Background and Literature Review

As the wzorld becomes more connected through technology, two competing hypotheses have emerged regarding the effect of globalization on culture. These hypotheses are discussed in more detail by Webber (1969), Yang (1986), and Ronen (1986). One hypothesis, convergence, suggests that cultures will become more similar as they modernize because they must pass through a "relatively fixed pattern of development" (Coughlin, 2000, p. 422). In addition, "the rapid growth of telecommunications and computing technology … holds profound implications for possible societal convergence" (Coughlin, 2000, p. 428). The Internet could play a key role in this process.

The opposite hypothesis, divergence, suggests that cultures tend to resist assimilation and adapt technologies in culturally distinct ways. Furthermore, the cost of technological assimilation can be considerable, leaving "have not" cultures unable to participate to the same extent even if the population might wish to do so. In some cases, the same technology can result in both convergent and divergent outcomes.

One of the difficulties in researching convergence theory is determining where to look for it. For example, Coughlin (2000) identifies industrialization (including means of production, occupational mobility and differentiation, increased emphasis on formal education, and growth of cities), stratification of class structures and occupational prestige, demographic patterns, family life, education, and the welfare state as areas of interest. Most quantitative research has been cross-sectional, although longitudinal studies seem more appropriate (Williamson and Fleming, 1977).

Most research on global or international information systems takes a divergence perspective, pointing out the problems that can occur when cultural differences are ignored. For example, Fernandes (1995) and Del Galdo and Nielsen (1996) provide guidance on user interface design. Both books point out problems that have occurred when user interfaces designed for one culture have been applied to another. More recently, researchers have been interested in cultural implications of global e-commerce issues (e.g., Davis, 1999) from a similar perspective. Simon (2001) provides empirical evidence that differences in the perception of web sites and satisfaction with them do exist among different cultures (and between men and women).

However, when Ein-Dor, Segev and Orgad (1993) investigated the effect of culture on international information system construction, they found considerable consistency and thus support for the convergence view. Ito and Nakakoji (1996) provide an interesting example of convergence, showing how Japanese word processors follow a Western typewriter model. They also note that Japanese programmers began with English languages (e.g., Fortran) and were reluctant to switch later to Japanese programming languages. Thus, information technology can also be a force for cultural homogeneity.

Web pages are much less expensive to build than programming language compilers and corporate information systems and so provide a much better opportunity for cultural customization. Johnston and Johal (1999) suggest that the Internet is a virtual culture and, using Hofstede's (1980) dimensions, conclude that this culture is both evolving and converging. Marcus and Gould (2000) reviewed selected web pages from various cultures, basing their analysis on updated work by Hofstede (1991). They conclude that culture, as expressed by Hofstede's dimensions, does affect the design of web pages. Thus, we can also see examples of divergence as cultures adapt new technology in different ways to suit their own needs. Simon (2001) also uses Hofstede's dimensions to analyze how students from different cultures react to select web sites.

Hofstede (1980, 1991), based on extensive research, suggests that culture may be differentiated via five major dimensions. These dimensions are described in Table 1 (adapted from Hunter and Beck, 1997).

Table 1: Hofstede's cultural dimensions

Dimension	Description
Individualism Collectivism	Individualistic cultures expect their members to be independent and look after themselves. Collectivist cultures have a tightly knit framework of mutual dependencies and obligations.
Power Distance	High Power Distance cultures accept unequal distribution of power within its society. Low Power Distance cultures strive for equalization and participation.
Uncertainty Avoidance	Strong Uncertainty Avoidance cultures attempt to control uncertainty by strict rules and codes of behavior. Weak Uncertainty Avoidance cultures are not as strictly controlled and deviation is more acceptable.
Masculinity Femininity	Masculine cultures emphasize achievement, success, and assertiveness. Feminine cultures emphasize caring, close relationships, and harmony.
Long-Term Orientation Short-Term	Long-term oriented cultures promote the family, respect for older people, and virtuous behavior such as hard work and frugality. Short-term oriented cultures develop equal relationships, emphasize the individual, and promote creativity and self-actualization.

This research builds on the work of both Hofstede (1980, 1991) and Marcus and Gould (2000) by focusing specifically on national portals. Web portals are Internet sites intended to be the starting point (i.e., a virtual launch-pad) to locate information and services on the Web.

Yahoo!, which began as a search engine in 1994, is perhaps the best known example. Full-service national portals are designed to appeal to a more focused audience within a country or culture, typically offering a search engine, directories of links on a set of selected topics, news items (including weather, sports, entertainment, and stock market results), advertisements and shopping, and other services such as free e-mail services and web pages. Thus, a full-service national portal is a platform for a collection of activities that are enabled by Web technology and that serve its intended audience (or culture). National portals are well suited for this research because they are quite common (many countries have multiple national portals), are intended for a particular culture or nation (rather than the worldwide Internet community), and are likely to endure and evolve over time. National portals will reflect the culture and values of developers, but they are also commercial ventures that must be responsive to the culture of their intended audiences to be successful. Just as a magazine's content says something about both those who produce it and those who read it, so does that of a national portal.

Paralleling the tremendous growth of e-commerce and other Internet services, the role of portals will continue to evolve rapidly. Portals may become the major link to entertainment and informational video (as the Internet and television converge), telecommunications (offering video phone calls), financial transactions, and other key services. Therefore, portals have the potential to become a major public policy issue. For example, Canada has tried to protect its magazine industry against split-run American magazines (Magder, 1998). Other countries have had similar concerns and will want to take steps to protect their interests. Yet, major portal providers often provide "split-run portals" for major countries and groups. For example, as of July 2002, Yahoo! offers links to 22 national or regional portals from its main page plus sites for American Chinese and Spanish speaking people. Some offer considerable local content, but there are currently few regulations or monitoring agencies to control this. Who determines the content is unclear. Nevertheless, split-run portals provide at least some local content. A recent survey by MMXI Europe found that about a third of the most popular sites in Germany, France and Britain were American portals (NUA, 1999).

Indigenous national portals catering to the needs of distinct national and cultural groups should reflect the socio-cultural, technological and economic characteristics of their cultures and countries both in their appearance and the list of services they provide. Of particular interest are national portals in Europe and Asia that are not in English. About half of Web content is expected to be non-English by 2003, up from about 20% in 1999 (Burke, 1999). Different countries and cultural groups have their own buying habits, social pastimes, ways of reporting and accessing news, values about public media, modes of business transactions, and ethics and taboos that restrict social practices. Not all countries are at the same level of technological adoption, either. As the world adopts Internet technology, portal developers must balance pressures towards convergence (i.e., providing a site much like Yahoo! which its intended customers may already be using) and divergence (i.e., to reflect cultural variations in their indigenous portals).

Some limitations of using portals to reflect culture should also be noted. In some countries, only a small percentage of the population has Internet access. Internet users are likely to be younger, wealthier and better educated than the average citizen (Johnston and Johal, 1999). Some will have gone outside their countries, often to Europe or the United States, for their education and are thus more familiar with these cultures. Some national portals may even target expatriates. Thus, until access to the Internet greatly increases, an argument can be made that national portals are directed at a small (and not necessarily representative) segment of some cultures.