GIS, Internal Colonialism, and the U.S. Bureau of Indian Affairs

Abstract

This article contains the first comprehensive empirical account of the history of geographic information systems (GIS) development within the U.S. Bureau of Indian Affairs (BIA), an account founded in part on a previously unused source of archival data. It also demonstrates the importance of linking a standard technical and institutional history of GIS with a topic neglected in most such histories, the history of the resource application crucial to GIS deployment. The main finding is that across four decades of effort, the BIA's pursuit of GIS is better understood as an effort to perpetuate its internal colonial agenda and its own bureaucratic existence during an era of rapid technological upheaval rather than as a trustee's effort to better manage resources for the greater good of American Indians. The BIA's quest involves various time-honored colonial practices: creating new forms of dependence, imposing complex bureaucratic procedures, misusing funds, distributing free commodities, developing obligatory points of control, and outsourcing both management and labor to a private sector with long experience exploiting Indian resource economies. We conclude that rather than revolutionizing institutions and setting them on new trajectories toward self-improvement as some have suggested, GIS development is often merely a part of a broader and historically consistent pattern of policymaking and behavior.

本文根据一份过去从未曾使用的部份档案资料来源, 首次对美国印地安人事务局 (BIA) 建置地理信息系统 (GIS) 的历史进行综合性的经验解说。本文同时揭露将地理信息系统的标准技术与机构历史连结至一个近乎被历史遗忘的主题— —地理信息系统部属中关键的资源运用历史— —的重要性。研究主要发现, 美国印地安人事务局致力发展地理信息系统将近五十年, 实际上是在快速科技变革的年代中, 为延续内部殖民的议程与维系官僚机构本身所进行的努力, 而非出自信托机构为美国印地安人的福祉进行更佳的资源管理所做的努力。美国印地安人事务局的探索包含了各式历史久远的殖民活动: 创造新的依赖形式、推行复杂的官僚程序、滥用基金、分派免税商品、建立强制控管地点, 并同时将管理和劳动外包给长期剥削印地安资源经济的私部门。结论中, 不同于地理信息系统的发展促成机构革新、将机构推向自我提升的新兴道路之主张, 我们指出, 地理信息系统的发展, 通常纯粹仅是较为广泛且具有历史一致性的政策制定与行为模式中的一部分。

Este artículo contiene el primer recuento detallado en los EE.UU. sobre la aplicación de los sistemas de información geográfica (SIG) por la Oficina de Asuntos Indígenas (BIA, por su sigla en inglés), historia basada en parte en una fuente de datos de archivo apenas ahora utilizada. El artículo demuestra también la importancia de ligar una historia estándar técnica e institucional de los SIG con un tema ignorado en la mayoría de tales estudios, o sea la historia de la aplicación del recurso crucial para el despliegue de los SIG. El hallazgo principal es que a lo largo de cuatro décadas de esfuerzo, la utilización de los SIG por BIA se entiende mejor como el esfuerzo para perpetuar su agenda colonial interna, y su propia existencia burocrática, en una era de rápida agitación tecnológica, más que como esfuerzo administrativo para manejar mejor los recursos en mayor provecho de los indios norteamericanos. Las acciones de la BIA incluyen varias prácticas coloniales de vieja data: crear nuevas formas de dependencia, imponer complejos procedimientos burocráticos, malversación de fondos, distribución de mercaderías gratuitas, desarrollo de puntos de control obligatorio y contratación externa tanto de administración como trabajo con un sector privado de larga experiencia en la explotación de economías de recursos indígenas. Concluimos que más que revolucionar instituciones y reorientarlas hacia nuevas trayectorias por el auto-mejoramiento, como algunos han sugerido, el desarrollo de los SIG es solo parte de un patrón más amplio e históricamente consistente de políticas de acción y conducta.

Key Words:

• American Indians
• Bureau of Indian Affairs
• forests
• geographic information systems
• internal colonialism

关键词:

• 美洲印地安人
• 印地安人事务局
• 森林
• 地理信息系统
• 内部殖民

Palabras clave:

• Indios norteamericanos
• Oficina de Asuntos Indígenas
• bosques
• sistemas de información geográfica
• colonialismo interno

The history of geographic information systems (GIS) and their role and influence in society has been an important research topic in geography and allied disciplines since the early 1990s. In the early years, this research helped clarify and, in most cases, resolve early contentious debates among positivist technoscientists, social theorists, and others (Sheppard et al. 1999; Schuurman 2000). This led to a blossoming of the field as important new understandings accumulated on both empirical and theoretical topics, including digital representations, ethics, geodemographics, and epistemologies (Curry 1995; Goodchild 1995; Goss 1995; Pickles 1995b; Rundstrom 1995; Sheppard 1995; Schuurman 2000); the social construction of GIS (Harvey and Chrisman 1998; Harvey 2000; Chrisman 2005); the development of public participation GIS (Harris and Weiner 1998; Craig, Harris, and Weiner 2002; Sieber 2006); GIS and qualitative research methods (Kwan 2002; Kwan and Knigge 2006); volunteered geographic information (Goodchild 2007; Elwood 2008); the impacts of GIS within institutions (Curry 1998); and the social and technical history of GIS development (Mark et al. 1997; Chrisman 2005; Sheppard 2005; O'Sullivan 2008). This article links these last two areas of inquiry by examining the history of GIS development at the U.S. Bureau of Indian Affairs (BIA), the major national institution responsible for fulfilling the treaty obligations of the United States to the numerous American Indian societies that live within it. ¹

As in other federal government agencies, the BIA has toiled with the independent development of GIS since the early 1970s (Palmer 2007). We provide the first comprehensive empirical account of the history of GIS development within this agency. How did it occur, who was involved, what were its avowed purposes, and what were the consequences? In answering these questions, we join two strands of research in the history of GIS, the technical and the institutional, with a third important element that has been largely neglected in studies of the history of GIS, a history of the resource crucial to GIS deployment. In the case of the BIA, that resource is timber. Linking resource history to GIS history at a federal agency is one important contribution of this research. Another is the uncovering of a previously unused source of archival data by the first author, data that proved vital to the empirical account that follows.

To be clear, the research reported here is narrowly confined to GIS history within one institution, the BIA. It is an agency history, one that we believe is a significant contribution to geographers' efforts to understand the social and bureaucratic implications of GIS. This is not an account of GIS development and implementation throughout Indian Country. ² Such an account has yet to be written but would no doubt include numerous examples of achievements in litigation, land-use change, resource conservation, and other endeavors in which geospatial technologies play an important role in benefiting Indian communities, tribal governments, and the trust lands and allotments on which they are based. Scores of tribes now operate their own GIS offices and gather and control their own geographic data, and many make use of BIA-contracted training, software, and data layers. Important as they are, we do not focus on those endeavors here, nor do we condemn the general use of GIS in Indian Country. Instead, our purpose is to provide an account of the machinations of a federal agency trying to please congressional committee members who control its funding, GIS users in other federal agencies, and, least of all, Indian users as it inched its way into the modern era of geographic data.

Our central argument is that GIS development at the BIA was not a linear march-of-progress episode in technological change as some have suggested but a continuation of internal colonial practices and uneven economic development. Unfortunately, the list of such practices is not a short one: creating dependency by distributing “free” commodities and by establishing obligatory points of control for geographic information; federally dominating reservation economies; imposing complex bureaucratic procedures; denying services based on a federally imposed land tenure system; imposing policy decisions, one of the implications of which was the attempt to complete an epistemological shift in how Indian lands are understood and managed; government and corporate profiteering from environmental resources on Indian lands; and outsourcing labor to a private sector with long experience at exploiting both federal funding and Indian resource economies. In sum, it appears that GIS planning and implementation on behalf of tribal governments marginalized the very people the BIA proclaimed to be helping. ³

The BIA, Corporations, and Tribal Government: A Brief Primer

One reason government agencies are important institutions for studying the history of GIS is that they constitute the bureaucratic context within which GIS can be deployed across very large areas involving great numbers of people. In the case of the United States, numerous federal agencies started to shift to GIS beginning in the 1970s, both independently and in collaboration. The most peculiar agency involved has been the BIA, the most important colonial institution in American Indian affairs since its inception (Champagne 1983). It is also one of the oldest government agencies in the United States. The BIA (originally, the Office of Indian Affairs) was created in 1824 to both “civilize” and assimilate indigenous populations by implementing federal Indian policy beginning with the epic land clearances conducted in the early part of the nineteenth century in the American South. Initially located in the War Department, Indian Affairs was transferred to the newly created Department of the Interior (DOI) in 1849, where it soon became one of the most neglected divisions among a seemingly random aggregation of them in the DOI (Cahill 2011, 9). Since then, the BIA has simultaneously attempted to assimilate American Indians, their land, and natural resources into bureaucratic inventories and maps centrally located in Washington, DC, and it has done so using a support network of Indian agencies, military forts, and Christian missionaries throughout Indian Country. Moreover, the BIA has acted in concert with corporations since at least 1870, when railroad construction through Indian lands was allocated to private corporations instead of to Indian promoters of railroad construction in tribal governments (Miner 1989, 26–27). That timber interests were aligned with those of the railroads at that time is of particular relevance to this article. This process began during a period when the U.S. government simultaneously pursued a ruthless military campaign that, with the aid of both congressional legislation and Indian leaders acculturated to industrialization, would eventually lead to the fairly rapid dispossession of most Indian lands and natural resources. Later, tribal governments developed in the twentieth century were modeled closely on the tripartite separation of powers operating in the U.S. federal government. With executive, legislative, and judicial branches united under a single “chief” or chairperson, most tribal governments comprise nonnative, alien institutions for political decision making and distribution of goods and services. Leaders of tribal governments frequently occupy a position as compromised go-betweens, officials who, once duly elected by tribal members, serve also at the pleasure of the BIA and Congress. Indeed, most such governments were born out of congressional legislation authored in the 1930s by BIA attorneys. Thus, they represent forms of government without much of the authority of a government (Castile 2006, 9). Officials of these comprador governments typically participate in the same circuits of capital as their federal overseers and corporate patrons. Often the result is a textbook example of class conflict wherein a leadership class complicit in the maintenance of dependency on colonial institutions is pitted against a large, broadly colonized lower class. Indian communities are often riddled with the in-fighting and subsequent fraud and corruption that such conflict creates. In sum, Indian Country has long been riven with the governmental and corporate institutions of a settler society pursuing a policy of internal colonization administered by the BIA. Reforming the BIA has been attempted innumerable times to no avail, and a consistently viable alternative to the BIA's colonial brief has never emerged (Dunbar-Ortiz 1979; Snipp 1986b; Deloria [1969] 1988; White 1988; Moore 1993; Wunder 1994, 66–78; Churchill 2003, 114–16; Ostler 2004, 2–5; Hartwick 2009; Cahill 2011, 3). ⁴

Methods and Data

Studying GIS development within private corporations, academia, or government agencies sheds light on the historical, social, and political contexts for technological change. Much of this research has emphasized innovation, diffusion, and implementation of the technology in international projects, universities, and North American government agencies, particularly those at federal cabinet or ministry levels (Chrisman 1988; Tomlinson 1988; Coppock and Rhind 1991; Mark et al. 1997; Curry 1998; Foresman 1998; Greenlee and Guptill 1998; McHaffie 2000; Harvey and Chrisman 2004). Some histories also have been written as pioneer memoirs by individuals personally involved with GIS development at their institution. One of the advantages of such accounts is that writers are able to draw on unique and valuable firsthand experiences to reveal knowledge of internal workings that is often unavailable in archived documents, although such accounts are also made problematic “because insider researchers have a personal stake and substantive emotional investment in the setting” (Brannick and Coghlan 2007, 60). Subsequent conclusions from such accounts might be highly skewed. On the other hand, institutional histories written from the “outside” are rare due to the absence of informative documentation, limitations imposed on access, or both. Where government documents do exist, they “are made and used in accordance with organizational routines, and depend for their intelligibility on shared cultural assumptions” (Hannam 2002, 113). In the case of the BIA, it has long been the lead agency charged with administering a large number of cross-cultural relationships—multinational relationships, many would say. Thus, a reading of BIA documents allows reconstruction of “the colonial order of things as seen through archival productions” (Stoler 2002, 87). The result is a reconstruction of how a colonial institution tries to organize the world through its own representations.

The first author of this article gained access to a unique and previously unexamined set of data for this research, an opportunity that proved crucial to constructing the historical account that follows. He is the first American Indian academic to have unlimited access to the central archives housed at the BIA Geographic Data Services Center (GDSC) at the federal government enclave in Lakewood, Colorado. There, he examined GDSC quarterly technical reports for 1987 through 2005, a Database Organization Guidelines manual (BIA 1997a), enterprise software distribution records, and helpdesk database records queried on request. The technical reports provide detailed summaries of GIS development activities, including many of the people, technologies, and forms of interaction involved at the BIA GDSC. Most of the quarterly technical reports are paper booklets consisting of approximately forty pages of text, figures, and tables. Digital reports for 2003 and 2005 exist on a compact disc. Each quarterly report summarizes the geospatial activities of in-house technical specialists as products fulfilling financial contracts first with the Bureau of Land Management (BLM), then jointly with the BLM and BIA, and finally only with the BIA. In general, the content of

Table 1 Important acronyms and other components of geographic information systems at the U.S. Bureau of Indian Affairs, 1970–2006

Agencies of U.S. Federal Government (in alphabetical order)
DOI	Department of the Interior
BIA	Bureau of Indian Affairs
	Division of Forestry
GDSC	Geographic Data Service Center
	Office of Trust Responsibilities
BLM	Bureau of Land Management
USFWS	U.S. Fish and Wildlife Service
USGS	U.S. Geological Survey
NASA	National Aeronautics and Space Administration
USFS	U.S. Forest Service
Corporations (in alphabetical order)
ACS	Affiliated Computer Services
Autometrics, Inc.
Battelle Corporation
Boeing Computer Services
CDSI	Computer Data Systems, Inc.
ESRI	Environmental Systems Research Institute
Lockheed-Martin Corporation
North American Rockwell
Raytheon Company
TGS	Technicolor Government Services
Federal legislation
PL 638	The Indian Self-Determination and Education Assistance Act (1975)
	The National Indian Forest Resource Management Act (1990)
GIS-related hardware, software, initiatives, and agreements (in alphabetical order)
ARC/INFO
BND	BIA Nationwide Database
ELA	Enterprise Licensing Agreement
IIRIP	Indian Integrated Resource Information Program
IRMP	Integrated Resource Management Plan
MOSS	Map Overlay and Statistical System
NRIS	Natural Resource Information System
TRIS	Timber Resource Information System
Congressional Advisory Panel of Academic and Professional Forestry Experts
IFMAT	Indian Forest Management Assessment Team
Other political organizations
ITC	Intertribal Timber Council

the reports includes summaries of data entry projects, GIS database creation and maintenance, GIS user support and training, technical assistance, remote sensing operations, and geospatial applications. Details of financial contracts are often omitted. Later reports also feature enterprise licensing agreement data. All of the reports, their content, agency affiliations, in-house technical specialists, and applications signify the BIA GDSC's position as the geospatial service center for Indian Country. ⁵

These sources indicate that GIS development at the BIA involved only the numerous government agencies charged with resource management and longtime federal contractors specializing in advanced technologies and military hardware and software. Table 1 presents the federal, corporate, and other components of GIS development at the BIA between 1970 and 2006, along with the plague of acronyms used in our account. The most important agencies and corporations are the BIA Division of Forestry, BIA Office of Trust Responsibilities, BIA area and agency offices, BIA forestry coordinators, BIA agency foresters, BIA GIS coordinators, U.S. Geological Survey (USGS), U.S. Fish and Wildlife Service (USFWS), the BLM, the Environmental Systems Research Institute Corporation (ESRI), the Battelle Corporation, Technicolor Government Services, Inc. (TGS), Computer Data Systems Inc. (CDSI), Affiliated Computer Services (ACS), and the Lockheed-Martin Corporation. We identify three eras: resisting self-determination while experimenting with GIS in the 1970s; implementation and expansion in the 1980s; and increasing privatization in the 1990s and 2000s. The technical and bureaucratic history of each era is summarized, along with the gateway resource used for GIS development, forests.

Three Eras of GIS at the BIA

Resisting Self-Determination While Experimenting with GIS in the 1970s

The BIA and its predecessor, the Office of Indian Affairs, took responsibility for forest management on American Indian land for most of the twentieth century. Until federal forest sustainability programs entered reservations in the 1970s, the BIA's mission was to help off-reservation commercial timber companies get signed timber leases on both tribal lands and individual Indian allotments, to assist logging operations, and, supposedly, to create jobs for Indians (Newell, Clow, and Ellis 1986; Smith 1989). This approach did not change appreciably even in the 1970s, despite three significant developments in law, research, and politics.

Major federal legislation in 1975 was supposed to have set the BIA on a different course. Among other features, Public Law (PL) 638, The Indian Self-Determination and Education Assistance Act of 1975, required that Indians receive preference in training and employment associated with federal programs designed to benefit Indians. Indian-owned economic enterprises are also to be favored by federal agencies awarding subcontracts and subgrants for work done in Indian Country.

As PL 638 was going into effect, policy analysts funded by the Ford Foundation were writing the results of a study of the federal welfare system and sector-by-sector employment opportunities in Indian Country, focusing on lands held in trust or allotted within approximately 310 federal reservations (Levitan and Johnston 1975). Unsurprisingly, they found reservation economies and social structures “totally dominated by the federal presence” (Levitan and Johnston 1975, 9). Widely cited for many years thereafter, the report presented evidence that what little economic activity there was overwhelmingly benefited private firms extracting resources on land leases controlled by and negotiated with the BIA. For example, Northwest timber cut from Indian lands was hauled to distant locations for milling. Non-Indian managers at those sites controlled and managed the resource and wages and profits accumulated to non-Indians (Levitan and Johnston 1975, 26). Getting an “Indian contract” from the BIA simply presented an opportunity for a firm to increase profit margins at the expense of ecosystems and economies located in Indian Country. Moreover, even if firms actually made their lease payments, and sometimes they did not, lease monies were often redirected or “lost” to fraud. Such were the time-honored practices involving BIA leases (Cahn 1969, 9, 52, 149).

In the 1970s, with 5.5 million acres of commercial-grade reservation timber available, only fourteen reservations accounted for 96 percent of all the revenue generated for Indians. Just two dozen sawmills operated on or near reservations, most of them located in the less productive regions of the Southwest and Midwest. No mills at all operated on or near Indian land in the larger and more productive forests of the Northwest (Levitan and Johnston 1975, 20–26; Snipp 1986a, 467). Among numerous recommendations, the Ford-funded report called on federal agencies to assist development on timber-rich reservations with the greatest untapped potential for resource development, to use BIA funds to hire technical advisors to work on the reservations to develop expertise there that would be vested in tribal interests, and to accelerate inventories of forest resources in the interests of “building Indian Nations” (Levitan and Johnston 1975, 78–80).

In 1976, one year after the report's release, five Indian leaders with concerns about resource mismanagement in the Northwest created a new political organization, the Intertribal Timber Council (ITC), to prod the BIA Division of Forestry into serving Indians better by helping develop more productive forestry practices on reservation lands, including assistance with creating tribal databases of forest-related information. A recalcitrant, inert BIA would need almost two decades of such prodding to put modest changes into effect, and only after being compelled to do so by more congressional legislation achieved in part through ITC efforts.

That the consequences of PL 638, the Ford Foundation report, and the formation of the ITC failed to motivate change at the BIA is perhaps unsurprising given the glacial pace of bureaucratic change in general and the agency's colonial history in particular. Indeed, until the middle of the 1990s, the BIA dug in, reinforced its insularity, and resisted both recommendations and legal requirements emanating from outside the bureau that were designed to promote Indian self-determination in the face of widely acknowledged failures in previous federal policy.

This resistance was particularly noteworthy in the practices of the GDSC (Marchand and Winchell 1992, 177). The archived quarterly reports and related memoranda at the GDSC from this period reveal a striking absence of correspondence and negotiations with Indian representatives of any sort, let alone the members of tribal governments. Thus, the bureaucrats and their corporate and academic patrons were laying the foundation for GIS development in Indian Country with little or no input from the Indian communities or tribal governments whose undeveloped environmental resources were being targeted.

Precursors or prototypes of what we now recognize as modern GIS began to be of interest to the BIA in the late 1960s and early 1970s. These early efforts concentrated on the interpretation of high-altitude aerial photography and 1960s-era satellite imagery for classifying forest cover and slope and had little statistical or other computational capability compared to today's systems (Boeing Computer Services, Inc. 1972a, 1972b). The National Aeronautics and Space Administration (NASA) was often the lead sponsor because the projects were aimed at developing reliable techniques for analyzing the wealth of new spatial data beginning to be widely available via satellite. Targeted tribal forests included those on the Quinault, Warm Springs, and Colville reservations in the Northwest and on the San Carlos and Fort Apache reservations in the Southwest. Typically, the BIA and BLM were teamed with large corporations in the vanguard of technological change who secured contracts to conduct feasibility studies and assessments of agency needs and to then develop systems for testing and eventual implementation.

Contractors for the development of geospatial technologies were, then as now, giants in aerospace, computing, and weapons industries. North American Rockwell Corporation produced the first feasibility study for a resource management information system for the BIA in 1970 (Boeing Computer Services, Inc. 1972b, 1). In April 1972, Boeing Computer Services, a subsidiary of the Boeing Corporation, concluded a twelve-month BIA/BLM contract with another more detailed study regarding the development of a comprehensive Natural Resource Information System (NRIS) for all land-related resources under the control of DOI agencies. Its purpose was to provide natural resource managers in every DOI agency with a universal, integrated system for manipulating and analyzing large volumes of map and satellite data to achieve their widely divergent goals (Boeing Computer Services, Inc. 1972a, 1972b). Between 1972 and 1975, the Raytheon Company made NRIS operational. This was done partly through a demonstration project involving forests on the Quinault Reservation and adjacent lands in Washington state. Termed the Timber Resources Information System (TRIS), this latter project also involved the forest service agency of the State of Washington and the U.S. Forest Service (BIA 1972, 1974). The Quinault and Queets Indians on the reservation and their tribal government were involved only in providing ground-truth samples for the satellite imagery (BIA 1973). Proprietary issues associated with Raytheon's control of the software coupled with the prohibitively expensive cost of running NRIS on DOI computers lead to implementation troubles, however. An effort at Washington State University to revise the software failed in 1975, resulting in abandonment in early 1976 (USGS 1978, 5, 9–11).

Resource extraction issues also were gaining attention at other federal agencies in the middle of the 1970s, notably the USFWS. New federal legislation tasked the USFWS with developing habitat mitigation measures for lands subjected to destructive resource extraction. To do so on BIA-managed lands, the USFWS collaborated with the BIA Northwest Area Office, in particular the Branch of Forest Resources and Planning in Portland, Oregon (Smith 1989). This collaboration unveiled the first modern GIS innovation to have an impact on the BIA, the Map Overlay and Statistical System (MOSS). Developed by scientists under contract to a nonprofit organization tasked to help the USFWS complete its new assignment, MOSS was the first open-source, interactive, vector-based GIS and the first product used by the BIA in the modern era of GIS. It was developed in 1977–1978 in Fort Collins, Colorado, first tested in 1978 for a USFWS habitat mitigation project, and declared ready for production use in 1980. Both the BIA and BLM adopted it soon thereafter. Later, it would undergo another iteration when combined with wetlands inventory software developed by Autometric, Inc. MOSS was developed to serve the needs of credentialed wildlife biologists, natural resource planners, and other agency professionals involved in defining wildlife habitats and mitigating the effects of resource extraction on those habitats. Specifically, MOSS was designed for the collection and analysis of environmental data associated with Western coal lands subjected to strip mining (BLM 1987; Greenlee and Guptill 1998; Reed 2004).

As with earlier prototypes like NRIS, the significance of MOSS for GIS history at the BIA lies not in its actual use but in the way it was adopted and deployed. MOSS was a significant technical innovation developed to help federal bureaucracies comply with new federal environmental law. Like other efforts that preceded it, MOSS was intended to assist federal environmental agencies in their mission of monitoring and managing resource-rich ecosystems under their supervision (Palmer 2012). But neither its design nor the institutional setting in which it was conceived were intended to incorporate the needs peculiar to lands and lives in Indian Country. MOSS simply represented a way forward for BIA bureaucrats looking to move into the new world of GIS. That way forward supported and perpetuated their own occupational existence more than it supported Indian economic development and self-determination. Subsequent developments in GIS at the GDSC would follow a similar path.

Implementation and Expansion in the 1980s

Although the three crucial developments in the middle of the 1970s—passage of PL 638, the Ford Foundation report, and the formation of the ITC—did not induce immediate change at the BIA, they did lead to small increases in congressional appropriations to the BIA's forestry division and to tribal governments for forest management programs in the 1980s (Rigdon 2007, 108). During this same period, early forest and timber inventory data were developed for multiple reservation areas, as shown in Table 2 (BIA 1991). Subsequently, marginal increases in timber production did occur in both the Northwest and Southwest, yet production still remained low relative to potential. By 1987, only ten reservations across the United States were generating annual timber revenues over $1 million. The Northwest region held six of those ten reservations where revenues from timber harvests totaled $48 million. The Yakama Reservation alone accounted for $20 million (BIA 1988a). Throughout the 1980s, the reservations in the Northwest accounted for 65 percent of total board-feet of commercially viable timber and 55 to 75 percent of the total timber harvested in Indian Country. Unsurprisingly, the vast majority of the revenue generated from timber sales derived from the Northwest, too. Some reservations in the Southwest were also very important local producers, but they were secondary on a national scale to those in the Northwest in all three categories (Indian Forest Management Assessment Team [IFMAT] 1993, IV–1,4). For example, the White Mountain Apache on the Fort Apache Reservation produced $11 million in revenue in 1987 (BIA 1988a). The Northwest and Southwest BIA regions have led in forest productivity ever since (IFMAT 2003).

Table 2 Forest and timber inventory geographic information system data created by the Bureau of Indian Affairs Geographic Data Services Center in the 1980s for select reservations

Reservation	State (for location only)	Forest/timber GIS data
Pine Ridge	SD	Forest study area boundary; timber types
Acoma	NM	Timber compartment boundary; timber sale units; timber types
Isleta	NM	Precommercial thinning; timber compartment boundary; timber sale units; timber types
Jemez	NM	Precommercial thinning; timber sale units
Jicarilla	NM	Precommercial thinning; timber compartment boundary; timber sale units; timber types
Mescalero	NM	Continuous forest inventory; timber types
Nambé	NM	Timber types
Northern Pueblos	NM	Resource value—forest
Picuris	NM	Precommercial thinning; timber sale units; timber types
Pojoaque	NM	Timber types
San Ildefonso	NM	Timber sale units; timber types
San Juan	NM	Timber types
Santa Clara	NM	Precommercial thinning; timber compartment boundary; timber sale units; timber types
Southern Ute	CO	Timber types
Taos	NM	Timber types
Tesuque	NM	Timber types
Zuni	NM	Timber stand improvements; timber types
Alabama-Coushatta	TX	Timber types
Blackfeet	MT	Forest habitat; forest suitability; timber compartment boundary; timber types
Crow	MT	Timber types
Fort Belknap	MT	Forest habitat; forest management; timber types
Northern Cheyenne	MT	Forest habitat; forest management units; logging units; timber thinning; timber compartment boundary; timber types; tree planting areas
Rocky Boys	MT	Timber types
Wind River	WY	Forest management units; proposed timber management; proposed timber transportation; timber compartment boundary; timber types
Menominee	WI	Continuous forest inventory; timber blocks; timber compartment boundary
Fort Apache	AZ	Continuous forest inventory; continuous woodland inventory; thinning units; timber compartment boundary; timber cut blocks; timber sale units; timber types
Hualapai	AZ	Continuous forest inventory; timber compartment boundary; timber types
San Carlos	AZ	Timber sale units; timber types
Uintah and Ouray	UT	Timber types
Colville	WA	Continuous forest inventory; timber compartment boundary
Flathead	MT	Continuous forest inventory; even age cut blocks; forest habitat; forest management system; forest photo points; forest road compartments; forest road units; lodge pole pine stands; logging units; timber yypes
Kalispell	WA	Continuous forest inventory
Spokane	WA	Continuous forest inventory; forest habitat; timber compartment boundary
Umatilla	OR	Forest road structures; forest roads; timber types
Warm Springs	OR	Continuous forest inventory; forest management units; forest use areas; timber types; total resource inventory
Yakima	WA	Continuous forest inventory; logging units; precommercial thinning; timber types
Hoopa Valley	CA	Timber compartment boundary; timber cut types; timber types
Source: Bureau of Indian Affairs (1992).

But were the congressional allocations, small as they were, used more effectively by tribal governments or by the BIA? The authors of an econometric study of seventy-five timber-producing tribes conducted in the early 1990s found that tribal government officials were usually better at managing, harvesting, and marketing timber because they had incentives the BIA did not, including an interest in the well-being of both the forests they lived in and the societies of which they were a part (Krepps and Caves 1994). In practice, BIA managers were actually restricting economic development by earmarking select special projects for a small number of targeted forests, rather than enabling Indian forest managers to work in response to their own tribal governments as PL 638 required. Ironically, such colonial hoarding of expertise by the BIA, and the gross inefficiencies they so often produced in the past, have often netted the BIA important budget increases from Congress (Krepps and Caves 1994, 134). Thus, the BIA could actually be more fairly characterized as having a disincentive when it comes to improving chances for beneficial Indian economic development. This appears closely related to Deloria's claim about the essential conundrum of the BIA: As long as it is expected to help redevelop Indian self-sufficiency, it will fail in its programs (Deloria [1969] 1988, 136).

How did these budget increases influence the labor market? Although Indian Country timber industry jobs for Indians increased during the 1980s, they were almost entirely low-wage positions paying on average 58 percent less than jobs that went to non-Indians working in the same trust lands. Moreover, the BIA took money for its own internal development that should have gone directly to either tribal governments or Indian-owned timber operations. Consequently, full-time staff at the BIA Natural Resources Division swelled to 575 employees, 96 percent of whom were in forestry (Smith 1989, 89). In short, even after the legal and political developments of the 1970s, BIA interests remained fundamentally tied as always to perpetuation of the agency itself. Its agenda for maximizing timber harvests was being used to maximize agency staff levels (Smith 1989).

Concomitant with its growth in budget and staff, interest in GIS development accelerated at the BIA's forestry division. Officials there pushed for the appointment of GIS coordinators at each BIA area office and for a national GIS director for Indian Country to be located at GDSC offices in Colorado (BIA 1989; Smith 1989). Thus, a lot of the congressional allocation in the 1980s, $9.7 million to be exact, went toward the development of what was termed the Indian Integrated Resource Information Program (IIRIP) and for the maintenance at the GDSC of the geographic information it consumed (BIA 2005b).

The IIRIP began in 1983 and eventually incorporated spatial data from just ten Indian societies with reservation lands (Marchand and Winchell 1992). IIRIP goals included creation of a digital database called the BIA Nationwide Database (BND), implementation of computer hardware and software centralized and controlled by the newly appointed program manager at the GDSC, and the establishment of GIS expertise at some BIA field offices (BIA 1988b). The IIRIP used MOSS, later switching to ESRI's ARC/INFO software (Marchand and Winchell 1992, 175–76, 178). The expenditures and effort focused primarily on forest resources on reservations like the Colville, yet centralization of all information and activity was retained at the GDSC. Local tribal members could neither get full access nor control of the data and applications once implementation had begun. Moreover, with GIS functions under federal control and focused almost entirely on forestry, tribal members were hard-pressed to widen the scope of applications beyond forest resources alone (Marchand and Winchell 1992, 176–78).

These were not the only limitations to self-determination that Indians faced in the GIS arena of the 1980s. BIA officials conceived of IIRIP as a comprehensive program for all of Indian Country and saw individual, reservation-specific Integrated Resource Management Plans (IRMPs) as the key components. These plans were designed for accelerating resource extraction from Indian reservations and allotments. All IRMPs were required to use the BIA's GIS. With continuation of the agency precariously dependent on often noncommittal and always-stingy federal funding, the notion of a GIS assembled from detailed geographical information gathered across all Indian lands was very attractive at the BIA, for it would assist the agency in its bureaucratic charge of managing federal programs on Indian lands. Standardization of resource information was deemed crucial to speeding its use at the agency, never mind the separate needs of the various Indian societies brought into the system. To achieve this goal, the BIA imposed a single required format and content for all IRMPs, with the BIA's national GIS coordinator overseeing this standardization and ensuring compatibility with IIRIP. In Section 1.3 of the memorandum outlining the requirements, the BIA also arrogated to itself the authority to (dis)approve these tribal plans and related activities, with or without the approval of tribal governing bodies (BIA 1990c):

The Agency Superintendent shall strive to secure plan approval by the tribal governing body. Should tribal approval not be obtained after a prudent period of time and effort, or after consideration of written objections to the plan, the Area Director may approve the plan to protect valuable resources. In effect, this will be exercising the Bureau's trust responsibility. … Once the plan has been approved, the Bureau shall support only those resource related activities which are within the scope of the plan's preferred alternatives.

By the middle of the 1990s, an organization of experienced forestry personnel assembled to assess BIA relations with resource-rich tribal governments declared the IRMP/IIRIP a failure (Gordon et al. 1997, 11–12). Only nine IRMPs had been completed by 1994 for the ostensibly nationwide database. Oddly enough, the program that was initially so attractive to agency officials failed, in part, due to the BIA's own bureaucratic inertia, which had rendered the project largely unfunded, unused, and a low priority within the agency. Officials of the few tribal governments that completed an IRMP were also dismayed about the difficulty they were having getting access to the results. In 1990, for example, “participation in the development of the Fort Apache IRMP was terminated … the [BIA's Fort Apache] agency is continuing the process without tribal participation” (BIA 1990c, 18). The White Mountain Apache on the Fort Apache Reservation had decided that they could manage their resources better themselves.

Despite these failures, the BIA pressed on with the development of its general BND by assigning technicians to continue to digitize the relevant maps in the 1:24,000 USGS topographic series. In retrospect, it seems that this was the most lasting technological component and data archive in 1980s-era GIS at the BIA. To produce it, the physical environment was disintegrated into the customarily piecemeal, analytical epistemological categories associated with GIS themes, layers, and applications (Rundstrom 1995). The maps yielded locations of buildings, campgrounds, pipelines and transmission lines, roads, wells, tanks, streams, lakes, ponds, reservoirs, springs, boundaries of states, reservations, and counties, surveyed townships and sections of the U.S. Public Land Survey, and areas where land tenure was designated, particularly those places managed by other federal or state agencies. This was done for approximately 151 reservations distributed across all BIA regions (BIA 1992; Palmer 2009). Although less than half of the reservations in the United States received this kind of coverage—to say nothing of other Indian Country jurisdictions, the status of some of which remain under dispute—the scattering across the country gave the impression that the BND was indeed nationwide.

One region excluded from GIS data development was the trust lands and highly fractionated individual land allotments characteristic of Oklahoma. There, the complex administration of allotment land hampers efforts at modern integrated forest management and GIS database development. In Oklahoma, the heterolocal nature of myriad small, individual landholdings intensifies fragmentation of what are often already scant forest resources, induces large per acre administrative costs, and leads to disagreements among multiple Indian land owners, the BIA, and other agencies (BIA 1990a, 1990b, 1990d, 1990e; IFMAT 2003).

For most lands referenced in the BND, separate, specialized layers were created solely for BIA managerial purposes, including those for timber, range, water resources, minerals and energy, fish and wildlife, land tenure, infrastructure and transportation, and an ambiguously named archaeology and culture layer (Palmer 2012). Because the ubiquitous USGS topographic series was used for the base, it could be that very few proprietary data concerning specific sites of especially high spiritual value or food- and medicine-gathering areas were included. Unrelated secular information located at such sites would easily be made part of the database.

Subsequent development of GIS applications for these data emphasized, for example, land use and land cover classifications in the Midwest and Northwest regions, range management applications in association with the BLM, wildlife habitat projects when working with the USFWS, and forest management and timber mapping applications in concert with the needs of the USFS (BLM 1989; BIA 1990a, 1990b, 1990d, 1990e). In essence, “Indian Integrated Resources” was defined in the database simply as digitized information from the universally available federal topographic sheets. This disintegrated information was pressed into service to meet the needs of myriad federal agencies whose responsibilities divide along strikingly similar lines far more than the needs of particular Indian societies for integrated data of their own lands and waters.

Most of the work on the BND was done by private contractors, a labor trend gradually sweeping through all federal agencies at the time. Software development, technical expertise, tedious data entry tasks, and the public relations work thought necessary to perpetuate the GDSC were gradually shifted to the private sector. These moves helped stabilize the BIA's GIS efforts for a time in the 1980s and 1990s while cosourcing and outsourcing Indian economic development with federal funds designated for GIS development. Three large technical contractors were involved: ESRI, TGS, and Battelle.

The BIA transitioned from MOSS to ESRI's software in 1988. According to the BIA memo of 20 July 1988, “the implementation of ARC/INFO within the BIA is underway. … No version of MOSS will be distributed from this office after September 30, 1988” (BIA 1988c). The major disadvantages of the MOSS software were its limited ability to process raster data and limited updates to the overall software package. Another limitation was its reliance on specific computer hardware. According to the corporation to whom data entry had been outsourced, “transportability of this software to more modern computer hardware technology [proved] to be a substantial task” (Battelle 1988, 13). Reintegration or “combinability” of the disintegrated geographic information in MOSS, both within BIA operations and in connection with other federal agencies, was prohibitively difficult. In contrast, ARC/INFO was compatible with the BIA's new hardware and, most important, the ESRI Corporation provided the BIA with the technical assistance to implement that software and to develop standardized GIS applications, as it continues to do so today. ESRI then entered into a DOI-wide contract for ARC/INFO in the early 1990s.

If ESRI employees actually assumed on-site work responsibilities inside the BIA's Lakewood, Colorado, offices, the archival sources are silent about it. But the documents indicate that the employees of another large contractor did. TGS workers conducted data entry work for the BND. Located in Sioux Falls, South Dakota, TGS employed agricultural specialists, geologists, hydrologists, and agronomists to support technical services for the USGS Earth Resources Observation Systems (EROS) data center located in that city (BLM 1986, 1987). As the BIA's new primary contractor for GIS technical support, TGS visibility was high in the 1986–1987 quarterly technical reports (BLM 1986, 1987). According to these reports, TGS staff began full-time support for the BIA IIRIP in 1987 and were hired to provide on-site technical capability at the BIA. TGS performed such tasks as land cover mapping, data processing, digitizing, and GIS technical support for area and agency offices, too.

The Battelle Corporation was another source of influence on the colonial direction of GIS development beginning in this period. ⁶ Their service for the BIA included a cost–benefit analysis supporting and justifying the GDSC's GIS position not only to those in Indian Country but also to those higher in the DOI administrative hierarchy, particularly those tasked with monitoring program spending and success (Battelle 1988). Battelle justified perpetuating the GDSC largely on economic grounds. Claims about office efficiencies, personnel ceilings, budget reductions, and the development and maintenance of the BND were all important in defense of the GDSC (BIA 1988b). In this way, revenue earmarked for GIS development was actually spent on establishing, maintaining, and justifying new expansion at the GDSC, thereby removing investment still further from Indian Country itself. And what of the forests? Battelle proposed using the profits from increased sales of reservation timber to recover most or all of the cost of developing the GIS used to manage them (Battelle 1988, 14).

Increasing Privatization in the 1990s and 2000s

Until 1990, the ITC was having little apparent effect on the BIA. Eventually, its members succeeded in spurring Congress to pass the National Indian Forest Resource Management Act of 1990, which required the DOI Secretary to get a report assessing both the condition of Indian Country forests and the problems in BIA–Indian relations. The secretary contracted back to ITC, which appointed seven outside experts, a majority of whom were university academics with forestry expertise. All seven had long experience in the forests of the Pacific Northwest. This IFMAT filed its first report in 1993. As expected, the IFMAT report was highly critical of the BIA, particularly the absence of any sense of what trust responsibility meant for managing Indian forests, its indifference to tribal governments when communicating and coordinating its own activities, the poor quality of its technical assistance, and its complete inability to assess its own activities impartially (IFMAT 1993; Gordon et al. 1997). Even after the report was published, the BIA forestry division continued to operate without any independent oversight because the BIA insisted on monitoring itself (Gordon et al. 1997).

For myriad reasons, the dynamics of GIS development at the BIA changed during the middle of the 1990s. Federal Indian policy shifted when Congress decided to reduce BIA funding and ostensibly redistribute resources to tribal governments directly. In 1994, Assistant Secretary for Indian Affairs Ada Deer announced a “shift of resources from the BIA to Indian tribes” concomitant with the Clinton administration's emphasis on Indian self-determination and government-to-government partnerships (DOI 1994). Consequently, the GDSC's budget declined a whopping 65 percent in 1995, a move that devastated operations, resulted in many layoffs of technical and data entry specialists, and led to the decline of the BND and IIRIP programs (BIA 2005b).

These changes significantly diminished the GDSC's power as a data development and management center. Instead, it began to shift into a new role as product distributor akin to a mid-level private wholesaler, only in this case it was distributing GIS software products freely to hundreds of tribal governments and all of the BIA field offices. These were products channeled through the BIA from corporate producers, products on which individual GIS offices based in many tribal governments already had become dependent. Often staffed with only one or two people on marginal budgets, GIS departments of tribal governments now quickly developed another kind of dependency on the labor of employees of ESRI, Battelle, and other contractors, not that of employees working directly for the BIA. By the late 1990s, tribal governments were receiving a lot “less assistance for forest inventory [including GIS database development], management planning, marketing, and economics” directly from the BIA (IFMAT 2003, 10). The rationale was clear. “Partitioning of BIA budgets to individual tribes under self-determination and constant or declining budgets for technical services have strained the capacity of the BIA to provide a critical mass of technical service capacity” (IFMAT 2003, 23).

The White Mountain Apache weathered the storm better than most, able as they were to pursue development independently. A decade-long study published by Stephen Cornell and an associate, codirectors of the acclaimed Harvard Project of American Indian Economic Development, found that the White Mountain Apache timber operation on the Fort Apache Reservation had been one of the most productive timber operations in the Western United States, one regularly outperforming private operators. It might seem reasonable to think that the GDSC played a significant role in their success because, as noted earlier, Fort Apache forest layers had been digitized from the topographic maps by the GDSC. However, the successful revenue generator that the ponderosa pine forests became for the White Mountain Apache was based primarily on their own internal development of a competent bureaucracy controlling the use of the resource, one that matched traditional, culturally prescribed decision-making practices with the current structure of tribal governance. This is an ideal to which few Indian communities can aspire. There are not many with marketable resources whose decision-making institutions survived colonialism and excessive political fragmentation sufficiently enough to make this sort of achievement likely. Thus, the circumstances for success at Fort Apache were seldom replicable, particularly where the BIA was managing the resource (Cornell and Kalt 1998). ⁷

Although some support for GIS deployment to tribal governments continued at GDSC offices between 1996 and 2002, according to the 1997 and 2003 GDSC quarterly reports, two new technical firms, CDSI and ACS, assumed majority control of both development and management of GIS until 2004. Headquartered in Rockville, Maryland, CDSI conducted 90 percent of its computer service business with the federal government during the 1990s and employed twenty-one people inside the offices of the GDSC in 1997 (BIA 1997b). A subsidiary of the Xerox Corporation, ACS operates in more than one hundred different countries with its worldwide headquarters in Dallas, Texas. ACS specializes in communication and information technology services, and it staffed the GDSC with approximately twenty technicians and managers in 2001 (BIA 2001). After 2004, majority control shifted yet again. This time, aerospace leader Lockheed-Martin Corporation assumed responsibility for the BIA's GIS operation (BIA 2005c). ⁸

In December 2002, the DOI and ESRI also signed a three-year contract for the purchase of GIS software to be distributed to tribal governments and BIA field offices. The program, an enterprise licensing agreement (ELA), grew out of the belief within the executive branch that GIS was the “glue for government management in the future” more than any other e-government initiative (Sinha 2003). According to the GDSC's first-quarter technical report for 2003, the ELA was an agreement between ESRI and the DOI for the blanket purchase of GIS software to serve all DOI bureaus. Now able to provide tax-subsidized GIS software to American Indian tribal governments that met ELA requirements, the GDSC found its position strengthened as the legal distributor of the requisite software licenses, technical assistance, and GIS training without which very few tribal governments could successfully implement GIS. By early 2006, the GDSC had distributed GIS software to 232 tribal governments under the ELA (BIA 2005a).

To the unwary, an ELA can seem a fairly straightforward contract between two parties, but the reality is more complicated, and notably so in Indian Country. After 2002, the numerous but greatly dispersed lands and officials in Indian Country became far more interconnected with the Office of the President, the Office of Homeland Security, the Pentagon, the Office of Management and Budget, and other federal agencies with their own ESRI ELAs and with still other technology corporations, computer hardware and software products, and even state government agencies with whom tribal governments very often contend bitterly.

The implied seamlessness among multiple GIS-using agencies can have other potentially disturbing consequences as well. For example, different DOI agencies take the administrative lead in ELA-related projects, with 1 percent administrative fees being charged to the other department agencies that participate. As a DOI bureau, the BIA is also required by the ESRI ELA to provide its own technical support for software applications (ESRI 2008, 25, 27). Thus, under the ESRI contract, the BIA became legally obligated to use public monies earmarked for execution of their tribal trust responsibilities both to pay the administrative costs incurred by other agencies and to pay large federal contractors like Lockheed-Martin. ⁹

By 2006, then, cosourcing of treaty-required government functions had completely shifted to fully outsourced privatization despite the location of the work in government offices. Even the GDSC technical reports written during this period acquired the appearance of a private firm's annual report to stockholders. Report titles are relegated to the bottom of the cover, BIA contact information is virtually invisible, and corporate logos stand out prominently on the front (BIA 1997b, 2001, 2003).

Discussion and Conclusion

The status of the BIA as the epitome of an internal colonial agency is widely understood, yet when the prospect of significant technological change arrived beginning in the 1970s, an alternative narrative developed. GIS at the BIA has often been portrayed as an episode in a triumphal march of progress in resource management. Optimum solutions to both real and perceived environmental problems in Indian Country would be made possible through adoption of the new technology. Such technophiliac narratives have been deployed by those preparing internal documents at the agency (e.g., Battelle 1988; BIA 1988b) and by both Indian and non-Indian BIA employees and private contractors writing for industrial as well as academic consumption (Marozas 1991; Goes In Center 2000; Bohnenstiehl and Tuwaletstiwa 2001; Seagle and Bagwell 2001; Bond 2002). On occasion, the narrative might also include a sales pitch for geospatial technologies, wherein Indians are seen as a “market for the industry to reach here, from sales of satellite imagery to the production of digital orthophotos in support of GIS databases” (Bohnenstiehl and Tuwaletstiwa 2001, 134).

BIA officials of the highest rank have not been immune to the progressive narrative either. By the late 1990s, the Assistant Secretary of the Interior for Indian Affairs was excited with anticipation of a comprehensive GIS-managed set of databases in “this huge computer” as an essential part of an “integrated resources management system” (Gover 1999/2000, 225). “It will look very much,” he said, “like the trust function in a private bank with, of course, the exception that the Bureau will continue not to charge for that service” (Gover 1999/2000, 225–26).

The secretary probably did not foresee that analogies to banks would become unpopular as widespread knowledge of corruption in private banking institutions soon surpassed even that of the BIA's century-long mismanagement of resource-related trust funds. ¹⁰ More important, his stated goal of using GIS to improve the resource management obligations in the BIA's portfolio of trust responsibilities was what BIA officials already had been claiming for almost two decades. Our research suggests that the idea of such improvements was secondary to the mission of perpetuating the agency and the colonial agenda it is assigned within the U.S. government. GIS deployment in Indian Country proceeded by following an internal colonial model featuring misuse of funds, top-down policy impositions without consultation, and maintaining Indian dependency by repeated efforts to sustain the agency's position as an obligatory point of control. The BIA did so by centering software distribution and training opportunities at a single geographic location in the federal enclave in Colorado and by cosourcing and outsourcing labor to a private sector that has availed itself of both Indian resource economies and the federal appropriations necessary for their infiltration across a long span of time.

As a lower-rung agency perpetually striving to sustain its relevance, particularly to the members of the congressional committees who oversee its operations (Deloria [1969] 1988, 132), the woefully underfunded BIA continues to be a very small tail wagged by the far larger dog of the federal–corporate nexus. The BIA, per se, now has very little ability to control its GIS destiny because Indian lands and resources are now a small part of the very large business portfolios of some of the largest military contractors in the world. Indeed, the lead scientist of the Integrated Systems Division at Computer Sciences Corporation observed that with regard to indigenous peoples, “the Federal Government [is] acting as forward scouts for exploitative industry” (Madsen 1995, 223).

Yet none of this was likely the result of intentional behavior by specific individuals so much as it was due to the institutionalized habitual practices of a peculiar and profoundly inert bureaucracy. Thus, historically consistent structural patterns of policy and behavior within a given institution might often exert a molding influence on the manner in which GIS will be adopted in helping that institution perpetuate itself. Rather than revolutionizing institutions and setting them on new trajectories into uncharted territory, technological change can often occur in a manner remarkably consistent with institutional history.

We have also tried to show that although technical and institutional history should be conjoined in the study of GIS, resource history comprises a crucial third element. For example, how would our account have differed if we did not include at least a partial history of Indian forests? First, we would not have been able to explain the regional bias in initial GIS development, nor why the BIA focused on the Northwest in particular. Second, we would have only partially understood the complaints some tribal representatives were lodging about the confining limitations the BIA had placed on their own use of GIS. Third, the White Mountain Apache counterexample would not have come to our attention at all. Finally, we would not have discovered how the perceived underdevelopment of a specific resource became the gateway for GIS development. Thus, the technical history of GIS and its place in institutions might be misunderstood if the history of the related resource—or, more generally, the targeted GIS application—is omitted. Such an omission limits our understanding of technological change and the manner in which institutions cope with it. It could even lead to a dangerously simplistic narrative outlining a progressive and triumphal linear march of progress.

Yet the story of GIS implementation in Indian Country remains far from complete. The portion presented here focuses solely on the institutionalization of GIS in a single government agency. We omit entirely at least three other parts to the story: the on-the-ground interactions among individual tribal governments, the BIA, and private contractors; the responses to GIS development and use emanating from local Indian residents themselves; and the details of GIS history inside tribal government offices that operated independently from the top-down impositions by the BIA and its contractors. All three are matters for future research.

Acknowledgments

We thank Joe Hobbs, Sandie Holguín, the anonymous reviewers, and the editor for their helpful suggestions on earlier drafts of this article.

Notes

1. The term Indian is used throughout this article in lieu of Native American and other similar labels because in both authors' experiences it is far more widely used when Indian people themselves refer to a supratribal identity. The term is also used in the Commerce Clause of the U.S. Constitution and by 1846 had become embedded as a specific legal status that people on American soil can have, or be denied, as a matter of course in federal law. Thus, “entire volumes of the United States Code” are consumed with the term (Biolsi 2007, 7). Tribe has had similar legal status since it was first defined by the U.S. Supreme Court in 1901 in Montoya v. United States (Cramer 2005, 37).

2. “Indian Country” is the epitome of a discontinuous region. It includes reservations, rancherías, allotments, other trust lands, isolated dwellings and businesses, and other Indian lands of all kinds. It also has legal standing in all U.S. court systems. The phrase originated in the British Royal Proclamation of 1763. The United States adopted it originally to refer to all Indian lands under the jurisdiction and control of the U.S. Congress. The federal government retains authority in Indian Country, whereas the various states do not, although this prohibition continues to be contested and compromised (Biolsi 2007, 7, 29–30).

3. The idea that implementing GIS can be a colonial or assimilationist practice was recognized by academic geographers beginning in the early 1990s (e.g., Pickles 1995a, 224; Rundstrom 1995; Taylor and Johnston 1995; Harris and Weiner 1998). The subject has gained far more attention from geographers since 2000 (Laituri 2002; Piper 2002, 1–8, 134–35, 147, 169–75; Weiner and Harris 2003; Crumplin 2007; Palmer 2007; Bryan 2009, 2010; Crampton 2010, 6–11, 98–111, 123–27, 137–42; Elwood 2011; Farman 2011; Rose-Redwood 2012; Spiegel, Ribeiro, and Sousa 2012).

4. The vast majority of scholars in American Indian studies, history, and anthropology, including those we cite here, explicitly recognize internal colonialism as a condition of both past and present Indian life in the United States. Shaw, Herman, and Dobbs (2006, 272) are among the far fewer U.S. geographers who do. The problem of creating a viable alternative to the BIA is an exceedingly difficult one, made all the more so by the fact that some Indians fear that competency in self-administration can, if too successful, raise the specter of termination of the unique political and legal status for which their ancestors fought so hard. The fact that some white support for sovereignty is essential to maintaining this status, yet dwindles when Indians become less impoverished or less culturally distinct, only compounds the irony and the intractability of the situation (Castile 2006, 113–14). Thus, the notion of simply abolishing the BIA is closely linked to fears of termination. Deloria ([1969] 1988, 132–36) famously and concisely explained how this works in practice.

5. The technical reports are not distributed outside the agency and can only be obtained directly from the GDSC. Software distribution records and helpdesk information also have not been released to the public. The absence of correspondence, negotiations, or other dialogues between individual tribal governments and the BIA regarding GIS is noteworthy.

6. Battelle Corporation was renamed UT-Battelle after acquiring comanagement of Oak Ridge National Laboratory along with the University of Tennessee in 2000. It is a science and technology corporation specializing in the commercialization of technological systems. It contracts with more than 800 federal, state, and local governments in the United States. Battelle has annual revenues over $3 billion and has been manager of the U.S. Department of Energy's Pacific Northwest National Laboratory since 1965, the National Renewable Energy Laboratory since 1998, and Oak Ridge National Laboratory since 2000 (Battelle 2005).

7. In the 1990s, the Apache also began their own cooperative education programs teaching students technical and other aspects of forestry and management of other natural resources (Sloan and Welton 1997). Only one other similar case during this period has come to our attention. The Hoopa in northern California contracted with University of California Cooperative Extension foresters who made available a GIS called UC–SARA. Extension foresters and GIS developers also conducted training sessions on the reservation expressly for Indian personnel (Harris, Nakamura, and Blomstrom 1997). Nothing of that project appears to have involved the BIA.

8. Two of the largest corporations in the U.S. aerospace industry, Lockheed Aircraft Company and Martin Marietta, merged in 1995 to form Lockheed-Martin. The new corporation became “the largest military contractor in the United States and largest arms exporting company in the world” (Juhasz 2006, 138). In 2004, Lockheed-Martin earned approximately $35.5 billion, almost 80 percent of which was paid by U.S. taxpayers (Juhasz 2006, 138). Its central role in the military–industrial complex is matched by its leadership in the development and implementation of information technologies in agencies of the federal government like the BIA.

9. ESRI is the largest geospatial resources and commodities corporation in the world, with more than $600 million in annual earnings from its GIS operations alone (Noble 2011, 97–101). That the control of Indian trust assets now moves to companies like ESRI represents a continuation of the BIA's long history of implementing such privatizations.

10. The now-infamous class action lawsuit, Cobell v. Salazar, has become the gold standard for such fraud. In December 2010, Congress finally approved a settlement of this 1996 lawsuit brought against the DOI. The suit stemmed from the unaccountable loss of land-lease monies managed by the BIA in approximately 300,000 personal accounts of Indians. Some of the accounts date back to 1887 (Cahill 2011, 265–66).