An entertaining and interesting look at how group email technology is used by and influences groups in workplaces. They authors formed some hypotheses about the impact email groups would have, particularly investigating whether email group processes would mimic conventional group processes. The authors gathered data by monitoring the email of a sample of workers in a large corporation. Employees were familiar with email and had good access to the email system. This is one of few studies we have seen which was performed in a setting where users are well acquainted with the technology under investigation. The authors relied on case studies to illustrate and reinforce key points. The results of their studies, while not surprising to regular email users, are worth keeping in mind: for example, "in each discretionary [email] group over half of the group messages contained information that would not have been received without electronic communication", discretionary groups tend to reassert their identity frequently ("this group is for discussing user interfaces only"), entertainment groups are more active than work-related groups, email groups help people feel connected to the organization, etc. One interesting thing they noticed is that email groups can get much larger than conventional groups without breaking down.

Overall, I found the paper to be an excellent preliminary investigation into the long-term, large-scale effect of email on groups.

This paper is a good overview of the effects of email, mainly on individuals. The authors list numerous differences in the way people behave when using email, and give their opinion of what about email causes that effect. A sampler of effects: People talk more frankly and more equally. Status gets ignored (whether that status is legitimate or not -- short people don't get dumped on as much, but experts get overlooked).. Networked groups generate more proposals for action. Groups take longer to arrive at a decision. People flame. A sampler of causes: "the social and contextual cues that usually regulate and influence group dynamics are missing or attenuated." It's harder to distinguish communication which is formal vs casual, official vs unofficial, humorous vs serious, authoritative vs opinionated, etc. The paper also discusses the impact of email on organizations. Email enables the creation of a distributed "knowledge pool" that employees can draw upon. It lets people have a wider scope of influence (and conversely causes a person to be influenced by a wider range of people). It requires different systems of management, particularly management of credit and incentives, to work effectively.

This is a fairly good overview paper with little original content (the one study is limited in scope and not very enlightening). The paper was published in 1991, so is rather dated. In short, it is a good reference paper but not useful as a source of ideas or data.

An important early paper on semistructured email and user-controlled filtering. The Information Lens system allows users to specify message types and message attributes. It gives an interface to specify filtering rules and actions. It has an inheritance hierarchy of message types. The system provides full backdoors for users who do not wish to use it in all cases, and it allows an incremental path for using more features of the system. This paper is primarily about the system, and only secondarily about the impact and importance of using a filtering system of some sort.

The paper is a must-read for anyone who wants to create a related system. This is a good reference system which has a number of qualities commonly lacking in proposed systems: ready ways to bypass the system when desired, user extensibility (and without editing configuration files or programming), and unintrusive operation.