Conclusion

---

Software has a unique place in the world of technology because its extreme malleability and complexity makes it embodies numerous aspects of a particular culture's methods, knowledge and philosophy. The importance of computational methods and of information and knowledge has pushed forward the development of information technology (IT). IT in turn has provided resources that change the individual, organizational and national environments. By creating and manipulating information and meanings people and organizations can frame the world to reach goals with IT mediating an interpretation of the world (Poole and DeSanctis, 1990). In this context culture and technology affect the information interpretation and framing.

The emergence of active decision support systems, software agents, meeting systems and brainstorming tools, knowledge management systems and other programs that contribute to the behavioral patterns of individuals and organizations affect the individual and collective "software of the mind". Information systems in any organization embed behaviours that constitute organizational and national cultures.

In the 1990s the necessity for adapting software to international markets was recognised and resulted in software internationalization architectures. Object oriented technology made it possible for the separation of GUI objects from the culture-specific locale. Current software development paradigms such as procedural and object-oriented programming fail at capturing units of software modularity that crosscut modules. However, the component-based architectures and those derived from "meta-level" development methodologies appear to be better suited for deeper localization. For example, a promising technology called aspect-oriented programming is arising from Xerox PARC, and has been demonstrated to capture crosscutting concerns in new units of software modularity called aspects (Kersten and Murphy, 1999). Since cultural concerns crosscut the system architectures, aspect-oriented programming may prove to be a valuable method of capturing these concerns.

In contrast to the implied universality of the user interface-based internationalization, the proposed application internationalization architecture is not culture independent or universal. We do not posit that separation of the Core Libraries from the culture-dependent libraries (see Figure 3) is possible for all cultures. We recognize the inherent cultural bias in every software product yet suggest that there are values, business processes and mechanisms shared by different cultures. Market-oriented economies and organizations employ many mechanisms across different cultures, the same goes for democracies and so on.

The recognition of the cultural bias embedded in software and the acceptance of the need for application internationalization presented earlier, require both a broader and a deeper perspective to recognizing the role of culture in software design and development-broader because interdisciplinary approaches are required to determine national and organizational requirements. Deeper approaches are necessary because software internationalization has to move beyond the surface manifestations of culture. The recognition of culture as a primary concern in the design of the application core is a key factor in the successful deployment of applications targeted at the international market.

---