40 meters or 4'000 kilometers apart - how much does it matter?

Thomas J. Allen, Architecture and communication among product development engineers, Engineering Management Society, 2000. Proceedings of the 2000 IEEE

One of the articles on collaboration, distance and technology that is most often recommended to me is Thomas Allen's 'Architecture and communication among product engineers'. It is most famous for its finding that people separated by more than a certain distance (40m is usually cited), are far less likely to work together than people whose workspaces are closer to each other.

Probability of communication

Allen finds a dramatic decrease even within the first 50m of separation.

Plotting ... results produces a curve that to no one's surprise shows probability of communication declining with distance. ... Communication probability declines to asymptotic level within the first 50 meters of separation.

Organizational affiliation modifies the effect. The decrease in communication is still found, but

the existence of a departmental relationship adds to communication probability by a constant amount, which is independent of distance. Project relationships produce a similar, but usually stronger, effect. This is due to the interdependence of project activities and the strong need for technical communication [for coordination].

Allen goes on to identify a more comprehensive set of influences.

So the probability that a pair of scientists or engineers will engage in frequent technical communication is a function of the degree to which they share a common base of knowledge; the rate at which that knowledge base is developing' the size of their organizational unit; the degree of interdependence in their work and the distance between their workstations.

Effect of telecommunications

When we had engineers report telephone and electronic mail as well as face-to-face communications, we found that, following a 'near field' rise the use of all media decayed with distance. ... One reason for the pattern observed in our data is that all of these media, as well as the written medium, are correlated in their use. We communicate with nearly the same people through all of these media. ... The more often we see someone face-to-face, the more likely that we will telephone that person or communicate by another medium.

Allen also observes qualitative differences between various media of communications. He cites a study by Hauptman (Allen & Hauptman, 1989) that shows that the telephone was used for less complex communication. Face-to-face was used for more complex information.

Many things, particularly technical ideas and problems, are difficult to communicate verbally. We need the assistance of diagrams or sketches. In addition, we often need the feedback that often comes from looking into the other person's eyes. The eyes communicate understanding. ... Similarly, in describing an idea or technical problem to someone, you can tell whether they are following you. If the indication is negative, you are prompted to restate the information in a different way. This feedback is invaluable in guiding communication.

Needless to say, it is strongest in face-to-face communication and weaker when speaking on the phone. The use of letters or e-mail not only reduces the possibilities for this kind of feedback, but also delays it.

3 types of communication

Allen uses a typology of communication to organize his work:
Type I communication is used to coordinate work.
Type II is necessary when the knowledge base of work changes rapidly (keeping up to date).
Type III is used for creative work.

While the article only touches in parts on the differences in communication behavior by type of communication, this would certainly be an avenue to explore further.