This paper examines the influence national culture has on the design of information systems through a comparative study of geographic information systems design in the US and Germany. Hofstede's (1980) dimensions of national culture provide the theoretical framework for this research. Applying Hofstede's dimensions, the paper evaluates differences in the design documents and actual practice of design of King County, Washington, US, and Kreis Osnabruck, Germany. The findings support Hofstede's characterization at the conceptual level of design documents, but indicate that the practice of design in the German county deviates considerably from Hofstede's characterization: whereas Germanic national cultural characteristics suggest a very regulated top-down design process, the actual practice of design in Kreis Osnabruck involves, in fact, a great deal of negotiations. These negotiations are obscured by the cultural emphasis on regulation, as Hofstede did indeed predict through high uncertainty avoidance. In comparison, the findings in King County support Hofstede's characterization that Anglo-American national culture involves negotiations at all stages of design.

Francis Harvey: Swiss Institute of Technology (EPFL), Lausanne, Switzerland

ACKNOWLEDGMENT: Portions of this text are based on a book chapter titled "National cultural influences on GIS design", which will appear in the book: Geographic Information Research: Transatlantic Perspectives, Craglia, M. and Onsrud, H. (Eds), Taylor and Francis, London, 1997.The author would like to thank the following individuals for their assistance in carrying out this research: Martin Balikov, Karl Johanson, and Thomas Remme. The University of Washington Graduate School's Western European Dissertation Grant provided financial support for completing the case study in Germany.

National cultural influence on information system design

Like all human activities, culture influences information system design in a myriad ways. The complexity of cultural influences eludes easy understanding and remains obscure in the complexity of daily life (Clifford and Marcus, 1986; Emerson et al., 1995). From the numerous facets of culture, national culture has received particular attention. Researchers and practitioners refer to substantial empirical work that differentiates the important roles national culture plays in organizations. The work of Hofstede (1980) stands out for the connection of design activities to national culture and organizational forms. His comprehensive study of over 100,000 questionnaires in 66 countries is the basis for a noteworthy theoretical explanation of the influence of national culture on information system design.

The research presented here examines the question of whether Hofstede's framework applies to the actual practice of information system design. It compares the designs of geographic information systems (GIS) in a German and a US county. The ethnographic research design sets out to evaluate Hofstede's quantitatively developed characteristics.

The questions examined in this research are:

- Do Hofstede's national cultural characteristics explain differences gathered from reviewing design documents?

- Does the formal representation of design in documents replicate national cultural characteristics?

- Do they also help explain the actual practice of design?

Beyond Hofstede's cultural dimensions, this research considers other literature from the information science field about national culture (Hofstede, 1980; Jordan, 1994; Willamson, 1975). These researchers all root their understanding of culture in the sociological work of Max Weber. Culture is commonly understood in Weberian sociology as the shared set of beliefs that influence what we consider to be meaningful and valuable. Disciplines, professions, and institutions in modern bureaucratic society nurture and transmit cultural values and meanings (Albrow 1990; Weber 1946, 1947). It is important to note that research in this vein ascribes ideal typical qualities to each culture in a Weberian sense: they are the strived for forms, not individual characteristics. In other words, research can only find distinctions between social group behavior in terms of these dimensions.

With this brief introduction to the theoretical background and research issues complete, the next section moves on to present Hofstede's framework and other relevant work. The framework consists of four national cultural dimensions (uncertainty avoidance, power distance, individuality and masculinity) each with particular characteristics that influence information system design. This research examines the two pertinent dimensions (uncertainty avoidance, power distance).

Following the presentation of the theoretical background for this work in section two, the methodology employed is described in section three. Section four evaluates the GIS designs of the two counties, Kreis Osnabruck and King County, and section five examines the differences between design documents and design practice. The final section summarizes the research findings and presents an explanation for the differences found between GIS design practice and Hofstede's formal framework.

Dimensions of culture

Out of a mass of empirical data Geerd Hofstede developed the above mentioned four dimensions of national cultural influence in information system design. Because of the wealth of data and deep theoretical interpretation, this work has received much attention in the information systems field as well as other fields (Berry, 1989; Jordan, 1994).

Hofstede specifically examined the role of national culture in work-related values and information system design (Hofstede, 1980). Hofstede constructed his framework on a review of sociological and anthropological theories and work including Geertz (1973), Kluckhohn (1951, 1962), Parsons (1951), Parsons and Shils (1951) and Weber (1946). Hofstede's four dimensions of national culture are:

- uncertainty avoidance: the extent to which future possibilities are defended against or accepted;

- power distance: the degree of inequality of power between a person at a higher level and a person at a lower level;

- individualism: the relative importance of individual goals compared with group or collective goals;

- masculinity: the extent to which the goals of men dominate those of women.

Uncertainty avoidance is the focus of information systems and decision support systems (Jordan, 1994). It is considered together with power distance because of interaction effects (Hofstede, 1980). The other two dimensions, individualism and masculinity, having little importance and relevance to German and US cultures, lie outside this research's focus. Germanic and Anglo-American cultures are strongly differentiated in terms of uncertainty avoidance; the power distance dimension is quite similar.

Uncertainty avoidance and power distance form critical interactions affecting organizations. According to Hofstede, in Germany and the USA, both characterized by low power distance, there are two possible ways to keep organizations together and reduce uncertainty. In Germanic cultures, with high uncertainty avoidance, "people have an inner need for living up to rules, ... the leading principle which keeps the organizations together can be formal rules" (Hofstede, 1980, p. 319). With low uncertainty avoidance (Anglo-American cultures), "...the organization has to be kept together by more ad hoc negotiation, a situation that calls for a larger tolerance for uncertainty from everyone" (Hofstede, 1980, p. 319). Figure 1 shows important organizational characteristics based on uncertainty avoidance. Hofstede makes detailed comments about these differences. The "Anglo" cultures "would tend more toward creating implicitly structured organizations" (Hofstede, 1980, p. 319). In contrast, German speaking cultures establish "workflow" bureaucracies that prescribe the work process in much greater detail (Hofstede, 1980, p. 319). Hofstede argues that problem-solving strategies and implicit organization forms follow: Germans establish regulations, Anglo-Americans have negotiations. Germans conceive of the ideal organization as a "well-oiled machine," whereas Anglo-Americans understand a well functioning organization as a "village market" (Hofstede, 1980).

Information transaction cost theory (Willamson, 1975) provides complementary insight into cultural influence on organizational structure and approaches to problem solving. In this theory, all business activity is a transaction between individuals and groups. Information serves as the controlling resource (Jordan, 1994). In this form the theory is overly reductionist and simplistic. Boisot (1987) extended this transaction cost theory to include cultural issues, distinguishing two characteristics of information that affects transactions:

- codification: the degree of formal representation;

- diffusion: the degree of spread throughout the population (Jordan, 1994).

Internalizing the transaction in the organization reduces the diffusion of information (Jordan, 1994). Centralized information requires a bureaucracy, whereas diffuse information is distributed in a market. These differences correspond to Hofstede's national cultural characteristics (Jordan, 1994). How information system design codifies or diffuses information will depend on the importance of uncertainty avoidance and ideal organization type. The codification or diffusion of information is complementary to Hofstede's dimension of uncertainty avoidance. Low uncertainty avoidance corresponds to information diffusion, whereas high uncertainty avoidance corresponds to codification.

Normally, highly integrated industries and commerce utilize the information transaction approach. GIS design approaches often begin with a similar structured systems approach (Gould, 1994). When considering heterogeneous public administrations, a different, highly diversified organizational structure is possible. In county governments the multi-disciplinary interests, missions, goals, and perspectives require special consideration of the values propagated by institutions and disciplines.

Methodology

This ethnographic research compares the GIS designs and implementations in King County, Washington, USA and Kreis (County) Osnabruck, Lower-Saxony, Germany. Although the two cases are quite similar, the research organization differed owing to different periods of study. This resulted in a conceptual division in two research phases. In the first phase design documents were examined and compared (see Harvey (1995) for the first report of these results). During the second phase, the author participated as an observer in the actual design process to validate the findings from the first phase and to test Hofstede's framework. The ethnographic research design was chosen for the detailed insight it provides into the distinct cultural and institutional context of each GIS (Onsrud et al., 1992). In the case of King County a strategy of contextual inquiry was followed, compared to naturalistic observation used during a shorter visit to Kreis Osnabruck (Wixon and Ramsey, 1996).

The case studies in these two counties were prepared following Hofstede's framework with a focus on uncertainty avoidance and assessing differences between design documents and actual design practice. Ethnographic approaches to differences in scientific practice (Anderson, 1994; Hayek, 1952, 1979; Hirschhorn, 1984; Latour and Woolgar, 1979; Nelson, 1994) also influenced the choice of a participant observation approach to collect data. The actual issues raised during document evaluation, open-ended interviews, written correspondence, and telephone conversations focused on GIS design and the construction of organizational, institutional, and physical components.

The case study in King County occurred over a longer portion of time (six months) during which the author participated in the system conceptualization project. Because of the distance to Kreis Osnabruck, key questions were posed in written format over several months time before the site visit. During an intensive one week visit, open-ended interviews were held with six project participants. These interviews were analyzed during and immediately following the week-long intensive observation. The author's training in German planning and administrative law plus experiences with GIS applications in Germany enabled the key research questions to be obtained rapidly. The longer period of contact in King County made up for an original lack of knowledge of US public administrative practice and theory.

The preparation for the visit to Kreis Osnabruck involved formulating specific questions and issues about design practice, uncertainty avoidance, and the role of regulations and negotiations. Questions focussed on filling gaps in the recent history of the county GIS, understanding the role of different administrative agencies in the design process, and examining the practice of GIS design and implementation. During the visit, several agencies were visited and discussions took place with county staff.

Because of the far longer duration of observation in King County and a more direct involvement with the project, the case study in King County was organized differently. The details of the research design were formulated parallel to the author's work there, so this case study involved retrospective and inquiry phases. After six months project participation at King County, there were several meetings, telephone calls, and written correspondence with project staff to discuss specific questions related to project history, design, and implementation.

The preliminary evaluation of documents and an ongoing exchange of discussions and/or e-mail to discuss various questions, permitted a solid entry into the complexity of each county's design approach before participant observation. The design documents for each county were examined and evaluated in terms of Hofstede's framework. Flood protection planning was chosen for more detailed examination because of the similarity of this mandate in both counties.

Designs on documents and designs in practice

Design documents

Both King County and Kreis Osnabruck started their respective GIS projects in 1989. King County's GIS design began after several failed attempts, characterized by break-downs in negotiations. Kreis Osnabruck's design (see Figure 2) involves a detailed examination of two departments' mandates (regional planning and environmental protection) that focusses on the identification of tasks the county GIS should support.

Kreis Osnabruck's design relies strongly on standards, particularly legislated standards. ATKIS - Automatisierte Topographisch-Kartographische Informationssystem (automated topographic and cartographic information system) is the most important standard for Kreis Osnabruck. State law requires that all public administration geographic data be available in ATKIS format. It is an object-oriented data model for provision of vectorized topographic data at three scales: 1:25,000, 1:200,000 and 1:1,000,000. This acronym is also used to refer to data provided by agencies in this format, particularly the state surveying departments. Other standards and guidelines are also of importance: ALK - Automatisierte Liegenschaftskataster (automated property cadastre) is the standard for the automatization of the Grundbuch (property book). MERKIS, the Massstaborientierte Einheitliche Raumbezugsbasis fur Kommunale Informationssysteme (map scale orientated uniform spatial co-ordination basis for communal information systems) lays the conceptual groundwork for developing GIS in an ongoing project to simplify German public administration. GIS occupies a central role in this reform as the replacement of diverse maps and centralization of information. It describes GIS at the communal level as a "...geographic database for agency specific, spatial communal information systems based on the national coordinate system, a unified data model for all topographic and agency specific spatial data..." (Der Oberkreisdirektor Landkreis Osnabruck, 1990). (See Table I for a comparison of design organization and data models.)

The reliance on standards fits Hofstede's high uncertainty avoidance characteristic for Germanic national culture. The organization of design and implementation further underscores the workflow bureaucratic approach. Kreis Osnabruck's GIS design documents lay out a plan that involves three phases. In the first phase, questions regarding administrative functions (following the respective legal mandates) and problems with the available cartographic products were raised. The results were the basis for the detailed breakdown of administrative functions into tasks and objects. These tasks and objects are finally implemented during the last stage of design, when all issues and conflicts are to be worked out.

Reflecting the role of negotiations, King County's GIS design process is far more complicated. Although it followed commonly accepted GIS design procedure (needs assessment, conceptual design, pilot study), the autonomy of participating agencies and county politics led to a very convoluted development. The final design (see Figure 3) involves a project that constructs the core data layers and infrastructure, but then stops there, leaving many issues open to further negotiation. The central group in King County is basically a steering committee. There is no regulation or standardization for the county GIS. The GIS design activities fit Hofstede's characterization of a village market.

As this suggests, the design of KC-GIS was, not surprisingly, complicated and convoluted. This brief overview of the development of the GIS design offers some insights into the myriad actors involved in design. After an internal proposal for a GIS fell apart due to internal strife, PlanGraphics was called in to carry out the design. Their project began with a needs assessment. The basic tenet of the PlanGraphics needs assessment report points to the requirement for coordination and a centralized organization. They are the presumed basis for effectively using GIS technology that provides information and services to fulfil county administrative and governmental functions. The design paradigm follows the line that because departmental functions and information are dependent and related to other departments, a centralization of the functions and information in a county GIS would improve the effectiveness of King County's administration.

The Needs Assessment report (1992d), adopting a strategy of limited centralization, focussed mostly on elaborating county needs for a GIS in terms of common, shared, and agency specific applications. The intent was to determine which elements of a single department's applications were common with other departments.

The PlanGraphics GIS design proposal left a great many issues unresolved. These gaps required an exhaustive study of the conceptual design document and discussions with the various agencies to design a project that would fulfill objectives: in other words, establishing the playing field and negotiation. Starting with the PlanGraphics documents, a special group in the Information Systems Department of Metro prepared a scoping report (1993) with a more exhaustive overview of design, but left the implementation to inter-agency negotiation. Maintenance questions were left open for even later negotiation.

Many potential GIS applications identified in the PlanGraphics reports were later eliminated, because the budget for the project was reduced from US$20 million to US$6.8 million. The project's focus was reduced to the creation of the infrastructure and essential layers for a county GIS. After successful creation, responsibility for the layers would return to the "stakeholders". From the PlanGraphics proposal only the names of the essential layers remained. The contents of the layers were left open to negotiation. The reduction in funding without a corresponding redefinition of mission and vague descriptions of mandates meant the design stage had to continue into the implementation phase, necessitating yet further negotiations. This fits Hofstede's characterization of Anglo-American culture perfectly.

Design practice

Each county's design documents stop short of identifying specific GIS operations or functions required to prepare data or carry out parts of a task. It is clear that in King County the preparation of design documents and the negotiation of implementation are inextricable. However, there was no exact indication before the participant observation in Kreis Osnabruck if the design documents were rigorously followed or received permanent residence in the county archives. The evidence from the design documents supported Hofstede's findings that, given the high uncertainty avoidance in Germanic culture, Kreis Osnabruck rigorously follows standards and regulations. However, the actual process of getting the design to work remains obscure.

Indeed, the case study indicates that the actual practice of GIS design in Kreis Osnabruck differs considerably from Hofstede's characterization of Germanic national culture and the suggested procedures described in the design documents. Basically, the research findings suggest that the transformation of regulations to design and implementation occurs in Kreis Osnabruck through negotiation. This was related by means of several examples. A good example is the case of database software. The object-oriented database software was abandoned by the company writing. Lacking it, the entire software design had to be reworked around an off-the-shelf product. This change was worked out through negotiations between participating agencies. Problems arose almost daily during implementation, requiring quick action and alterations.

Contrasted with Kreis Osnabruck, with only an implicit framework of regulations and guidelines for GIS design and implementation, the design of King County's GIS project relies heavily on negotiations between departments. Since design work concludes only by pointing out the many loose ends to be dealt with by the respective departments (PlanGraphics, 1992b), negotiation will always be the crucial step in project design. The puzzle pieces, illustrating how different parts of the county GIS should "fall into place", in Figure 3, suggest graphically the importance negotiations have even at the end of formal design.

Comparison of flood protection planning

The detailed examination of the use of GIS for flood protection planning, a mandate similar in both counties, illustrates the influence of national culture on information system design. In both counties, flood protection planning is formulated in regulations (laws, legal codes, ordinances). Its goal is to reduce human and environmental risk by designating areas (i.e. flood plains) that cannot be built on, or only under certain circumstances. The design of GIS to support this planning reflects the national cultural dimensions. In both counties the practice of flood protection planning is a result of negotiations that makes do with existing constraints.

Flood protection planning in both counties is by overlaying transparent maps. The GIS implementation foreseen to support this mandate does not alter this procedure in any way. GIS overlay will be used in the same way as the overlay of map transparencies is used now. The results of the overlay, now displayed on a computer monitor, are reviewed by a planner in the same manner.

Flood protection planning is legally and administratively different in both countries. In Kreis Osnabruck it is defined in laws and administrative guides that define the regulations that must be fulfilled. The design of the GIS on paper reflects these laws and guides. The practice follows the established experience in dealing with planning projects. In King County the ordinances leave most of the GIS design open. In essence, the ordinances only define criteria. The application, design, and implementation are left open to the responsible agency. The GIS implementation basically requires the digitization of the respective plans, and automatization of the overlay operation. The agency is consulted during the earliest negotiations to determine that the base county data layers include the flood protection zones.

The differences in GIS design documents for flood protection planning fit Hofstede's national cultural uncertainty avoidance characteristics. Kreis Osnabruck develops the GIS operations around established regulations and King County employs GIS overlay in a manner consistent with the agency's established practices following agreements negotiated with the other county agencies. The importance of negotiations in Kreis Osnabruck still applies to the practice of designing the flood protection planning. Clearly there is a discrepancy between organizational self-representation (that Hofstede describes) and the actual practice of design in Kreis Osnabruck.

Different documents and practices

Whereas the importance of negotiation suffuses the entire process of preparing design documents and bringing them into agreement with the diverse interests of multiple county agencies in King County, the GIS design in Kreis Osnabruck appears to follow strictly the three-stage design process laid out at the onset of the project. The design practice in King County is in fact characterized by the negotiations Hofstede identified and explained as low uncertainty avoidance. On the other hand, the participant observation in Kreis Osnabruck turns up an interesting contradiction between documents and practice. Although regulations dominate the representation of design in documents, all insights into the practice of design suggested the strong role played by negotiations. However, although negotiations are so important, they still occur in the context of fulfilling standards and regulations for the most part. Ad hoc design occurs, but is downplayed, even when it means a complete alteration of the information system architecture. These differences in the GIS design approaches at the conceptual level go back to Kreis Osnabruck's reliance on standards and regulations (ATKIS, ALK, MERKIS), whereas King County develops their GIS from the ground up.

Hofstede's national cultural dimensions of information system design are clearly recognizable in each county's GIS design documents. Kreis Osnabruck describes its GIS in terms of a clear and concise framework of laws, regulations, and accepted standard operating procedures. Before any product or GIS function is implemented in this German context, it is first formalized and codified.

At a very early and obscured phase this usually involves negotiations (for example the modified ATKIS used in Kreis Osnabruck), but these negotiations are completed before the formalization of the design and accepted as regulations by other institutions. This reliance on regulations slows down the development of the county GIS to the rate at which regulations can be put in place.

King County, on the other hand, continually negotiates the design and implementation of the county GIS. The loose ends in the design documents and the importance awarded "stakeholders" interests reflect the "village market" approach. Piece-by-piece, portions of the county GIS are agreed to and implemented. This leaves design issues and, in particular, maintenance issues open or, simply, unresolved until implementation, reflecting the national cultural characteristics that lean towards negotiation as a design strategy. Design documents are just another part of negotiations. Agreement is only established for a very limited portion of the whole GIS design. Additionally, the vagueness of the design documents also leaves many courses of action open, requiring negotiations before any further work is done. Because of these complexities, the agencies still operate largely from one another.

Missing the representation of design and the practice of design in Hofstede's framework may result from the respective cultural emphasis on negotiations or regulations. In Kreis Osnabruck the GIS is implemented by fulfilling standards. In King County work on the design aims to fulfil negotiated requirements and retain institutional and disciplinary positions, and ensure a viable position for future rounds of negotiations. The work practice in the Kreis Osnabruck project was in fact strongly compartmentalized according to the regulations they needed to implement. Due to this compartmentalization, much work was required to resolve discrepancies between different regulations and constructing working GIS software. Even in this compartmentalized world of German public administration, individuals rely on a web of contacts with co-workers and knowledgeable outsiders in the practice of design. This non-formalized part of GIS design remains a tacit component of their work lives and is scarcely mentioned in discussions. Regulations are essential in a formalized hierarchy. Informal meetings and arrangements with coworkers and outsiders are only the backdrops for design activities. The dominant view is that if these practices do not culminate in regulations, they are not important to the project, nor worth reporting.

The project manager in Kreis Osnabruck was aware of these issues and conundrums. He indicated the difficulties of implementing broad standards and pragmatically solving the problems of implementation. In his words, "Kreis Osnabruck strives for an 80 per cent solution" (Remme, 1995). Although the documents reflect the uncertainty avoidance of Germanic national culture, the design practice clearly shows the necessity of negotiation to get things to work, even just 80 per cent.

Summary and conclusion

Regardless of national culture, the diversity of perspectives and purposes in any public administration means the design of an information system will always require some type of negotiation. Regulations shift the focal points, and lend a strong structure, but even regulations are ultimately implemented through some form of negotiation. Hofstede's national cultural characteristics seem to be a good indicator of cultural values and representation, but not of practice.

In King County negotiations and renegotiations of the GIS are ongoing. Compared to Kreis Osnabruck, the county GIS is not as stable, but agencies are extremely flexible in their response to institutional, legislative, and political contingencies. In Germany issues are negotiated and then codified as regulations or laws. The results are robust institutional solutions that offer an explicit framework, but bind the agencies involved to already established approaches that can lead to idiosyncratic solutions. New applications, consequences, and new actors' roles must be addressed and formalized in existing institutional structures before action is taken. This takes up many institutional resources and delays the response of institutions to new opportunities.

In the strong hierarchy of German public administration, emphasis on regulations and fulfilling legal mandates dominates the participants' representation of design activities. This attests to the high uncertainty avoidance in Germanic cultures that Hofstede identified. The centralization of information in Kreis Osnabruck contrasted with diffusion in King County also corresponds to the complementary measures of information codification or diffusion. In a culture so engrossed with regulations, it is no wonder the outside observer, employing quantitative research techniques, only turns up the aspects emphasized by the national culture. Going beneath the veneer of regulations and standards to the practice of design reveals a complex practice of negotiation and ad hoc problem solving. In spite of the emphasis on regulations and standards, the actual work constructing the GIS in Kreis Osnabruck involves negotiations as much as regulations.

Qualitative research can lead to valuable insights that illuminate the influence national culture has on information design practice. Benefiting from technology requires broadening our cultural understanding at the theoretical and practical levels (Kaye and Little, 1996). The ethnographic case studies of King County and Kreis Osnabruck show that, in spite of similarities, national cultural factors help explain design practice as well as the formal representation of design. The GIS techniques used (overlay of flood protection zones) may well be similar, but national cultural values lead to completely different designs of GIS technology.

The finding that negotiations are crucial to design practice in both the counties bears further research towards a reconsideration of Hofstede's framework. It suggests Hofstede's dimensions of national culture are a good basis for understanding the influence of national culture on organizations' self-representation, but miss the actual practice of social activities. Hofstede warned against the tendency to understand cultural out of the context of one's own national culture, but clearly focusses too strongly on the theoretical dimensions of abstracting culture's influence to organizations which veiled the nitty-gritty of design practice. This may rise from Hofstede's theoretical emphasis on systematic approaches in sociology (i.e. Parsons) that discount the individual. An ethnographic-based reconsideration of Hofstede's framework is necessary. Essential in this task is an emphasis on negotiations and the web of relationships between cultures, institutions, and disciplines in practice.

References and further reading

1. Albrow, M. (1990, Max Weber's Construction of Social Theory, St. Martin's Press, New York, NY.

2. Anderson, R.J. (1994, "Representations and requirements: the value of ethnography in system design", Human-Computer Interaction, Vol. 9, pp. 151-82.

3. Berry, B.J.L. (1989, "Comparative geography of the global economy: cultures, corporations and the nation state", Economic Geography, Vol. 65, pp. 1-18.

4. Boisot, M. (1987, Information and Organizations: The Manager as Anthropologist, Fontana, London.

5. Clifford, J. and Marcus G.D. (Eds) (1986, Writing Culture. The Poetics and Politics of Ethnography, University of California Press, Berkeley, CA.

6. Der Oberkreisdirektor Landkreis Osnabruck (1990, Das Kommunale Raumbezogene Informationssystem (KRIS) Eine Arbeitspapier zur Realisierung, Referat A, Osnabruck.

7. Der Oberkreisdirektor Landkreis Osnabruck (1992a, Situationsbericht 12/2/92, Der Oberkreisdirektor, Osnabruck.

8. Der Oberkreisdirektor Landkreis Osnabruck (1992b, Das Kommunale Raumbezogene Informationssystem Osnabruck (KRIS) Gemeinsamer Abschlussbericht der Projekt- und Entwicklergruppe (Final Report) 20 May1992, Der Oberkreisdirektor, Osnabruck.

9. Der Oberkreisdirektor Landkreis Osnabruck (1993a, Fachkonzept (Entwurf), Der Oberkreisdirektor, Osnabruck.

10. Der Oberkreisdirektor Landkreis Osnabruck (1993b, Losungsvorschlag, Der Oberkreisdirektor, Osnabruck.

11. Der Oberkreisdirektor Landkreis Osnabruck (1993c, Systemkonzept, Landkreis Osnabruck, Osnabruck.

12. Emerson, R.M., Fretz R.I. and Shaw, L.L. (1995, Writing Ethnographic Fieldnotes, University of Chicago Press, Chicago.

13. Geertz, C. (1973, The Interpretation of Culture: Selected Essays, Basic Books, New York, NY.

14. Gould, M. (1994, "GIS design: a hermeneutic view", Photogrammetric Engineering and Remote Sensing, Vol. 60 No. 9, pp. 1105-15.

15. Harvey, F. (1995, "National and organizational cultures in geographic information system design: a tale of two counties", in Peuquet, D. (Ed.), Proceedings of the Twelfth International Symposium on Computer-Assisted Cartography (AutoCarto 12), Charlotte, ACSM/ASPRS, Charlotte, NC, pp. 197-206.

16. Hayek, F. (1952, 1979, The Counter-Revolution of Science, Liberty Press, Indianapolis, IN.

17. Hirschhorn, L. (1984, Beyond Mechanization. Work and Technology in a Postindustrial Age, The MIT Press, Cambridge, MA.

18. Hofstede, G. (1980, Culture's Consequences. International Differences in Work-Related Values, Sage Publications, Beverly Hills, CA.

19. Jordan, E. (1994, "National and organisational culture: their use in information systems design", Faculty of Business, Report, City Polytechnic of Hong Kong, Hong Kong.

20. Kaye, R. and Little, S.E. (1996, "Global business and cross-cultural information systems: technical and institutional dimensions of diffusion", Information Technology and People, Vol. 9 No. 3.

21. King County Metropolitan Services (1994, "King County Consolidated GIS Project Requirements Specification", (Report KCReqs1a.mac) 10/18/94, King County, Seattle, WA.

22. King County Metropolitan Services (1995a, Request for Qualifications and Proposals (RFQ/P CS/M03-95), January 1995, King County Metropolitan Services, Seattle.

23. King County Metropolitan Services (1995b, "Agreement for professional services for the King County GIS Project Data Conversion", Contract No. CS/M03-95 between King County Department of Metropolitan Services and Smartscan, Inc., WA.

24. Kluckhohn, C. (1951, "Values and value-orientations in the theory of action: an exploration in definition and classification", in Parsons, T. and Shils, E.A. (Eds), Toward a General Theory of Action, Harvard University Press, Cambridge, MA.

25. Kluckhohn, C. (1962, "Universal categories of culture", in Tax, S. (Eds.), Anthropology Today, University of Chicago Press, Chicago, IL.

26. Landkreis Osnabruck Regionalplanung (1989, Die Moglichkeit der Einfuhrung eines graphischen Informationssytems auf der Grundlage von ATKIS beim Landkreis Osnabruck, Regionalplanung, Osnabruck.

27. Latour, B. and Woolgar, S. (1979, Laboratory Life: The Social Construction of Scientific Facts, Sage, Beverly Hills, CA.

28. Municipality of Seattle (1993, King County GIS Scoping Project, Municipality of Metropolitan Seattle, Seattle, WA.

29. Nelson, A. (1994, "How could scientific facts be socially constructed?", Studies in History and Philosophy of Science, Vol. 25 No. 4, pp. 535-47.

30. Onsrud, H.J., Pinto, J.K., and Azad, B. (1992, "Case study research methods for geographic information systems", URISA Journal, Vol. 4 No. 1, pp. 32-44.

31. Parsons, T. (1951, The Social System, Routledge & Kegan Paul, London.

32. Parsons, T. and Shils, E.A. (1951, Toward a General Theory of Action, Harvard University Press, Cambridge, MA.

33. Plangraphics (1992a, King County GIS Benefit/Cost Report, PlanGraphics, Frankfort.

34. Plangraphics (1992b, King County GIS Conceptual Design, PlanGraphics, Frankfort.

35. Plangraphics (1992c, King County GIS Implementation and Funding Plan, PlanGraphics, Frankfort.

36. Plangraphics (1992d, "King County GIS Needs Assessment/Applications", Working Paper, PlanGraphics, Frankfort, KY.

37. Remme, T. (1995, "The GIS of Kreis Osnabruck", (Interview) August 17-19.

38. Scherder, B. (1995, "Flood Protection Zones in Kreis Osnabruck" (Interview), August 18.

39. Schulze, T. and Remme, T. (1995, "Kommunale Anwendungen beim Landkreis Osnabruck", in Kopfstahl and Sellge, H. (Eds) (1995), Das Geoinformationssystem ATKIS und seine Nutzung in Wirtschaft und Verwaltung, Niedersachisches Landesvermessungsamt, Hannover, pp. 193-8.

40. Smartscan, I. (1995, Scope of Work for King County, Smartscan Inc., Seattle, WA.

41. Weber, M. (1946, "Bureaucracy", in Gerth, H.H. and Mills, C.W. (Eds), From Max Weber: Essays in Sociology, Oxford University Press, New York, NY, pp, 196-224.

42. Weber, M. (1947, The Theory of Social and Economic Organization, (trans. Henderson, Parsons, A.M. and Talcott), 1st American edition, The Free Press, New York, NY.

43. Willamson, O.E. (1975, Markets and Hierarchies: Analysis and Antitrust Implications, Free Press, New York, NY.

44. Williams, R. (1976, "Culture", Keywords, Oxford University Press, Oxford.

45. Wixon, D. and Ramey, J. (1996, Field Methods for Software and Systems Design, John Wiley & Sons, New York, NY.