One important factor to remember when considering the cultural biases that may affect the usability of digital libraries is that biases are, in fact, embedded in development tools themselves. Until recently, program development was a skill for few to master. It required detailed knowledge and sometimes years of experience to make a computer perform exactly the way it was intended. Since then, IT-development tools have become more and more sophisticated. Today various tools on the market cater to a variety of purposes and different degrees of sophistication.

Authoring tools are high-level, easy-to-use programming tools for creating interactive multimedia applications. They synthesize multimedia content from various sources into a single presentation, they provide interactive and presentational objects like buttons, windows and scroll boxes, and they offer various options for publication. The general look and functionality of potential applications are more or less pre-set. Therefore, multimedia applications authored with a particular tool have an easily distinguishable style -- they look very much alike and they have identical peculiarities and limitations. This means that many important decisions - such as those about templates, icons, color schemes and computing metaphors -- have been made before the application developer starts working. These culturally laden decisions have consequences not only for application developers as users of authoring tools -- they may permeate the applications they create to the end-users of the applications.

Aided by colleagues here at Middlesex University, I have begun a program of research designed to provide a stronger base of knowledge about cross-cultural usability issues in digital libraries and other information system applications. As part of our first phase of work, we recently completed a study - using participant observation, interviews and a survey - of a Middlesex University class of 60 students with various cultural backgrounds. About one quarter of the class were British English, while three-quarters of the students had foreign backgrounds from as many as 13 countries distributed over several continents. The participants were also part of many different subcultures of Greater London.

To illustrate the nature of findings that socially grounded cross-cultural usability studies can contribute I will highlight just two findings from our study. One is that students with different cultural backgrounds have different color preferences. But a more subtle (and perhaps more important) cultural issue we uncovered is that students with an international background did not wish to be distinguished from their British English counterparts.

Color preferences were revealed when students - who had no prior education in human-computer interaction (HCI) principles or the use of colors for human-computer interfaces - set about developing an educational technology module. English students used pastel color schemes with a lot of gray and low contrast, Scandinavian students tended toward dark colors also with low contrast, whereas students with a Jamaican background chose strong and bright colors with high contrasts and combined them into very colorful schemes. African students usually chose black as the ground color and added some brighter colors. European and U.S.-American students more or less chose a bright background, black text and few moderately colorful objects. No particular pattern seemed to emerge in the use of colors by Asian students.

The survey that we launched later in the semester showed that the culture of the country of origin mixes and mingles with the local (sub-) cultures. Our survey results indicate that "Black-British" students with an African or Caribbean background showed very similar color preferences, although the color preferences in their countries of origin are very different. Nearly all of them chose black as the primary preferred color. Whereas most of the students with an African background chose reddish-brownish-ocher colors as secondary colors, three of the four students with Caribbean backgrounds preferred strong and bright colors to go with the black.

This preference for interface colors seemed mirrored in personal clothing color preferences. We observed that most students with Caribbean and African backgrounds dress in black from top to toes. Their dressing black is very much part of their subculture.

Students with a cultural background in Africa or Jamaica defined their nationality as "Black-British" or "Black-English," thus defining themselves as part of English society and part of a particular subculture. The color preferences seem to reflect this process and the colors of the country of origin become deflected by the color preference of the UK-subculture they belong to. If the preferred color of their country were golden yellow, they chose black with yellow. The originally preferred color mixed with the dominant black of the subculture.

The survey we launched with the intention to collect data about the various cultural backgrounds of the participants showed some unexpected secondary effects regarding cultural issues and caused us to rethink, in part, our approach to the subject. Knowing that we were approaching a sensitive topic (that of ethnic minorities in the United Kingdom), our definition of students with foreign backgrounds (not foreigners!) was that they were one of the following:

1)  students who had recently come to the United Kingdom for study,

2)  students who had come to the United Kingdom with their parents, or

3)  students who were born in the United Kingdom as children of first generation immigrants.

Comments in the margin of the questionnaire and subsequent interviews revealed that students were unhappy with this attempt to categorize them. They simply did not want to be differentiated from the norm, the norm being British English.

However, the reasons for this attitude amazed us as much as the attitude itself. In addition to bringing forward the issue of discrimination in the United Kingdom, students - especially those who had been in the United Kingdom for less than 10 years - expressed a strong belief that "British is Best" and especially better than everything in their countries of origin. We heard some tragic stories of multiple cultural memberships, alienation from and transition between cultures. One ethnic Chinese interviewee from Malaysia had been brought up "English" partly to identify with the former colonial power and partly to avoid Malaysian discrimination against ethnic Chinese. For these students, computers, electronic information systems, the Internet and some recognizable interfaces were status symbols. The idea of culturally sensitive information systems, they feared, would make them second class citizens in the world of computing.

These two examples from our research suggest that both primary and secondary cultural dimensions will affect the development and use of digital libraries and other information systems. They also reveal that user studies in multicultural contexts are extremely sensitive. In fact, secondary issues of post-colonial tension between the cultural periphery and the cultural center raise the question of whether it is sensible to have culturally sensitive interfaces. If this question is answered in the affirmative, the next question is to what extent the primary cultural issues are counteracted by secondary issues and where the balance lies for a well-designed, politically correct and internationally usable interface.

We hope that these two examples more generally illustrate the value of cross-cultural usability studies of information systems. Results from such work will assist researchers and practitioners in HCI who wish to integrate cultural aspects in information system development. It also delivers relevant results for domain specialists and content providers. In the long term, the consideration of cultural aspects of digital library design and use will prevent industry losses caused by culturally offending and inappropriate technologies. Also in the long term, this line of work should produce positive outcomes for the broader community of digital library users around the world. It has potential to benefit the culturally diverse users of digital libraries as they access and use networked information sources.