and Coalescent groups had differential access to, or preferences for, non-local lithic materials (Ahler 1977; Johnson 1984). Conversely, if the obsidian artifacts from Breeden and Dodd belong to later protohistoric occupations, then it may be that Post-contact communities in South PLAINS ANTHROPOLOGIST 13 Dakota accessed different networks of interaction than that of northern villages such as Leavenworth, Rock Village, and Hintz. The differential procurement and consumption of obsidian between the two sub-regions would be no surprise. By the Post-contact period the proto-Arikara, Mandan, and Hidatsa were already well established in South and North Dakota (Johnson 2007:191). Future work with NMNH archival materials may be able to answer this question. A multi-regional perspective When looking at the NMNH and CPO datasets, patterns of obsidian use common to both the Central and Northern Plains become apparent. First, although the frequency of individual source-types fluctuates through time and across space, as noted above, the range of dominant source-types within each region remains relatively constant in each region (with the exception of the shift in usage during the Middle Ceramic period in Kansas discussed above). Furthermore, the dominant obsidian source-types found in Kansas and the Middle Missouri originate from major obsidian sources directly west of each region (Figures 1 and 4) (for similar observations see Gregg et al. 1996; Vehik and Baugh 1994). For example, while the frequencies of individual source-types fluctuated between village sites in the Middle Missouri, the predominance of northern Rocky Mountains sources remained constant throughout the region. Likewise, while the relative frequencies of Valle Grande and Cerro Toledo fluctuated through time among sites in Kansas, the overall predominance of Jemez obsidians remained constant after the Middle Ceramic period. This pattern of west-to-east obsidian movement suggests that while localized, interpersonal and inter-group interactions related to obsidian procurement were somewhat fluid between groups and generations, regionally these interactions tended to be more stable and focused in a specific direction. Interestingly, this pattern of obsidian utilization remains relatively stable through multiple technological and demographic shifts within each region (Hoard and Banks 2006; Lehmer 1971), suggesting one or more features common to both may have contributed to structuring the directionality of obsidian movement. We argue this commonality is the occurrence of major drainages that would have contributed to the directionality, frequency, and abundance of specific obsidian source-types within each region. Most sites in this study are located along major easterly flowing rivers and their tributaries. The Arkansas River and its tributaries in Kansas cuts through the center of the state. In the Middle Missouri region, the Missouri River and its numerous tributaries crosscut both North and South Dakota. These drainages likely facilitated intra-drainage interactions that extended westward toward the northern Rocky Mountains and Southwest. We are not suggesting groups in the Northern and Central Plains acquired obsidian through the same means or that individuals actively chose obsidians based solely on their distance from sources or ease of access. We are merely stating that long-standing relationships between social groups participating in networks of interaction across the Plains and the subsequent availability and variety of obsidian would have been at least partially conditioned by local geography. Other studies also discuss the importance of river valleys and drainages in structuring group 14 JONES ET AL. interactions (e.g., Anderson and Hanson 1988). Given the temporal stability of dominant source-types within Kansas and the Middle Missouri, the relative ease of movement within easterly flowing drainages, and the association of these drainages with heavily utilized obsidian sources, it is reasonable to suggest that groups in both regions were accessing obsidian through intra-drainage networks of interaction. The second pattern we see when reviewing the NMNH and CPO datasets is that the percentages of source-types in both the Northern and Central Plains mirror one another (Figure 5). Artifacts from Central Plains contexts originate primarily from Southwestern sources, specifically Valle Grande (66%) and Cerro Toledo (26%), for all periods with only a small percentage (5%) of obsidian originating from the northern Rocky Mountains and Great Basin. Conversely, these same source-types, specifically Obsidian Cliff (63%) and Bear Gulch (23%), are the most common source-types in the Northern Plains, with only one piece of Jemez obsidian occurring. Interestingly, Signal Butte, located near the boundary of these two regions in west-central Nebraska, has near equal amounts of Obsidian Cliff (60%) and Valle Grande (40%) obsidian (Figure 4). This multi-regional pattern in obsidian use may represent a dynamic narrative for social interactions occurring at this boundary. As stated in the results section above, stone tools and debitage from the Late Prehistoric to Protohistoric occupational layer (SBIII) at Signal Butte cannot be securely placed within either Central Plains tradition (Upper Republican) or Dismal River contexts. Two possibilities can explain the presence of both northern Rocky Mountains and Southwestern obsidians in equal amounts. First, Upper Republican and Dismal River groups at Signal Butte may have obtained obsidian through distinctly different networks of interaction – one oriented west and the other southwest. Although this may have been the case another scenario is equally likely. Obsidian artifacts found at Signal Butte cumulatively represent practices indicative of both Upper Republican and Dismal River occupations. The CPO and NMNH datasets combined with other provenance studies from Upper Republican and Dismal River sites in the High Plains demonstrate that obsidian and other materials from the northern Rocky Mountains, Great Basin, and Southwest occur within both Central Plains tradition and Dismal River assemblages, often simultaneously within the same site (see Scheiber 2006 for an overview). The presence of western and southwestern materials at Upper Republican and Dismal River sites indicates both groups participated in interaction networks oriented in similar directions, even if we cannot identify the specific occupation or occupations associated with these obsidian artifacts. If the obsidian from Signal Butte are in fact representative of both occupations, the inability to place them into specific occupational periods is not necessarily problematic. Instead of focusing on a particularistic culture historical narrative (i.e., attempting to identify specifically who was exchanging what, and with whom), such a scenario indicates a long history of access to broad, diversified networks of interaction by the multiple occupants of Signal Butte. Perhaps by occupying a geographic position between the Northern and Central Plains during the Late Prehistoric and Protohistoric periods, the later inhabitants of Signal Butte, whether associated with Upper Republican or Dismal River practices, accessed a greater diversity of socioeconomic relationships associated with obsidian than groups to PLAINS ANTHROPOLOGIST 15 the north and south. Often, areas positioned at the edges of physiographic, political, or cultural boundaries are simultaneously economic and social centers (sensu Marquardt and Crumley 1987:8–9). Certainly, a relatively greater variety in obsidian usage at Signal Butte supports such a scenario. Conclusion The goals of this article have been to: (1) present new obsidian provenance information generated from Northern and Central Plains artifacts curated by the Smithsonian NMNH; (2) integrate the new data into a multi-regional database for obsidian in the eastern Plains, and; (3) provide limited interpretations based on both new and extant data contained in the database. The data generated from the NMNH samples can be found in Supplemental Table 2. This new data will be combined with the data presented by Hoard et al. (2008:Table 1) and it is our hope to add to this database through future obsidian provenance studies and to collaborate with other researchers in order to bring extant obsidian provenance data together into a unified, accessible multi-regional database. As for our interpretations, the evidence from both the NMNH dataset contained in Supplemental Table 1 and the CPO dataset presented in Hoard et al. (2008:Table 1) suggests Late Prehistoric and Protohistoric groups in both Kansas and the Middle Missouri region engaged in far-reaching networks of interpersonal and intergroup relationships that were at least partially mediated through the movement of obsidian across the landscape. At broader spatiotemporal scales, interactions associated with obsidian from specific regions (e.g., the northern Rocky Mountains or Southwest) appear to have persisted for long periods of time while access to, or preferences for, specific source-types from each region fluctuated across time and space. Clearly, additional work is needed to better understand the multiple spatiotemporal scales at which interactions occurred and the socioeconomic mechanisms that drove them in the eastern Plains. Specifically, Nebraska in the Central Plains and the entire Northern Plains datasets require a greater number of sites with adequate provenience data and chronological control. Continuing to add to regional datasets and continually reinterpreting trends in the movement of obsidian is a productive avenue for advancing research focused on the long-distance movement of materials and the socioeconomic practices associated with them. Acknowledgments The authors would like to thank K.C. Jones and Jennifer Birch for their help and thoughts prior to submitting this manuscript for review as well as the two reviewers who gave substantive comments that only served to improve the clarity of this article. We also want to thank the Smithsonian Institution for allowing us to analyze their collections. Travis Jones would like to thank the National Science Foundation as this material is based upon work supported by the National Science Foundation Graduate Research Fellowship Program under Grant No. (049347-06). Any opinions, findings, and conclusions or recommendations 16 JONES ET AL. expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. Note 1 A Bruker Tracer Series III pXRF was utilized for this analysis. The instrument’s X-ray tube consists of a rhodium (Rh) target and a beryllium (Be) create their own calibrations. The benefit of this is that the user can create a more accurate calibration based on the specific properties of the window. The detector is a silicon drift detector (SDD) with a resolution of 145 eV. Samples were material he or she wishes to analyze. The calibration used in this study measures elements in analyzed for 100 seconds each at 40 kV and 30 μA. Bruker’s primary green filter (0.006’ Cu, the mid-Z range of ionization energies on the periodic table. This range of elements have been .001’ Ti, .012’ Al) was utilized to reduce background radiation within the mid-Z elemental shown to be the most important in obsidian sourcing studies and included manganese (Mn), iron range to better quantify trace elements, the most valuable for sourcing obsidian. Prior to analysis, samples were examined for flat, smooth surfaces, (Fe), zinc (Zn), gallium (Ga), rubidium (Rb), strontium (Sr), yttrium (Y), zirconium (Zr), niobium (Nb), and thorium (Th) (see Speakman which are most suitable for analyses because the low angle and lack of surface variation produces 2012; Speakman and Shackley 2013 for further discussion). A set of 40 obsidian standards com- the least amount of surface scatter upon contact with an X-ray beam. Sample size also was con- missioned from the University of Missouri Research Reactor (MURR) was used to create sidered, as only samples with the minimum surface area and thickness will produce accurate this matrix-matched calibration (Speakman 2012). Elemental data were tabulated in Excel results (Shackley 2011:52–57). All samples had an acceptable surface and/or minimum size suitable for analysis. Though all pXRF instruments and chemical groupings created based on elemental concentrations. Chemical groups were identified using bivariate plots of measured elemental come with preset calibrations that allow operators to analyze materials, some instruments, like the concentrations (Figure 3). These groups were then compared to known elemental signatures instrument used in this study, allow users to for obsidian sources (cited in the text). Supplemental data Supplemental data for this article can be accessed at https://doi.org/10.1080/ 00320447.2018.1458274 ORCID Travis W. Jones Robert J. Hoard http://orcid.org/0000-0001-8237-1726 http://orcid.org/0000-0002-9577-5321 References cited Ahler, Stanley A. (1975) Pattern and Variety in Extended Coalescent Lithic Technology. PhD dissertation, Department of Anthropology, University of Missouri, Columbia. Ahler, Stanley A. (1977) Lithic Resource Utilization Patterns in the Middle Missouri Subarea. Plains Anthropologist 78 (Part 2):132–150. PLAINS ANTHROPOLOGIST 17 Ahler, Stanley A. (1979) Functional Analysis of Nonobsidian Chipped Stone Artifacts: Terms, Variables, and Quantification. In Lithic Use-Wear Analysis, edited by B. Hayden, pp. 301–328. Academic Press, New York. Ahler, Stanley A. and Herbert Haas (1993) The KNRI Phase I Chronometric Subprogram. In The Phase I Archeological Research Program for the Knife River Indian Villages National Historic Site, Part I: Objectives, Methods, and Summaries of Baseline Studies, edited by Thomas. D. Thiessen, pp. 115–161. Occasional Studies in Anthropology No. 27. US Department of the Interior, National Park Service, Midwest Archaeological Center, Lincoln. Ahler, Stanley A. and Denis L. Toom (1993) KNRI and Upper Knife-Heart Region Lithic Artifact Analysis. In The Phase I Archeological Research Program for the Knife River Indian Villages National Historic Site, Part III: Analysis of the Physical Remains, edited by Thomas. D. Thiessen, pp. 173–262. Occasional Studies in Anthropology No. 27. US Department of the Interior, National Park Service, Midwest Archaeological Center, Lincoln. Anderson, David G. and Glen T. Hanson (1988) Early Archaic Settlement in the Southeastern United States: A Case Study from the Savannah River Valley. American Antiquity 53:262–286. Bamforth, Douglas B. (2009) Projectile Points, People, and Plains Paleoindian Perambulations. Journal of Anthropological Archaeology 28:142–157. Bauer, Alexander A. and Anna S. Agbe-Davies (editors) (2010) Social Archaeologies of Trade and Exchange: Exploring Relationships Among People, Places, and Things. Left Coast Press, Walnut Creek, California. Baugh, Timothy G. and Fred W. Nelson, Jr. (1988) Archaeological Obsidian Recovered from Selected North Dakota Sites and Its Relationship to Changing Exchange Systems in the Plains. Journal of the North Dakota Archaeological Association 3:74–94. Baugh, Timothy G. and Jonathon E. Ericson (editors) (1994) Prehistoric Exchange Systems in North America. Plenum Press, New York. Beck, Margaret E. and Sarah Trabert (2014) Kansas and the Postrevolt Puebloan Diaspora: Ceramic Evidence From the Scott County Pueblo. American Antiquity 79:314–336. Billeck, William. T. (2000) A Glass Bead Sequence for Late 17th to 19th Century Arikara Sites in North and South Dakota. Paper presented at a joint meeting of the Plains Archaeological Society and Midwest Archaeological Society, St. Paul. Carter, Tristan and Mariana Milic (2013) The Consumption of Obsidian at Neolithic Çatalhöyük: A Long-Term Perspective. In Stone Tools in Transition: From Hunter-Gatherers to Farming Societies in the Near East, edited by Ferran Borrel, Juan José Ibáñez, and Miguel Molist, pp. 495–508. Servei de Publicacions de la Universitat Autònoma de Barcelona, Barcelona. Chartkoff, Joseph L. (1989) Exchange Systems in the Archaic of Coastal Southern California. Proceedings of the Society for California Archaeology 2:167–186. Craig, Nathan, Robert J. Speakman, Rachel S. Popelka-Filcoff, Michael D. Glascock, J. David Robertson, M. Steven Shackley, and Mark S. Aldenderfer (2007). Comparison of XRF and PXRF for Analysis of Archaeological Obsidian from Southern Perú. Journal of Archaeological Science 34:2012–2024. Dillian, Carolyn D. and Carolyn L. White (2010) Trade and Exchange: Archaeological Studies from History and Prehistory. Springer, New York. Donnelly, H. (2015) The Celtic Question: An Assessment of Identity Definition in the European Iron Age. Archaeologies-Journal of the World Archaeological Congress 11:272–299. Earle, Timothy K. (1982) Toward an Explanation of Exchange. In Contexts for Prehistoric Exchange, edited by Jonathon E. Ericson and Timothy K. Earle, pp. 1–12. Studies in Archaeology. Academic Press, New York. Ferguson, Jeffrey and Craig Skinner (2003) Colorado Obsidian? Preliminary Results of a Statewide Database of Trace Element Analysis. Southwestern Lore 69(4):35–50. Fitzhugh, Ben, S. Colby Phillips, and Erik Gjesfjeld (2011) Modeling Hunter-Gatherer Information Networks: An Archaeological Case Study from the Kuril Islands. In Information and Its Role in Hunter-Gatherer Bands, edited by Robert Whallon, William A. Lovis, and Robert K. Hitchcock, pp. 85–116. UCLA/Cotsen Institute of Archaeology Press, Los Angeles. 18 JONES ET AL. Forbis, Richard G. (1985) The McKean Complex as Seen from Signal Butte. In McKean/Middle Plains Archaic: Current Research, edited by Marcel Kornfeld and Lawrence C. Todd, pp. 21–29. Occasional Papers in Wyoming Archaeology, Vol. 4. Recreation Commission, Laramie. Fredrickson, David A. (2003) Pretribelet Cultures in the North Coast Ranges, California. Proceedings of the Society for California Archaeology 16:83–87. Fry, Robert E. (editor) (1980) Models and Methods in Regional Exchange. SAA papers No. 1. Society for American Archaeology, Washington, DC. Golitko, Mark and Gary M. Feinman (2015) Procurement and Distribution of Pre-Hispanic Mesoamerican Obsidian 900 BC-AD 1520: A Social Network Analysis. Journal of Archaeological Method and Theory 22:206–247. Golitko, Mark, James Meierhoff, Gary M. Feinman, and Patrick Ryan Williams (2012) Complexities of Collapse: The Evidence of Maya Obsidian as Revealed by Social Network Graphical Analysis. Antiquity 86:507–523. Gregg, Michael L., David Meyer, Paul R. Picha, and David G. Stanley (1996) Archeology of the Northeastern Plains. In Archeological and Bioarcheological Resources of the Northern Plains, edited by George C. Frison and Robert C. Mainfort, pp. 77–90. Arkansas Archeological Survey Research Series, Vol. 47. Fayetteville, AR. Griffin, James B., Adon A. Gordus, and G. A. Wright (1969) Identification of the Sources of Hopewellian Obsidian in the Middle West. American Antiquity 34:1–14. Gunnerson, James H. (1960) An Introduction to Plains Apache Archaeology-The Dismal River Aspect. Bureau of American Ethnology Bulletin 173, Anthropological Papers No. 58. Smithsonian Institution, Washington, DC. Gunnerson, James H. (1968) Plains Apache Archaeology: A Review. Plains Anthropologist 13:167–189. Hoard, Robert J. (editor) (2012) Archeological Investigations at Arkansas City, Kansas. Kansas Historical Society Contract Archeology Publications No. 26. Kansas Historical Society, Topeka. Hoard, Robert J. and William E. Banks (editors) (2006) Kansas Archaeology. The Kansas State Historical Society, Topeka, and the University Press of Kansas, Lawrence. Hoard, Robert J. and Jeffrey R. Ferguson (2011) Source Determination of an Obsidian Projectile Point from the Massacre Canyon Site (25HK13), A Keith Phase Occupation in Southwest Nebraska and Implications for Social Connections During the Early Ceramic Period. Plains Anthropologist 56:47–52. Hoard, Robert J., C. Tod Bevitt, and Janice McLean (2008) Source Determination of Obsidian from Kansas Archaeological Sites Using Compositional Analysis. Transactions of the Kansas Academy of Science 111 (3–4):219–229. Hodder, Ian (1982) Toward a Contextual Approach to Prehistoric Exchange. In Contexts for Prehistoric Exchange, edited by Jonathon E. Ericson and Timothy K. Earle, pp. 199–212. Academic Press, New York. Hughes, Richard E. (1984) Obsidian Sourcing Studies in the Great Basin: Problems and Prospects. In Obsidian Studies in the Great Basin. Edited by Richard E. Hughes, pp. 1–20. Contributions of the University of California Archaeological Research Facility, No. 45. University of California Archaeological Research Facility, Berkeley. Hughes, Richard E. (editor) (2011) Perspectives on Prehistoric Trade and Exchange in California and the Great Basin. University of Utah Press, Salt Lake City. Hughes, Richard E. and Fred W. Nelson (1987) New Findings on Obsidian Source Utilization in Iowa. Plains Anthropologist 32:313–316. Hughes, Richard E. and Donna C. Roper (1999) Source Area Analysis of Obsidian Flakes from a Lower Loup Phase Site in Nebraska. Plains Anthropologist 44:77–82. Ibáñez, Juan José, David Ortega, Daniel Campos, Lamya Khalidi, and Vicenç Méndez (2015) Testing Complex Networks of Interaction at the Onset of the Near Eastern Neolithic Using Modelling of Obsidian Exchange. Journal of The Royal Society Interface 12:1–11. Jenkins, Richard (2008) Social Identity. 3rd ed. Routledge, New York. Johnson, Craig M. (1984) Time, Space, and Cultural Tradition as Factors in Lithic Resource Exploitation in the Middle Missouri Subarea. Plains Anthropologist 29:289–302. Johnson, Craig M. (2007) A Chronology of Middle Missouri Plains Village Sites. Smithsonian Contributions to Anthropology No. 47. Smithsonian Institution Scholarly Press, Washington, DC. Jones, George T., Charlotte Beck, Eric E. Jones, and Richard E. Hughes (2003) Lithic Source Use and Paleoarchaic Foraging Territories in the Great Basin. American Antiquity 68:5–38. PLAINS ANTHROPOLOGIST 19 Jones, George, Lisa Fontes, Rachel Horowitz, Charlotte Beck, and David Bailey (2012) Reconsidering Paleoarchaic Mobility in the Central Great Basin. American Antiquity 77:351–367. Jones, Travis W., Todd Kristensen, and Jeff Speakman (2016) Western Canadian pXRF Obsidian Sourcing. Poster presented at the 82nd annual meeting of the Society for American Archaeology, Orlando. Keyser, James D. and Mark D. Mitchell (2001) Decorated Bridles: Horse Tack in Plains Biographic Rock Art. Plains Anthropologist 46:195–210. Krause, Richard A. (1972) The Leavenworth Site: Archaeology of an Historic Arikara Community. Publications in Anthropology No. 3. University of Kansas, Lawrence. Lehmer, Donald J. (1971) Introduction to Middle Missouri Archeology. Anthropological Papers No. 1. National Park Service, Washington DC. Mac Sweeney, Naoise (2011) Community Identity and Archaeology: Dynamic Communities at Aphrodisias and Beycesultan. University of Michigan Press, Ann Arbor. Marquardt, William H. and Carole L. Crumley (1987) Theoretical Issues in the Analysis of Spatial Patterning. In Regional Dynamics: Burgundian Landscapes in Historical Perspective, edited by Carole L. Crumley and William H. Marquardt, pp. 1–18. Academic Press, New York. Renfrew, Colin (1977) Alternative Methods for Exchange in Spatial Distribution. In Exchange Systems in Prehistory, edited by Timothy K. Earle and Jonathon E. Ericson, pp. 71–90. Academic Press, New York. Scheiber, Laura L. (2006) The Late Prehistoric on the High Plains of Kansas: High Plains Upper Republican and Dismal River. In Kansas Archaeology, edited by Robert J. Hoard and William E. Banks, pp. 133–150. Kansas State Historical Society and University Press of Kansas, Lawrence. Shackley, M. Steven (2005a) Obsidian: Geology and Archaeology in the North American Southwest. University of Arizona Press, Tucson. Shackley, M. Steven (2005b) Source Provenance of Obsidian Artifacts from Prehistoric Sites in Kansas. Letter report prepared for Robert J. Hoard, June 1, 2005. Copy on file, Kansas State Historical Society, Topeka. Shackley, M. Steven (2005c) Source Provenance of Obsidian Artifacts from Prehistoric Sites in Kansas and Nebraska. Letter report prepared for the Kansas Anthropological Association, the Kansas State Historical Society, and the University of Kansas Museum of Anthropology, September 2, 2005. Copy on file at Kansas Historical Society, Topeka. Shackley, M Steven (2006) An Energy-dispersive X-ray Fluorescence Analysis of Artifacts from Three Sites in Kansas. Letter report prepared for Robert J. Hoard, October 17, 2006. Copy on file at Kansas Historical Society, Topeka. Shackley, M. Steven (2011) X-ray Fluorescence Spectrometry (XRF) in Geoarchaeology. Springer, New York. Shackley, M. Steven (2013) The Geochemistry and Archaeological Petrology of Volcanic Raw Materials in Northern New Mexico: Obsidian and Dacite Sources in Upland and Lowland Contexts. In From Mountain Top to Valley Bottom: Understanding Past Land Use in the Northern Rio Grande Valley, New Mexico, edited by Bradley J. Vierra, pp. 17–32. University of Utah Press, Salt Lake City. Skinner, Craig E. and Jennifer J. Thatcher (2009) X-Ray Fluorescence Analysis of Artifact Obsidian from the Craters of the Moon National Monument and Preserve, Butte and Blaine Counties, Idaho. Northwest Research Obsidian Studies Laboratory Report 2009-75. Speakman, Robert J., (2012) Evaluation of Bruker’s Tracer Family Factory Obsidian Calibration for Handheld Portable XRF Studies of Obsidian. Report prepared for Bruker AXS, Kennewick, Washington. Speakman, Robert J., Charles E. Holmes, and Michael D. Glascock (2007) Source Determination of Obsidian Artifacts from Swan Point (XBD-156), Alaska. Current Research in the Pleistocene 24:143–145. Speakman, Robert J. and M. Steven Shackley (2013) Silo Science and Portable XRF in Archaeology: A Response to Frahm. Journal of Archaeological Science 40:1435–1443. Spielmann, Katherine A. (2004) Communal Feasting, Ceramics, and Exchange. In Identity, Feasting, and the Archaeology of the Greater Southwest, edited by Barbara J. Mills, pp. 210–232. University Press of Colorado, Boulder. Spielmann, Katherine A. (2014) The Emergence of Forager-Farmer Interaction in North America. In The Oxford Handbook of the Archaeology and Anthropology of Hunter-Gatherers, edited by Vicki Cummings, Peter Jordan, and Marek Zvelebil, pp. 881–900. Oxford University Press, Oxford 20 JONES ET AL. Strong, William Duncan (1933) Signal Butte, a Prehistoric Narrative in the High Plains. In Explorations and Field-Work of the Smithsonian Institution in 1932, pp. 69–72. Smithsonian Institution, Washington, DC. Strong, William Duncan (1935) An Introduction to Nebraska Archeology Miscellaneous Collections Vol. 92, No. 14, pp. 224–236. Smithsonian Institution, Washington, DC. Strong, William Duncan (1940) From History to Prehistory in the Northern Great Plains. In Essays in Historical Anthropology of North America in Honor of John R. Swanton, edited by Charles G. Abbot and Julian H. Steward, pp. 353–394. Smithsonian Institution, Washington DC. Vehik, Susan C. (1976) The Great Bend Aspect: A Multivariate Investigation of Its Origins and Southern Plains Relationships. Plains Anthropologist 21:199–205. Vehik, Susan C. (2002) Conflict, Trade, and Political Development on the Southern Plains. American Antiquity 67:37–64. Vehik, Susan C. and Timothy G. Baugh (1994) Prehistoric Plains Trade. In Prehistoric Exchange Systems in North America, edited by Timothy G. Baugh and Jonathon E. Ericson, pp. 249–274. Plenum, New York. Wheeler, Richard P. (1963) The Stutsman Focus: An Aboriginal Culture Complex in the Jamestown Reservoir Area, North Dakota. Bulletin No. 185, River Basin Surveys Papers, No. 30, Bureau of American Ethnology Bulletin, Smithsonian Institution, Washington DC. Notes on contributors Travis W. Jones is a PhD student in the Anthropology Department at University of Georgia and a research assistant at University of Georgia’s Center for Applied Isotope Studies. His interests are in utilizing elemental and isotopic analyses to answer, as well as ask, archaeological questions. Robert J. Speakman is an archaeologist and geochemist. His research interests include radiocarbon dating, isotope geochemistry, and elemental analysis of cultural materials (pottery, stone, and metal). He is currently Director of the Center for Applied Isotope Studies at the University of Georgia. William T. Billeck is the head of the Repatriation Office in the Department of Anthropology at the Smithsonian Institution’s National Museum of Natural History in Washington, DC. In addition to repatriation work throughout the US, he conducts research on glass trade beads and their modification by Native Americans and also undertakes archaeology studies in the Northern and Central Plains. Robert J. Hoard is the State Archaeologist for Kansas and is affiliated with the University of Kansas and Washburn University. He conducts research on ceramic technology and the identification of source materials using replicable and quantitative methods as a tool for studying patterns of human interaction. Correspondence to: Travis W. Jones; Department of Anthropology, University of Georgia, 355 South Jackson Street Baldwin Hall Room 250, Athens, GA 30602, USA; Center for Applied Isotope Studies, University of Georgia, Athens, USA. Email: diomed1@uga.edu. Article received 21/05/2017; accepted 22/05/2018