ASM Occasional Electronic Papers No. 1: Homol'ovi IV
E. Charles Adams
The archaeological survey by Lange (1998) determined that there were two periods earlier than the Homol'ovi villages when there was substantial use of the Homol'ovi area. These were labeled the Early Period, which dated from A.D. 620 to 890, and the Middle Period, dating A.D. 1000 to 1225. Dates for these periods are based on radiocarbon dates and tree-ring dated ceramics imported from surrounding areas (Adams 2002; Lange 1998:43-46; Young 1996a). Young has conducted extensive excavations at site 36 (AZ J:14:36 (ASM)), one of the villages occupied during both these periods (Young 1996a, 1999b). She has also tested the Creswell Pueblo (AZ J:14:282 (ASM)) located in the floodplain near Homol'ovi II (Young 1999a). It is clear from these excavations that there is no overlap in terms of ceramics or in terms of stratigraphy between the Middle Period occupation of the Homol'ovi area and the massive Late Period occupation associated with the Homol'ovi pueblo villages (Young 1996a). This startling discovery indicates that the Late Period occupation was a totally immigrant population into the Homol'ovi area. Archaeologically, there are no indications of anyone in the Little Colorado River valley in the vicinity of Homol'ovi prior to this immigration.
The excavations at site 36 by Young (1999b) were designed to evaluate the nature and level of occupation of the Homol'ovi area prior to the settlement of the large villages and the evidence for or against their occupation up to the settlement of the Homol'ovi villages. This evidence was evaluated using ceramic dating, radiocarbon dating, stratigraphic association, context, and continuity of cultural traditions. None of these data sets supported the argument that occupation of this or any of the other Middle Period hamlets continued into the Late Period. The ceramic assemblage of the late phase of the Middle Period is dominated by Walnut Black-on-white, which is variously dated at ending between 1200 and 1250 (Colton 1955; Downum 1988:490; Goetze and Mills 1993:41; Wood 1987:100). However, the virtual absence of mid-1200s types, including Leupp Black-on-white, Tusayan Polychrome, and Tusayan Black-on-white, suggested to Lange (1998:46) that the occupation of the area, which defines the Middle Phase, ended by 1225. Radiocarbon dates, even at two standard deviations, do not suggest occupation beyond the early 1200s (Lange 1998:45; Young 1996a). Nowhere in the village are there deposits of Late Period artifacts that suggest a continuity of use of this site. In fact the only Late Period artifacts in the area are from nearby field areas not directly associated with site 36.
Pit houses characterized site 36 during all occupations. By the 1100s the hamlet had scattered square pit houses 3-4 m on a side and about 1.5 m deep. The hamlet also had a large, circular community structure and a small jacal surface pueblo used for storage and some exterior cooking activities. A small circular kiva was apparently never completed (Young 1996a, 1999b). The plan view of site 36 (Figure 2.8) reveals the structure of this ca. 1150-1200 hamlet, which formed part of a community of hamlets that occurred on neighboring terraces up and down the east side of the river from Homol'ovi II to south and east of Homol'ovi I (Lange 1998). None of the late features at the site have any suggestion of association with the Late Period, and there are no Late Period ceramics (Young 1999b). Finally, none of the pit house architecture, surface architecture, or ritual architecture at site 36 is represented in any of the Homol'ovi villages. Therefore, the traditions of site 36 end with its abandonment, almost certainly no later than the early 1200s (Young 1999b). The most likely period of intense occupation for site 36 and contemporary hamlets in the Homol'ovi area is during the 1130-1180 drought when the river and associated floodplain would have been most conducive to use by small-scale farmers. As noted previously, the period from 1180-1205 was wet and would have been less likely to be occupied, following Van West's (1996:29) model. This follows the pattern repeated throughout the occupation of the Homol'ovi area into the late 1200s.
A Show Low Black-on-red outlined in white pottery vessel on the floor of the latest structure at site 36 indicates the village may have been sparsely used past 1225. Additionally, excavations in 1999 at the Creswell Pueblo in the floodplain below Homol'ovi II revealed a small pueblo of four rooms and two pit structures, possibly kivas. The kivas were both used as middens and based on the ceramic assemblages from each, were probably used sequentially rather than at the same time. The latest ceramic assemblage at the site clusters around AD 1225 (Young 1999a). As research progresses on the smaller, Middle Period villages, we are realizing that occupation continued into the early 1200s. But, as at site 36, no Late Period ceramics occur in the middens of Creswell indicating no overlap in occupation.
As has been the case throughout the Early and Middle Period occupation of the Homol'ovi area, what we are most likely seeing is a break in occupation, perhaps in the late 1100s, with a smaller reuse of site 36 and the construction of Creswell Pueblo occurring briefly at about 1225. The much longer occupation span of site 36 makes it difficult to determine the extent of this late Middle Period occupation, but fairly extensive excavations to date suggest it is quite small and short-lived. The extensive midden in the Creswell kivas indicates a larger or longer occupation. The fact that Creswell is located on the floodplain, rather than the terraces, makes it mandatory, as with Homol'ovi III, that occupation be during a dry, low stream flow period. Van West (1996:29-30) identifies the period 1205-1227 as an ideal time for floodplain use, while the periods just before (1183-1204) and just after (1228-1249) were times when the floodplain should be avoided. At present the hiatus between the Middle and Late Periods would appear to be from about 1225 to 1260.
From the foregoing discussion, it is already clear that the Homol'ovi area was for all intents and purposes, vacant in the mid-1200s. It is also obvious from the discussion on the physical environment that the Homol'ovi area is a prime region to conduct intensive agriculture, that is, agriculture involving the river, but it is not ideally suited for long-term use relying solely on dry farming. The lengthy history of use of the area beginning in the A.D. 600s testifies to the high risk factors in settling in the area. No continuous occupation of more than a few years is evident between first intensive use of the area in the 600s through the early 1200s. The large size and rapid growth of the Homol'ovi villages leaves no question that the 13th century occupants were immigrants. We also have an understanding of approximately when migration took place based on the reconstruction of stream flow and tree-ring dated pottery traded into the Homol'ovi area.
Following Anthony (1990), Cameron (1995), Haury (1958) and others, the tasks before us, then are to identify the source area, to determine the size of the migrating population or population segment, to determine the conditions that made migration feasible, and to determine its distance. Additionally, migrating groups would have prior knowledge of the area and previous contact with groups in the area prior to migration. With this information, we will have a full understanding of all facets of the migration into the Homol'ovi area during the late 1200s.
The first question is, what are the conditions that make migration feasible? There are two parts to answering this question. We must look at the conditions at the home of the migrant group and their destination, in this case, Homol'ovi. We have touched on the environmental reasons why groups may have chosen to migrate into the Homol'ovi area in the late 1200s. These include permanent water, a reduced risk of flooding, and possibly abundant driftwood. Van West's (1996:29-30) stream flow reconstruction points to the 1276-1299 period as ideal for settlement of the Homol'ovi area and use of the floodplain for farming. Kolbe (1991) suggests a slightly longer period from 1263-1300. In the late 1200s a massive dislocation of Pueblo people occurred in the Four Corners region (Adler 1996; Cameron 1995; Dean et al. 1995). This has been documented extensively and attributed to warfare, social breakdowns, as well as drought (Adler 1996; Cameron 1995; LeBlanc 1999; Lipe 1989; Varien 1999); although, Van West (1994a) has demonstrated that the Mesa Verde area was still capable of supporting agriculturally based populations. Multiple causes, perhaps in a domino effect, probably created the circumstances that caused the Four Corners abandonment. The impact in the Little Colorado River and neighboring regions has been much discussed and debated. Immigration of groups from northeastern Arizona, primarily the Kayenta Anasazi, into the upper reaches of the Little Colorado and below the Mogollon Rim has been documented (Haury 1958; Lindsay 1992; Mills 1998; Reid and Whittlesey 1982). It is certainly possible that some of these immigrants, or the results of their immigration, caused settlement of the Homol'ovi area.
Groups continued to occupy three areas around Homol'ovi: the Hopi Mesas, Anderson Mesa, and the Silver Creek/upper Little Colorado River area. But the structure of this occupation changed significantly. Just as at Homol'ovi during the Late Period, the occupation of these areas changed from many small villages spread over the landscape to a few discrete clusters of larger villages. Therefore, it is possible that political or social conditions were such that groups could have emigrated from any area surrounding Homol'ovi. There is no evidence that the changing environmental conditions of the late 1200s directly caused the immigration to Homol'ovi from any of the surrounding areas. In fact, evidence suggests just the opposite, the areas were recipients of new populations (Adams 1996a; Adler 1996; Mills 1996; Pilles 1996).
What are the sources of these immigrants? The first question to be asked is where are large populations located in the late 1200s that would be in contact with or knowledgeable of the Homol'ovi area. The areas would be the Hopi Mesas/Hopi Buttes, the Flagstaff/Anderson Mesa region of the Sinagua, and the upper Little Colorado/Zuni area. Given the long history of use of this area by Early and Middle Period occupants, knowledge of the middle Little Colorado River valley was extensive by all surrounding groups, but in particular groups living sporadically or seasonally in the Hopi Mesas/Hopi Buttes region 30-80 km north of Homol'ovi. The primary ceramic and architectural traditions at Homol'ovi derive from Hopi Buttes populations (Daifuku 1961; Gumerman and Skinner 1968; Lyons 1998a, 2001; Van West 1994b). When the Hopi Buttes was abandoned in the early 1200s, groups on the Hopi Mesas and Bidahochi Pueblos in the eastern Hopi Buttes absorbed their populations. They brought their extensive knowledge of the resources of the Homol'ovi area with them. Even without these immigrants, the highly mobile populations surrounding the Homol'ovi area would have known the unique riparian resources of the Little Colorado River and major tributaries, such as Chevelon and Clear Creek. This knowledge would have informed any potential immigrants of the resources at Homol'ovi.
The best lines of evidence for the identity of the immigrants are architectural and ceramic technology. As noted with Site 36, an excellent material indicator of a society is how it expresses itself on the landscape in its settlement pattern and site layout. T. Mitchell Prudden (1903) recognized this a century ago in the unit pueblos of Mesa Verde. Studies of settlement patterns and site layouts have occupied the interest of Southwest archaeologists since (Adler 1996; Leonard and Wills 1994; Varien 1999).
Several attributes of fixed material culture have been considered in examining sources of migration. These include room size; construction material; feature location, especially hearths/firepits; layout of the village, including habitation rooms, storage rooms and ceremonial structures, especially kivas; location and organization of space and features within the kivas; and location and use of communal or plaza space. Lyons (1998a, 2001, 2003) has studied the ceremonial architecture of the region and concludes that all the varieties of rectangular kivas in the Homol'ovi villages occur in the Hopi region at the same time or earlier than at Homol'ovi or the upper Little Colorado. These include simple rectangular kivas, kivas with corner ventilator systems, and kivas with southern benches or platforms. The small room size of most of the villages is also much more typical of the plateau rather than the mountains.
As was described in chapter 1, the architectural layout of Homol'ovi IV in addition to room size, location, and kiva architecture all point to Homol'ovi IV being settled by immigrants from the Hopi Mesas. This, combined with the strong continuity of style and manufacture technology of the Homol'ovi IV ceramics, as discussed in chapter 8, clearly argues for Hopi as the source area. The majority of ceramics at Homol'ovi IV were imported from the Hopi Mesas area and consist of Hopi white wares and Jeddito Black-on-orange and Polychrome (Smith 1971). The locally made ceramics, termed Winslow Orange Ware, are identical in technology and decoration to the Jeddito series, varying only in the use of local clays and local sand tempers (Hays-Gilpin, Bubemyre, and Senior 1996). The dating of Homol'ovi IV will be the subject of the next chapter; however, it is important to note that the accumulation of ceramic and paleo-environmental evidence points to Homol'ovi IV being the first and only village occupied in the Homol'ovi area until the 1280s. As described previously, there is no continuity with previous occupations of the area, which ended about 1225. As argued by Adams (2002), the stimulus for this migration may have been as much to solidify ownership of the area by occupants of the Hopi mesas area, as to exploit resources in the area that are unique in the region, such as the ability to grow cotton reliably.
Other evidence points to Homol'ovi IV being an important trade center by virtue of its location midway between the Hopi Mesas and villages on Anderson Mesa to the south (see Figure 5.1). Evidence of exchange with Anderson Mesa villages includes abundant ceramics in the form of Alameda Brown Ware, obsidian from the Government Mountain source (Harry 1989), and shell imported from the Hohokam region. Relations with the Hopi Mesas probably involved the exchange of cotton grown locally, but also included many varieties of pottery manufactured at Hopi Mesa villages and points beyond, including the Kayenta Anasazi region. Therefore, the cultural context for Homol'ovi IV was one of isolation with contact with nearest neighbors expressed through long distance exchange. Homol'ovi IV probably served as a middleman in this exchange network, trading obsidian and shell to Hopi and pottery to Anderson Mesa. The demise of Homol'ovi IV coincides with the settlement of nearby Homol'ovi III and more distant, Homol'ovi I, Cottonwood Creek, Chevelon, and Jackrabbit villages. It seems likely that Homol'ovi IV occupants stayed in the area and moved to one of the larger villages, probably Homol'ovi I, simply due to its proximity. This is suggested by the general absence of floor assemblages at Homol'ovi IV, including metates and other heavy items that are not transported long distances.
I (1996b) have used these same arguments for the settlement of Homol'ovi II. The nearly 90 percent frequency of yellow-firing decorated ceramics from the Hopi Mesas, the size and layout of the village, the nature of the kiva architecture all point to source populations from Hopi. Lyons (1998b, 2001, 2003) work on the ceramics at Homol'ovi I clearly tie the earliest ceramics to stylistic traditions from Hopi, especially Jeddito Orange Ware. He has termed this the Jeddito style (Lyons 1998b, 1998c, 2003). The early white wares at Homol'ovi I are also clearly made at Hopi, rather than in the upper Little Colorado River area. This is consistent with the situation at Homol'ovi IV (see Chapter 8) and indicates the strong possibility that the original settlement of Homol'ovi I was principally the relocation of Homol'ovi IV, rather than an immigration of new populations into the area.
The criteria marshaled to consider the source populations for immigrants, the nature of the migrant groups, and the push and pull factors encouraging migration have been presented in this chapter. It is clear that the immigrant groups came almost totally from the north, people occupying villages on or near the Hopi Mesas. Pueblo people from north of Homol'ovi had knowledge of the area going back at least to the 600s. The Little Colorado River has always provided a unique habitat and resources for the entire Little Colorado River basin. Analysis of village architecture suggests most groups immigrated in supra-household groupings, almost certainly in related groups, possibly lineage segments or extended families, although household migration is also indicated on a limited basis.
The landscape in the late 1200s was alive with groups migrating to new areas; however, as noted, the primary migrants to Homol'ovi were probably relatively local, Tusayan (Hopi) Anasazi, rather than more northern groups, the Kayenta Anasazi. Three related reasons suggest it was pull rather than push factors that caused the migration to Homol'ovi. The first factor is environmental. Conditions for farming the floodplain were improving and the deposition of building and heating wood from driftwood would have been the missing ingredient to making Homol'ovi a viable area for large, sedentary populations. Dry periods in surrounding areas had attracted immigrants to the Homol'ovi area time and again from the 600s onward. Secondly, the scale of the villages and their organization differed greatly this time. The settlers were not in dispersed, independent farming hamlets. They were organized into labor-rich villages that were able to cope with the extreme circumstances of trying to farm the floodplain of the Little Colorado River. The final factor is the probability that the immigration was politically rather than subsistence motivated. According to Adams (2002), the occupants of Hopi mesa villages may have been protecting or asserting their rights and control of an economically and ritually important area by settling it in the face of immigrant groups from many areas looking for locations with farming potential. Homol'ovi would have been highly desirable. It is unlikely people already living at Hopi would have left their homeland due to crowding from immigrant groups settling in or near their villages. Why bother? The immigrant groups would have been encouraged to continue migrating to Homol'ovi to settle on their own rather than at Hopi. They were not. The fact that people already living at Hopi, judging from their material culture, chose instead to move to Homol'ovi supports the political/territorial explanation for expansion into the middle Little Colorado River valley in the late 1200s.
In this Section
©1995–2014 Arizona Board of Regents