In this paper, we describe the influence of physical proximity on the development of collaborative relationships between scientific researchers and on the execution of their work. Our evidence is drawn from our own studies of scientific collaborators, as well as from observations of research and development activities collected by other investigators. These descriptions provide the foundation for a discussion of the actual and potential role of communications technology in professional work, especially for collaborations carried out at a distance.

How has the proliferation of communications media changed the volume of communication and the distribution of information in large organizations? There is reason to think that the availability of more communication media increases the amount of communication employees receive, with positive effects on their organizational knowledge and commitment, but negative effects on their perception of being overloaded. Prior research suggests that electronic mail may differ from other media by more effectively spreading organizational information to peripheral employees, and doing so while interrupting them less than other styles of communication. This paper uses survey data from a large US-based multi-national corporation to examine the effects of communication by electronic mail and other media. Results are that employees who used electronic mail extensively, net of their communication over other media, were better informed about their company and more committed to its management's goals. One reason for their superior organizational knowledge seems to be that electronic mail promotes "information spillover" from a focal recipient of a message to others who are less directly interested in a message, but does so without subjecting the marginal parties to the burdens of interruption and information overload.

Junk e-mail or spam is rapidly choking off e-mail as a reliable and efficient means of communication over the Internet. Although the demand for human attention increases rapidly with the volume of information and communication, the

this report. This section provides a conceptual framework for future applications of the network. The framework borrows heavily from a report summarizing a National Science Foundation (USA) workshop on the National Collaboratory, a vision of a nation-wide structure that provides tools enabling researchers to perform their work and collaborations without regard to geographical location (Lederberg and Uncapher, 1989). The framework is a multi-tiered hierarchy (see Figure 1, adapted from Lederberg and Uncapher, 1989). At the top are the research functions in the scientific process that researchers and developers cycle through. Scientists achieve these functions through a variety of collaboration functions, which in turn are implemented via collaboration tools and underlying technologies.

this article, we make a conceptual distinction between a preference for information and entertainment as compared with a preference for interpersonal communication. We operationalize this distinction by comparing people's use of the World Wide Web with their use of personal electronic mail (e-mail). This operationalization is not perfect. When people use Internet services---the World Wide Web, electronic mail, Telnet, Usenet groups, MUDs, and so forth, they may be finding out the weather in the Bahamas, downloading games, chatting with friends, learning about hobbies, or sending coworkers a report. A single session on the Internet can fulfill multiple information, entertainment, and communication goals. Nonetheless, the World Wide Web and e-mail are the most popular of all Internet services and most distinctly represent people 's preference for information and entertainment as compared with their preference for interpersonal communication, respectively. Our prior research (Kraut et al. 1996, Table 2) suggested that use of the Web and email are sufficiently independent of each other to have some distinctive causes and effects