Volume 1: Introduction to the Contact-Period Native California Community Distribution Model

By:

Randall Milliken,Consulting in the Past

With:

John Johnson,Santa Barbara Museum of Natural History
David Earle,Antelope Valley College
Norval Smith,University of Amsterdam
Patricia Mikkelsen,Far Western Anthropological Research Group, Inc.
Paul Brandy,Far Western Anthropological Research Group, Inc.
Jerome King,Far Western Anthropological Research Group, Inc.

Submitted to:
California Department of Transportation, District 6, 2015 East Shields Ave, Fresno, CA 93726

It may be anticipated that future scholars, undaunted by the huge mass of available published and manuscript data on California Indians, will work over the information on a tribe-by-tribe basis and prepare maps showing the domains of the identifiable or inferable tribelets. -R. F. Heizer 1966

Abstract

The Contact-Period Native California Community Distribution Model (CDM) reconstructs California Indian community ethnogeography at the time of Spanish settlement. The CDM package currently includes: (1) a GIS map layer portraying 663 independent landholding communities/year-round habitation regions; (2) a group of monographs documenting ethnogeographic information and sources for the regions; and (3) a mission register database tracking the vital statistics of individuals who moved to Franciscan missions from approximately 420 of the 663 native regions. The digital map and text are dynamic, designed to be used, updated, and revised by academic scholars, tribal scholars, government agency planners, and culture history interpreters.

Most of the CDM mapping regions—especially those across most of central California—represent the lands of true territorial sedentary village-communities or tribelets. Some other regions represent distinct clusters of loosely affiliated villages, sedentary in Northwest California but with extensive seasonal population shifts in areas east of the Sierra and in the dry inner Coast Ranges. Mapping regions in parts of southern California are merely arbitrary divisions applied for purposes of local vicinity analysis to portions of extensive open networks of small, inter-marrying village-groups. Confidence in the accuracy of the regions and their boundaries varies greatly due to differences in the richness of existing data and to differences in scholarly interpretations of those data.

The database, narratives, and regional boundaries are presented in a GIS application and Wiki format. This will allow future scholars to reconsider regional boundaries, to expand or annotate existing monographs, and to contribute new monographs for the currently unfinished areas.

Introduction

The Contact-Period Native California Community Distribution Model (CDM) is a digital atlas and wiki encyclopedia that models the socio-political landscape of native California at the time of first contact with the Spanish, a rolling moment from the 1770s to the 1830s. The CDM atlas portrays a model distribution of 663 community regions (inferred or known village communities or tribelets) across California on a GIS digital map layer, divided into 14 analytical zones that combine regions on the basis of mutual histories, shared language, and similar land-use patterns. The associated encyclopedia consists of "wiki" monographs that gather together archival information for each of the community regions (i.e., presented within a collaborative website that allows for creation and controlled editing of interlinked web pages). An additional key element, a Mission Register database, provides locational information for the CDM regions from which the people were entirely removed to the Franciscan missions between 1770 and 1835. These separate elements together form the CDM (Figure 1).

The CDM Project is being developed by Dr. Randall Milliken for use by academic scholars, tribal scholars, government agency planners, and culture history interpreters. It brings together decades of research and mission record analysis. It has been produced by Far Western Anthropological Research Group, Inc. with support from a number of district environmental branches within the California Department of Transportation. As of spring 2010, monographs have been prepared for more than 80 of the regions, primarily in the San Francisco Bay area, the Sacramento Valley, and the San Joaquin Valley. Our future goal is to place the CDM within an academic setting so that the GIS map and wiki monographs will be available for further development—expansions, modifications, and annotations—by scholars knowledgeable about the contact-period ethnogeography of every local area of California.

Project Purpose

Figure 1. Creating the Community Distribution Model.

No statewide "tribelet" map has ever been constructed for contact-period ethnographic California. Robert Heizer commented on that fact in his 1966 mapping project entitled Languages, Territories, and Names of California Indian Tribes, a study that introduced the only composite statewide map of local groups based upon C. Hart Merriam’s many field notes, and compared against A.L. Kroeber’s 1925 map of the language groups of California. At the time, Heizer (1966:9) wrote:

Kroeber estimates that California held between 500 and 600 … ’independent and separate definable groups.’ It may be anticipated that future scholars, undaunted by the huge mass of available published and manuscript data on California Indians, will work over the information on a tribe-by-tribe basis and prepare maps showing the domains of the identifiable or inferable tribelets. Such a task is far too complex and time-consuming to be attempted here.

The digital process used in the CDM makes it easier than ever before to create a statewide map of contact-period California ethnogeography:

• We are not committed to a published, static, one-size map.
• It is especially valuable for mapping communities mentioned in mission records that can only be located indirectly and tentatively by date of baptism and marriage ties.
• The digital map can be modified as scholars examine the implications of one or another interpretation of the archival record for a given local area; i.e., the network placement of groups in relation to one mission may need to be redesigned when the inferred pattern is compared with those developed for the communities of adjacent missions.

Assessing the Regional Model

Figure 2. Community Distribution Model Analytical Zones within California.

It is a down side of this "cubby hole" approach that some readers will fail to read caveats presented here and believe that the regional boundaries are all precisely documented local group borders. But that problem is less serious than the tyranny of "language group as tribe" that informs the understanding of the public and some scholars today. The upside of the model presented here lies in its exposure of the variation in quality of ethnographic information from one region to the next and in its flexibility for adding new information about regions and their possible boundaries through in-depth local studies. In the future, thorough studies of one region, or a group of contiguous regions, might be the subject of entire academic symposia or doctoral dissertations.

Ethnographic Landholding Groups in California

The groups of California Indians portrayed on historic tribal maps of the United States—Cahuilla, Chumash, Pomo, Washoe, and dozens of others—are actually language groups, delineated and named by western scholars. Landholding groups in California at the time of European contact were local communities—people closely bound by blood, marriage, and proximity of residence. Anthropologist A.L. Kroeber introduced the term "tribelet" in 1932 for the type of territorial multi-family landholding community that prevailed across most of California. At the end of his career Kroeber considered the tribelet to be one of four landholding group types, the other three being: (1) small, ceremonially linked village clusters of northwest California; (2) small, single-lineage villages in the southern desert; and (3) high-population Colorado River multi-village tribes (Kroeber 1955, 1962; see also Bean 1978; Gifford 1926; Heizer 1966). While these general categories are maintained here, we have used broader terminology to identify them and have assigned groups to slightly different categories than identified by Kroeber (1955, 1962).

• Tribelets

Bay Miwok, Coast Miwok Ohlone, Achomawi, Rumsen Costanoan; some Patwin, Pomo, and Yokuts; perhaps Plains Miwok and Nomlaki.

• Loose Regional Communities (includes Kroeber’s ceremonially linked villages)

Washoe, Hupa, Karok, Yurok

• Completely Unbounded Networks of Small, Single-Lineage Villages

Uto-Aztecan, Ipai/Tipai; perhaps Sierra Miwok, Hill Nisenan, Hill Northwest Maidu, Yana

• Ambiguously Bounded Regions of Large Sedentary Villages

Sacramento Valley (Patwin and Maidu), Santa Barbara Channel (Coastal and Island Chumash), and Colorado River (Cocopa, Mohave, Yuma)

The Tribelet

Local tribelets had fixed territories, a single head person, and a large enough population (200 or more people) to have been at least 50% endogamous. They varied in number and relative size of villages. Most Patwin, Pomo, Valley Nisenan, Valley Northwest Maidu, and Yokuts tribelets seem to have had capital villages and outlying suburbs (Kunkel 1962). Some Patwin, Pomo, and Yokuts tribelets, and all Bay Miwok, Coast Miwok, and San Francisco Bay area Ohlone tribelets had multiple villages of relatively even size (Milliken 1995:20-24), as perhaps did the Plains Miwok (Bennyhoff 1977) and the Nomlaki (Kroeber 1932:373).

Kroeber coined the term "tribelet" for multi-family political communities that held most of the California landscape. He summarized his thinking about the tribelet and lineage as follows in a 1955 monograph:

The tribelet was first called the "village community" in the ethnological literature, to distinguish it from the village as a mere physical settlement; and its size was underestimated at only around 100 members. Later, I deliberately coined the name tribelet to designate it as a sovereign though miniature political unit, which was land-owning and maintained its frontiers against unauthorized trespass. At the average of 250-300, there would have been a full 500 tribelets in the later American state of California (Kroeber 1955:307-308).

Kroeber (1962:34-36) described three case studies of California tribelets to illustrate local variation in tribelet size and population density:

1. Pomo-speaking tribelets of the North Coast Ranges averaged 240 persons in territories about ten miles by ten miles in size, for a population density of 2.4 persons per square mile.
2. Yokuts-speaking tribelets of the San Joaquin Valley averaged 350 members each in territories about four times larger, with a suggested population density of 1.4 persons per square mile.
3. Achomawi-speaking tribelets of northeast California contained about 275 persons and held such large territories that their populations were only 0.5 persons per square mile (when both their restricted winter territories and expanded summer territories are included).

Loose Regional Communities

Loose regional communities are clusters of family groups that shared numerous short-term villages within tribelet-sized regions that lacked defended boundaries and central leadership. The loose regional community model is well illustrated by the Washoe language group (D’Azevedo 1986:485). Among the Washoe, daily life and the yearly life cycle revolved around two levels of community—the local village level and the regional multi-village level. The local village community, the basic unit of social organization, was a group of closely related households that shared the same winter camp and identified with a single local leader. The Washoe themselves referred to this local village community level as "the bunch." A local community populated one winter village that generally consisted of a cluster of two to ten houses. While the local community did not spend the entire year together in that single village, it did consider itself the basic cooperative unit for communal hunts, defense, and group ceremonial expression year-round. Membership in "bunches" was fluid, as individuals and families often shifted residence temporarily or permanently "to the households of other relatives in the same or a distant community" (D’Azevedo 1986:483).

The Washoe regional community included all the local communities within a region approximately 10-15 miles in diameter. An example regional community included the four local communities of the Woodfords-Markleeville area of Alpine County, California. Together the four local communities formed the "dwellers in the corner where rivers flow away out" regional community (D’Azevedo 1986:468). Constituent local communities united into regional communities "by identification with place and by loose ties of kinship and constituted a population of hundreds of persons with whom there was some degree of familiarity and mutual trust" (D’Azevedo 1986:484). The regional community, as a cluster of local communities, had a number of headmen, but in ethnographic times it seldom acted as a single political unit.

In Northwest California the speakers of Hupa, Karok, and Yurok lived in numerous small villages along major rivers. Nuclear families were stratified between the wealthy and the poor. Extended family lineage affiliation did not seem to be important. Nor did groups of hamlets cluster into local landholding tribelets under centralized political authority. However, there was some regional organization at the ceremonial level:

There were practiced a series of "world-renewal" rituals, each made separately and with a fair measure of differentiation, at designated spots, and supported by the inhabitants of a recognized tract surrounding the sacred spots. Both in extent and in population these tracts resemble tribelets [Kroeber 1955:311].

Unbounded Networks of Small Single-Lineage Villages

A third form of land use was based upon small, semi-sedentary villages of 40-100 inhabitants that intermarried with near neighbors in overlapping spheres of outreach that extended for great distances, without establishing any tribelet-sized regions. Such land-use patterns were clearly present among many Uto-Aztecan groups of southern California, probably among the Ipai/Tipai of San Diego County, and perhaps among the Sierra Miwok, Hill Nisenan, Hill Northwest Maidu, and Yana. In all these areas, evidence is either lacking or contradictory for fixed territorial tribelets. Populations lived in numerous small villages, many with their own ceremonial dance houses, scattered along ridges and in favorable valleys. The marriage networks of such communities overlapped with one another. Under this form of organization, each local village was politically independent, but each was so small that it was completely exogamous to near neighbors, much smaller than the 10-15-mile-diameter areas of classic Yokuts or Pomo tribelets. The independence of each local village community is upheld in the classic works of Kroeber (1925) and Strong (1929) for southern California as a whole, as well as a later study by Bean (1972) for the Cahuilla. Villages had complex intermarriage patterns driven by moiety considerations among the Uto-Aztecans (Bean 1972; Strong 1929). However, blanket terms were not applied to distinct multi-village communities because distinct multi-village regions do not seem to have existed:

The California lineage group was what the name implies, a line of male kinsmen who were autonomous in a territory sufficient to support them. They took their wives from and married their daughters into other lineages. The bonds of kinship might be transcended if one lineage became reduced and took refuge with another. A desert or infertile habitat, necessitating a spread of population, tended to preserve political organization on the lineage basis. Contrarily, in habitats rich in food, it is presumed that several lineages tended to coalesce into permanent village—that is tribelets [Kroeber 1955:308].

Ambiguously Bounded Regions of Large Sedentary Villages

The last of the four land-use regions are the two areas of California where there were closely spaced, large permanent villages, each seemingly independent of the other, but so close together that they must have shared hinterlands in different ways than the classic Yokuts or Pomo tribelets. These two areas were the central Sacramento Valley (Patwin and Maidu) and the Santa Barbara Channel (Coastal and Island Chumash).

Similarly, tribelet-level regional differentiation seems to have been absent along the Colorado River, including southern California. Groups of 2,000 to 3,000 farming people (Cocopa, Mohave, Yuma) held tracts of riverine lands far more extensive than any of the small lineage or tribelet areas elsewhere in California. During the Spanish era in the southwest these groups were capable of mobilizing under a central leadership to make war upon one another. Yet, "it is possible that even these tribes were conglomerations of earlier tribelets," Kroeber (1955:310) suggested.

Summary for Land-Use Types

These different land-using groups are one aspect to be considered when delineating regional boundaries. There is the contrast between sedentary groups with fixed boundaries (tribelets) and mobile groups with fluid boundaries (loose regional communities). Then there are small, single-lineage villages with overlapping outreach areas (single-lineage villages), in contrast to closely spaced, independent villages with shared hinterlands (ambiguously bounded large villages).

We discuss how these landholding groups were used to create regional boundaries in Section 1.5.

Franciscan Mission Registers and Ethnogeography

This chapter describes the use of Franciscan mission records of baptism, marriage, and burial for reconstructing the local-level ethnogeographic tribal landscape. These mission records are the only systematic source for such reconstructions for a large portion of California. By the 1830s all tribal people of the San Francisco Bay area, the Los Angeles basin, the South Coast Ranges, the Delta, and the western side of the San Joaquin Valley—more than 50,000 people—had left their lands and moved to the missions. The mission records contain the only information regarding the original home groups of the vast majority of those people. And because the missionaries were obliged to document basic information about each individual that they baptized, the mission registers provide a systematic and nearly comprehensive tally of native groups for much of California.

Of the 663 CDM mapping regions identified in California, 420 are in areas that were completely or partially disrupted by the Franciscan mission system. Unfortunately, the mission record data on native homelands is uneven and often opaque for most of those areas. Missionaries did write down the names of thousands of communities that they called "rancherias," but they seldom clarified whether they were referring to a specific village, on the one hand, or a regional multi-village group on the other. Nor did they often provide explicit clues regarding the locations of the communities they listed. This chapter describes the techniques that have been developed over time to overcome these and other weaknesses in the mission register ethnogeographic information.

Rancherias in the Mission Register

Figure 3. Father Serra’s Leather-bound, Mission Carmel Libro de Confirmaciones. Diocese of Monterey Chancery Archives, Monterey.

Figure 4. Record of the First Baptism Performed by Father Serra, listed in Mission San Carlos Borromeo’s Libro de Bautismos. Bernardino de Jesus of the Rumsen Tribe, Born in the Achista Rancheria.

Alternate spellings of rancheria names by different priests are a major problem as they can be quite extreme. This issue can be aided by bringing together the mission register information for nuclear and extended families so that names spelled out by different priests over many years can be seen at one time. In central California, for example, approximately 4,000 distinctly spelled rancheria names are listed in the registers of baptism, burial, and marriage of the 11 Franciscan missions (San Francisco Solano south to San Miguel). These can be reduced to 1,300 names when alternative spellings of specific rancheria names are recognized. The processes of re-sorting mission register information by family group and of tracking "standardized" rancheria names in specific mission register entries will be documented in subsequent sections of this chapter.

A problem in understanding the socio-political meaning of the many rancheria names is that of synonymy. Some missionaries labeled groups by the name of the group’s headman, others labeled them by the name of their largest village, and still others named them by some directional or regional term. One missionary might label a group by a word used by its neighbors of a different language, while another might name the same group by a label that it called itself; this was common among groups that sent people to more than one mission. Through study of extended family groups, some synonymous rancheria names become apparent.

Another problem in interpreting the socio-political landscape from mission register information is that of scale. It has been mentioned that the word "rancheria" means "community." From the earliest days of exploration in California, Spanish diarists used the term in describing specific villages, or clusters of grass houses, inhabited by the tribal people they encountered. That usage corresponds to the modern Spanish-English dictionary definition of rancheria as "a collection of huts, like a hamlet" (Velásquez 1974:551). But the English term "community" can also mean a group of people who share a number of villages within a fixed territory, and the Spanish term "rancheria" came to be used that way in California as well (Milliken 1987:59, 1995:21, 233). Both in military diaries and in the mission records, the term "rancheria" may signify either a specific village or the community of shared identity that utilizes one or more specific villages. This conflation of two meanings partially explains why some rancheria names appear only five or six times in all the mission registers, while others appear hundreds of times.

In sum, some rancheria names in mission records refer to mobile bands, others represent specific sedentary villages, and still others represent the names of multi-village regions. The process that has developed over time to overcome problems of scale and synonymy in the mission registers, to track individuals as members of identifiable communities (distinct single village or multi-village groups), and to choose "standard" names for those groups, will be discussed in the following sections of this chapter.

Pioneering Mission Register Studies

Figure 5. Map showing the Progressive Collapse of Autonomous Tribal Areas during the Years 1790 to 1806.

Figure 6. Marriage Ties Between the Chupcan and Their Nearest Neighbors Based on Baptisms of Married Couples at Mission San Francisco and Mission San Jose, 1795-1811.

James A. Bennyhoff advanced mission register study by making two of Cook’s approaches explicit. He wrote:

By plotting the tribelet names and the numbers baptized against the years, a significant constellation appears for each tribelet by which one can judge the approximate distance from the mission and the year of most intensive contact [Bennyhoff (1977:20].

Modern mission register-based ethnogeography follows three principles discussed by Bennyhoff:

• Rancherias close to missions generally sent their people for baptism earlier than villages at greater distances, resulting in a "domino" effect outward from each mission.
• Rancherias near one another had a greater number of intermarriages than rancherias separated by great distances (the basis for "marriage pattern" studies of group proximity).
• Rancherias that sent people to two or more missions usually existed in areas equidistant between the missions in question. (This rule does not apply north of San Francisco Bay due to the complex history of mission development and outreach in that area.)

When the "wave-front" of mission outreach is mapped, the relative distance of otherwise-unlocated groups baptized in specific years can be determined (Figure 5). The "marriage pattern" technique looks for strongly intermarried groups, then infers that such groups may have held lands contiguous to one another (Figure 6). Bennyhoff (1977) used a combination of these two techniques, together with clues from classical ethnography, to map ethnographic central California.

In southern California, Alan K. Brown (1967) pioneered the rigorous use of mission register clues to help determine Chumash village sizes and locations. While Brown did not systematically evaluate quantitative family links among villages, he carefully extracted quotations from specific baptismal entries to emphasize inter-village relationships that supported locational surmises. Techniques introduced by Bennyhoff and Brown were further developed by Chester King. King’s (1969, 1975) initial studies of Chumash village locations and their populations relied on J. P. Harrington’s (1986) field notes, other early texts, and aggregative summaries of Santa Barbara County vicinity mission register entries. King (1973) also gleaned information from the Mission San Juan Bautista registers for T. King and Hickman’s (1973) reconstruction of tribelet locations in the Mission San Juan Bautista vicinity.

Family Reconstitution to Solve Rancheria Riddles

Figure 7. Family Reconstitution Chart: Onesimo and Maria de Los Angeles Baylon.

The backbone technique for identification and resolution of such problems is the family reconstitution method. Family reconstitution, first introduced for social geography studies in Europe (Henry 1976; Wrigley 1966), is the process of amalgamating dispersed bits of information about individuals, married couples, and extended families into composite information sets (Figure 7). In mission register studies, family kinship charts are reconstituted from various register entries, resulting in composite data sets that illuminate: (1) synonymous terms for rancherias; (2) relationships of rancherias that are villages to rancherias that are regional names; (3) patterns of intermarriage among communities; (4) timing of family and community movements to missions; and (5) numerous demographic processes that do not emerge from aggregative mission register studies.

Family reconstitution was first applied to California mission data sets by Chester King (King 1974, 1977, 1978). King constructed kinship charts, exposed alternative names of villages that sent people to more than one mission, and increased the number of identified inter-village marriages. The reconstituted evidence allowed King to infer the locations of poorly documented Northern Chumash locations around Mission San Luis Obispo.

Other researchers soon began constructing kinship charts following King’s lead (Horne 1981; Milliken 1981). John Johnson (1982) documented marriage relationships among Santa Barbara Channel Island villagers and between them and mainland populations. Randall Milliken (1981) used mission register information to identify the relationship between the Rumsen tribelet of the Carmel Valley and its five constituent villages. John Johnson (1988:248-288) applied quantitative techniques borrowed from social-network analysis and cultural geography to examine inter-village social relationships among Chumash of the Santa Barbara Channel. Johnson’s diagrams of inter-group interaction, for all people baptized from forty Chumash towns and villages within the Santa Ynez Valley and along the Santa Barbara Channel coast, suggest some alternate ethnogeographic placements to those in King’s 1984 study.

Milliken (1991) used Bennyhoff’s indirect mission register techniques, as well as family reconstitution, to identify San Francisco Bay area group locations and local group migrations to missions Dolores, Santa Clara, and San Jose between 1777 and 1810 in a work subsequently published as A Time of Little Choice: The Disintegration of Tribal Culture in the San Francisco Bay Area, 1769-1810 (Milliken 1995). During the same time period, Chester King (1994) applied family reconstitution methods to Mission Santa Cruz register data to examine inter-group marriage patterns in the Santa Cruz Mountains.

Beginning in 1993, a team of ethnohistorians led by Johnson and C. King conducted a comprehensive study of Chumash ethnogeography, marriage patterns, and family survival into the American period. The team built a six-mission database that included information on all baptized Indian people from Chumash territories between missions San Fernando and San Luis Obispo. The database was used by McLendon and Johnson (1999) to document the genealogical relationships between the Spanish-contact Chumash and the Chumash people of the twentieth century. In that study, McLendon and Johnson (1999:31) used an enhanced version of the domino and marriage pattern principles to re-evaluate village locations within the Northern Chumash region (the Mission San Luis Obispo outreach area), adjusting King’s 1984 map of that western Santa Barbara/southwestern San Luis Obispo county area.

Computers for Mission Register Data Management

Initial Development of Computer Databases

The base data for the present CDM process in the coastal portions of California south of the Russian River are computerized mission records that augment family reconstitution and contain fields that assign individuals and married couples to groups with standardized group names. Computer databases have been key tools in California ethnogeography since the mid-1980s. Gary Coombs (1975) was the first person to use a computer database to track and cross-refer thousands of bits of information in the mission registers. For his Ph.D. dissertation, Coombs used computer punch cards to collate and sort information about the history of baptism of Chumash villages in the present Santa Barbara County area. A short time later, Milliken (1981) developed computer punch card sets for the first 800 baptisms at Mission Carmel and the first 1,800 baptisms at Mission Dolores (Milliken 1983). Johnson (1988:248-288) used the database to apply quantitative techniques on the Santa Barbara Channel.

Milliken moved the Mission Carmel and Mission Dolores data from punch cards to a dBASE database format in the early 1980s, then continued to expand the Mission Dolores database in the late 1980s to include information on more than 5,000 individuals. The Mission Carmel database aided construction of kinship charts for people from rancherias whose members went to Mission Carmel from Esselen-speaking areas in the northern Santa Lucia Range and the Salinas River Valley (Milliken 1990). For other projects, including his dissertation, Milliken developed separate databases for all 11 of the central California missions from San Miguel north to San Francisco Solano from 1988 to 2004.

Current Comprehensive Databases

Rapid development in mission register digitization since the year 2002 has led to the emergence of two distinct comprehensive California mission databases. One is the Early California Population Project (ECPP) database, with more than 100,000 records available on-line through the Huntington Library in San Marino, California. That system was developed by historian Steve Hackel, with initial design advice from John Johnson and Randall Milliken, among others. The ECPP database, while excellent for genealogical research, lacks rigorous standardization of "rancheria" information and does not tie rancheria names to actual locations on the ground.

The second comprehensive California mission register database, and the one utilized for this report, is the California Mission Database built by Randall Milliken. It took form in 2006, when Milliken consolidated the series of independent mission-specific databases he had built over the years for the 11 central California missions into a single Microsoft ACCESS database. One of the key tables is the records for baptisms, also called the "Individuals" table (see Appendix Table A-1). In 2007 John Johnson allowed Milliken to modify the structure of Johnson’s Santa Barbara Database (six missions from San Fernando to San Luis Obispo) and consolidate it (for limited research purposes) into Milliken’s ACCESS database structure. The result is the Mission Database, consisting of records up through the year 1840 for the 17 missions from San Fernando north to San Francisco Solano. The Missions Database has an individuals table that includes 55,603 records representing 32,375 tribal converts, 13,696 mission-born California Indians, 7,919 non-Indians, and 1,543 records marking duplicate or skipped entries. This table also contains linked death records of 33,210 death records of the baptized people except in the case of non-Indian immigrants baptized elsewhere. Finally, a separate table includes 16,333 marriage records (14,689 Indian couples, 1,535 Hispanic and other foreign couples, and 109 Hispanic/Indian couples) cross-referenced to the combined baptisms/deaths table.

As of 2010, the Missions Database also contains a separate "Regions" table that cross-links to the 663 GIS-mapped California habitation regions (see Appendix Table A-2). The Regions Table in the Missions Database also cross-links internally to baptismal records in the Individuals Table through a "Regions" field. The nature and use of the "regions" concept is explained in next chapter of this document.

Building the Community Distribution Model

This chapter presents how the CDM of ethnographic California, with 663 year-round habitation regions, was built. The first section illustrates how the model marks off a large segment of California that did not support year-round populations. Then we show the variation of regional identification from one portion of California to another due to differences in the nature of political communities and in the nature and quality of the information available today. The final portion of this chapter points out the results of printed and digital products and explains the reason for building a structure with both local region monographs and 14 analytic zone reports. Details of computer use in this process are laid out in Appendix B.

Delineating Seasonal-Use Areas

It is for the most part neither necessary nor desirable to delineate small local regions for parts of California that were not utilized intensively year-round, i.e., parts that were either covered with snow for much of the winter or very dry for much of the summer. These areas were utilized seasonally by people who lived in hospitable landscapes directly bordering them. Identified in the CDM as "seasonal regions," these harsh landscapes tend to be larger, as mapping units, than the regions that had year-round inhabitants. They are further distinguished from the year-round regions as they were often shared, and occasionally contested, by people from multiple year-round regions, often speaking different languages.

For cismontane California, the CDM upland boundary has been set at 2,780 feet elevation at the Oregon border, rising to 5,800 feet elevation at the Mexican border. As one travels ten miles north or south, the upland boundary moves downward or upward some 45 feet. This is due to changing day length and solar radiation exposure. It is, of course, only a general rule; distance from the ocean, weather patterns, and local topography render reality much less orderly. Nevertheless, the CDM boundary rule works for cismontane California because it was based upon the actual distribution of ethnographic villages in uplands across California. The rule has been ignored in the Pit River watershed and areas of transmontane California north and east of Walker Pass; in those areas year-round villages existed at elevations of 1,000 feet or more, higher than those over the mountains to the west.

Of special note, there are some places in California where small numbers of people seem to have been willing to live in very harsh environments year-round. These areas are all in transmontane upland and desert California. The Truckee Mono Lake regions are examples, as is the Mojave Desert. Most of these were at great distances from the typical year-round regions of California, suggesting an alternative, and more efficient, adaptation than the yearly "transhumance" found elsewhere in the state.

Shaping Year-Round Local Group Regions

Four different approaches were taken to delineating regions based upon two different sets of variables. One key variable is the nature of the land-using organization. Region determination is a different activity in areas where people lived in bounded territories (e.g., Pomo, Pit River, Yurok) than in areas were no regional territories seem to have existed (e.g., Cahuilla, Foothill Nisenan).

The second key variable is the nature of the data. The problems and opportunities for reconstructing ethnogeography are much different in areas strongly impacted by mission outreach (west central and west southern California) than in areas where the available information emerges primarily from classical ethnographic fieldwork (northern and eastern California).

In all four types of areas, the digital mapping process is preceded by working with small paper disks of known villages and groups placed on a large map. From that point forward, the decisions about coalescence of locations and placement of region boundaries is highly variable. Below, the different kinds of data and mapping procedures are summarized; Volumes 2-15 fully document these procedures for each analysis zone.

Bounded Regions North and East of Mission Areas

The following areas were the subject of good field ethnography that established core homelands, and in some cases mapped specific boundaries (some porous, others strict) between distinct territorial groups:

4. Modoc speakers around Tule Lake and Palaihnihan speakers of the Pit River drainage, northeast California.

5. Tolowa, Karok, Yurok, Wiyot, and Hupa lands, from the Eureka vicinity to the Oregon border in northwest California.

6. Most North Coast Range areas where Pomo and Yukian languages were spoken (the exception being the missionized Santa Rosa Plain and northern Napa County near the San Francisco Bay area).

7. East side of the South San Joaquin Valley and adjacent Sierran foothills where Yokuts languages, Mono and Tubatulabal, were spoken.

8. Lake Tahoe vicinity, where Washoe group winter territories were politically unbounded but well-defined.

9. Owens Valley where Mono/Northern Paiute groups had loose territorial boundaries.

In all these areas, regional boundaries are drawn with confidence that they generally reflect a factual ethnographic situation.

Open Village Networks and Ambiguous Regions North and East of Mission Areas

For large areas of California, field ethnographers gathered dozens of village names and locations, but found little or contradictory evidence of regional coalescence and boundedness. Such areas include:

1. Athabascan-speaking portions of the North Coast Ranges in the Eel River drainage.
2. Yana, Nomlaki, and Wintu areas of the northern Sacramento River drainage.
3. River Patwin, Northwest Maidu, and Valley Nisenan villages of the Sacramento Valley (said to be tribelets, but boundaries absolutely opaque).
4. The entire Sierra Nevada from the Feather River south to the headwaters of the San Joaquin River.
5. The Kawaiisu area south of the Kern River.
6. Shoshone and Southern Paiute-speaking areas of desert southeast California.
7. The Mountain and Desert Cahuilla areas from San Gorgonio Pass to the south and east.

In areas where flexible home-base groups varied in size and locations from year to year, and sent their people far out into the Mojave Desert seasonally, we have divided the "year-round" habitation portion of the landscape into "tribelet-sized" regions to bring focus to what local ethnographic clues are available and to consider differences from one local region to another. These types of regions have significant year-round wells or springs and are often adjacent to uplands covered in snow for parts of the winter.

For the basin and range north of Mojave River, Panamint and Chemehuevi people had such flexibility that the concept of seasonal "home bases" may not even apply. Nevertheless, we have attempted to isolate regions that had the best resources for semi-permanent camping sites all year-round, and to leave the surrounding areas as "flexible use areas."

Mission-Impacted Areas that were Bounded Territories

Figure 8. Ethnographic Regions by Major Language Group.

Figure 9. Variation in Mission-Induced Depopulation across California.

Figure 10. Ethnographic Regions by Tentative Population Density.

1. The San Francisco Bay Costanoan/Ohlone, Bay Miwok, and Coast Miwok speaking groups.
2. The Plains Miwok and Yokuts groups of the Delta and central riverine corridor of the San Joaquin Valley.
3. Some southern Costanoan-Ohlone groups of the Monterey Bay area, all Esselen groups, and some Salinan groups of the South Coast Ranges.
4. Takic speakers of the Los Angeles Basin and the San Luis Rey area.

It is easiest to map such groups in vicinities that have geographic boundaries, such as the San Francisco and Marin peninsulas. In all other areas, regions are inferred using classic indirect mission register analysis.

Mission-Impacted Areas with Numerous Independent Villages

Mission records give no indication at all of the presence of multi-village regional groups in some areas of west-central and southern California. Areas include:

1. Coast Miwok-speaking areas of western Marin County along Tomales Bay in the San Francisco Bay area.
2. Much of the densely populated Chumash-speaking Santa Barbara Channel.
3. Serrano Takic-speaking areas along the north edge of the San Gabriel Mountains.
4. Luiseño Takic-speaking areas around Mission San Juan Capistrano and eastward toward San Jacinto.
5. Ipai/Tipai-speaking village areas through San Diego County.

Although all these areas are listed together, the mapping problems that each present are unique. The techniques and results for each will be discussed in appropriate Analysis Zone volumes.

Regions and Analysis zones

After factoring in landholding community types, Franciscan mission registers of baptism and death, rancheria delineation, domino effects of missionization, marriage patterns, family reconstitution, mortality rates, inferred population densities, classical ethnographies, and seasonal-use areas, we have delineated 663 year-round local group regions in California. These are locations on the landscape that represent the known or inferred homeland of contact-period communities. The quantity of mapping regions, their sizes, and their precise boundaries, as presented here, underwent many changes during an iterative process, and will continue to change with future research.

We have grouped these regions into 14 analysis zones on the basis of mutual histories, shared language, and similar land-use patterns (Figure 8). The zones also combine some of the various factors that can affect region delineation, such as community types, the quality of mission record data, and the variation of ethnographic field data.

Appendix C presents the current array of 663 regions, broken down by analysis zone. Although the region count is larger than the 500 suggested by twentieth-century authors, it certainly is in the same order of magnitude. It is impossible to know how close the CDM model approaches ethnographic reality in the absence of a completely accurate survey in the year 1770. Thus the model can be criticized, even rejected, by perfectionists. But one of its main points is to illuminate assumptions about group locations and stimulate careful review of the evidence behind each of those assumptions.

Naming Conventions for Regions and Native Groups

An attempt was made to be systematic in applying name labels, both the modern geographic regional names and the chosen "tribal" group labels, for the regions. For the modern regional names, an attempt was made to pick something broadly recognizable, usually the largest town in the region. In rural areas a geographic feature, such as a creek or mountain peak, was used as a label, always one that is recognized by Google Maps. Modern rancheria names were typically not chosen, in order to reduce any fuzzy area between the ethnographic information and the map locations of today.

In choosing the ethnographic tribal or "rancheria" group label for each region, a probably hopeless attempt was made to minimize controversy and best reflect the preponderance of evidence that might be portrayed in a well-developed monograph. There is little problem in areas with minimal ethnographic data where only one field ethnographer worked; the name supplied by that source is used. Where two ethnographers supplied very different group names, but both with merit (e.g., Kroeber and Merriam), the label chosen followed that of the earliest (or only) formal publication.

In mission-impacted regions where only mission register group names are known, the most common spelling of a group in the Spanish orthography of the mission registers was used as the label. In other cases, where the published literature has slightly modified the spelling, the published standard was used (most common for Yokuts-speaking groups that went to the missions).

Results as Digital Data

Regional attributes can be used to generate numerous maps that will aid in the task of moving between close local study and analysis of broad regional patterns, another goal of the CDM. Two maps generated from those attributes are presented here, as a matter of illustration, rather than full explanation. Figure 9 shows the relative impact of the missions in the regions. Figure 10 is a draft representation of the population density of contact-period California.

Moving Forward with the CDM

The Community Distribution Model is a program, rather than a final product, for mapping and documenting the geography of the local socio-political groups of ethnographic (western-contact) period California. Kroeber (1962:3) estimated that there were 500 to 600 independent definable political groups in the State. Heizer pointed out in 1966 that no systematic attempt had yet been made to map such groups, due to the "huge mass of available published and manuscript data on California Indians" (1966:9). He anticipated, however, that future scholars would map out the domains of those groups across the state. The CDM sets out along the path to accomplish that goal.

As it turns out, the goal of a perfect map of contact-period California local tribe areas can never be reached. One reason is that there are irreparable holes in the ethnographic record and the other reason is that there were areas of California that were not organized into definable local political groups. Recognizing these problems, the current version of the CDM separates the state into 663 lowland year-round local group regions, many of which are based on strong data, some of which are reconstructed from opaque data, and a few of which have been cut out from the landscape in the absence of any real data, just so that the map can be finished and local ethnographers can be challenged to fix it.

The concept of "editing and fixing" is critical to the CDM. The project is set up so that the groups and their regions are, or will be, described, defended, and questioned, in "wiki" text documents to which proven scholars from academia, tribal communities, and agencies will be able to improve on the work through primary field notes of the early ethnographers. When consensus is reached among the scholars in certain zones of the state, the GIS model boundaries themselves can be changed to reflect this improved knowledge. There will be problems, especially in defining and mapping the parts of California where small local village groups intermarried in overlapping outreach areas across wide landscapes, without developing true central leadership and territorial boundaries comparable to the Pomo or Yokuts groups of central California.

In summary, the CDM can be utilized in the future in the following areas:

• Bibliographic source: a place to check for references when conducting studies in any local region.
• Forum for working out conflicting ethnogeographic interpretations, using the consensus rules of Wikipedia Commons.
• Model for historic demography studies: separate data columns can be used to build and map regional population densities under alternative assumptions.
• Model for geographical social interaction studies: evidence on endogamous and exogamous marriage patterns can be translated to a map model, and tied to studies of genetic and cultural flow over time.
• Model for the initial colonization of California: algorithms can be built to model "budding off" and migration during the first human entry into California and, by map expansion, over all of North America.

These are just a few of the potential uses of the flexible package of tools that we are initiating in this statewide iteration of the Contact-Period California Community Distribution Model.

What follows, then, are 14 volumes, each one representing an analytical zone, focusing on multi-regional issues including a mutual history, shared language and land-use patterns, and relevant references, followed by anywhere from 13 to 68 regional monographs for each zone (Table 1).

Bibliography

Acsádi, Györgi, and J. Nemeskéri

Adams, John W., and Alice Bee Kasakoff

Bean, Lowell John

Bennyhoff, James A.

Bennyhoff, James A.

Brown, A. K.

Cook, Sherburne F.

Coombs, Gary

D’Azevedo, Warren L.

Gibson, Robert O.

Gifford, Edward W.

Harrington, John P.

Heizer, Robert F.

Henry, Louis

Horne, Stephen P.

Johnson, John R.

King, Chester A.

King, Chester D.

King, Thomas F., and Patricia P. Hickman

Kroeber, Alfred L.

Kunkel, Peter H.

McLendon, Sally, and John R. Johnson

Merriam, C. Hart

Milliken, Randall

Milliken, Randall, and John R. Johnson

Morgado, Martin J.

Priestley, Herbert Ingram

Schenck, W. E.

Strong, William Duncan

Velásquez de la Cadena, Mariano

Wrigley, E. A.

Additional Files

• Appendix A: Coding Manual for the Contact-Period Native California Community Distribution Model
• Appendix B: Process for Producing the Community Distribution Model Regions' Mapping Layers
• Appendix C: Contact-Period Native California Community Distribution Model Zones and Regions