and emerging from cooperative ventures, new complexities are introduced as government, industry, and academia explore joining forces as a catalyst to innovation. The varied cultures of each of these sectors will require flexibility to accommodate the interests of the others. Perhaps most important, a stable body of consistent law is required on which all parties can rely, thereby avoiding yet one additional complicating factor that would be introduced if sui generis laws were prevalent.

As to the third point, the catalogue of approaches to research is expanding: it includes large corporate programs, national laboratories, research done under contract, state-funded programs, university research, and individual research done "at the kitchen sink." We also see technology alliances and networks that do not necessarily depend on equity ownership or even contractual relationships. We see on-line computer-connected research done at two or more locations, even on opposite sides of the world. We see platform building through knowledge sharing that includes "casual" research relationships, in which one company or laboratory will volunteer information to another (on a confidential basis) with an expectation of some feedback. Also, there is more transient employment in laboratories as researchers move from place to place. More attention is being paid to how to conduct research to maximize creativity, and the computer is playing an ever greater role.

I have noted that one of the major ways R&D has changed is by being globalized. This is true of research being conducted by single companies as well as by research consortia. Computer linkages through satellites, data bases, and networks mean that research of all kinds can be conducted in different locations as a single effort. This is already far advanced. It is not unusual for teams operating in different time zones to relay information through electronic mail and other techniques so that research continues around the clock. These linkages are not confined to business. Some of the biggest networks are shared by private industry, universities, and government agencies. Some are public or quasi-public networks, whereas others are private or classified government communication channels.

There are several factors that are prompting research through global networks, now that the capability has been demonstrated. A desire for greater proximity to customers, suppliers, and university talent accounts for some global activity. In other areas, relatively lower personnel cost is a notable reason for going overseas. Another reason is the desire to tap talent found in other countries. A leading example is access to programmers in India through satellite links.

There are elements of risk sharing, cost sharing, and economies of scale in these arrangements, but the desire to pool knowledge is increasingly evident as a motivating factor. IBM's new arrangements with companies such as Siemens and Apple illustrate this. Whatever the mixture of motiva-