This is the ideal of interface design that many North American multimedia producers are moving towards. The locations and images on the screen that respond to contact with the cursor are called hot spots. Hot spots are activated when clicked, or simply by rolling the cursor over them. The latter are commonly referred to as rollovers, and the sophisticated use of rollovers has become a hallmark of good multimedia design over the past year. Examples of this abound in this year's new titles, one of the most remarkable being Mark Canter's Media Band.

Japanese titles, however, make much more sparing use of rollovers, and in many cases they are absent. The typical Japanese interface relies heavily on the point-and-click metaphor, and on navigation or manipulation using direction buttons, rather than through direct interaction with the screen or screen objects. In a title currently being developed by a Tokyo producer, 3D models of airplanes can be rotated using left-right/up-down buttons at the bottom of the screen, but not by interacting with the plane directly. A North American designer by contrast would probably have let the user orient the plane by rolling the cursor directly over it, pulling the nose to the left or right, up or down.

Viewing such interfaces, many Western developers will say that Japanese design is behind the times. A few will falsely assume that Japanese programmers do not know how to program rollovers or interactive screen objects, or believe that the Japanese do not understand their significance, or how to use them effectively. But of course Japanese developers do know how to program, and more importantly, they know what their users want.

Extensive testing of three multimedia titles--two reference, one edutainment--has shown that Japanese users tend to prefer to interact through point-and-click rather than through rollovers. They also prefer to manipulate an object using tools (buttons) rather than by manipulating the object directly. Rollovers are generally not expected to do anything more than to indicate when an area is hot. This is probably the most common use in North American titles as well, although the older option of having the shape of the cursor change seems to be making a comeback in multimedia. Rollovers that change large areas of the screen are often disliked by Japanese users, as are the use of a large number or very wide rollovers, which make it difficult to move the cursor without triggering some sort of screen event. The latter is actually the goal of some North American developers, as this is what makes the screen seem to come alive and respond to the user's touch.

It is easy to assume that Japanese users are not as experienced as those in North America, and that as they become more accustomed to interacting with the screen they will want the same type of interactivity that sells in North America, but this may just be wishful thinking. The above testing was carried out by two classes of users: naive users, mostly young mothers and retired couples, and by the staff of a software and multimedia developer. There was no difference in the affective response, although the latter group was able to grasp the concepts behind the interface more quickly. Given the prevalence of cartridge games in Japan and the extensive use of kiosks, one should not assume that Japanese users are naive.

The forms of interactivity expected in both markets are likely to evolve over time, but they will not necessarily evolve in the same way. An analogy might be the pop music industry. Much Japanese pop music sounds childish and cute to ears accustomed to industrial or grunge, and even the more sophisticated bands tend to go for a clearer, more diffused sound. This is the type of music that dominates the charts not only in Japan, but throughout East and South East Asia as well. And Japanese pop music is outselling American pop music in a number of Asian countries. Little is learned by assuming that the forms dominant in one's own market are superior or will necessarily be successful in a different market.

To date, relatively little research has been done on cross-cultural differences in how people most naturally interact with the screen, and most of the evidence is anecdotal, as is what I have presented here. This is a subject that needs serious study, as it affects issues such as database structure, screen interactivity, how multimedia information is to be referenced or indexed, and how hypertext links are to be made.

At my own company we are currently developing a multimedia title called Silk Road , based on the ancient trade route of the same name that linked China with Europe from the 2nd to the 16th centuries. A path over mountains and across deserts, for the exchange of ideas and information as much as for goods, this was the Internet of its time. It is also a subject close to the Japanese heart, the topic of countless books and a major multi-part television documentary by the national public broadcaster NHK.

We are developing both a Japanese and English version in-house, as we have a fairly strong team of Japanese writers, designers, and programmers to complement our English- speaking team. Trying to reconcile Western ideas about interactivity with Japanese ideas, while negotiating the widely differing cultural assumptions about the significance of goods such as silk and nylons, or emotional and intellectual issues such as the essence of Buddhism, has been fraught with difficulty, and has not always made for amiable relations between the two teams.

What we have tried to do is to incorporate Japanese views of the subject in the English version which challenge our own established assumptions, and to challenge some of the Japanese assumptions in the Japanese version. For the Japanese interface we have added more user feedback and made decision paths clearer. In some cases we have modified rollovers, making the screen real-estate effected by a rollover smaller, or even changing it to a point- and-click button. Certain Japanese users have also found it difficult to use the drag-and-drop metaphor, a technique which is so common in Apple. We have kept this metaphor in Silk Road, but in the future we may experiment with the use of elements that work both with drag- and-drop and select-move-deselect (in this case the mouse button is up during the movement phase). This is an approach suggested by several different Japanese programmers who come from Windows or NEC 9800 series backgrounds.

Interactive multimedia is still in its infancy, and new paradigms for interaction appear almost every month as more and more elements of the media mix are made to respond to the user. Inevitably, the evolution of interactivity will progress differently across cultures, as it does with the novel, the comic book, music, and film. North American developers must take this into account when they develop products for Asian markets, and they need to follow trends and new developments in interface and interactivity design as they arise in Japan, China, India, and other countries and cultures.

The issue goes much deeper than screen design, however. Different cultures prefer to access and present information differently, in ways that are determined as much by language as by the inherent structure of the information. Indeed, it is quite difficult to decompose the two. In the near future, we will be accessing databases and manipulating spread sheets using the natural language query techniques developed in the artificial intelligence community. These will pose new challenges to software localizers, who will have to perform radical re-organizations of the way in which an application structures, references, and accesses information. I suspect that it will be easier to localize information structures in software which has been designed using the object- oriented paradigm than with the structured- design paradigm. In the future, the process of localization may be more a question of modifying the behavior of objects than of linear translation.

Designers of mainstream applications will have to follow trends in multimedia as well. The aesthetics of multimedia interfaces and the concepts of interactivity incorporated into multimedia and games will become the standards for all computer applications. The first exposure children have to a computer already tends to come from multimedia titles such as The Manhole, by the team that went on to develop Myst, or the Broderbund Living Books series. These future users of computers at the office and on the shop floor will expect screen behavior and information structure to be based on models experienced in the home, or will at least feel more comfortable and consequently be most efficient in those environments.

It is no coincidence that one of the best loved tools of designers, Kai's Power Tools, has one of the most interesting interfaces of any application on the market. And it is possible that the aesthetics of its interface helped KPT, and Photoshop itself, gain acceptance in the Japanese market. Future interfaces are as likely to be influenced by titles from Cyan or The Residents as they are by Microsoft Word or ClarisWorks, and by young designers in Asia and Europe whose first efforts have not yet made there way into the market in North America.