From One to Many: Egocentricity and Allocentricity in the Design of Collective Intelligence Systems

In his essay Betwixt and Between: The Liminal Period in Rites of Passage, the great anthropologist Victor Turner wrote, “as members of society, most of us see only what we expect to see, and what we expect to see is what we are conditioned to see when we have learned the definitions and classifications of our culture.”

This is a good definition of egocentricity; the process of experiencing the world from a single point of view. As human beings, we tend to see what we expect to see (which, ironically, is often given to us by others) and then define the world based on this view. This is strongly related to our sense of self, as well. “I am this, but and I am not that.”

Allocentricity, on the other hand, comes from experiencing the world from all possible points of view at once. It is being aware of the relationship of many things relative to each other, not just from one point of view. “This is, and that is, and this is how they relate.”

Psychologists call this perspective intersubjectivity; “I am aware that you are aware that I am aware. We exist in awareness of each other.” This is quite different from egocentrism, which in its extreme forms can be seen in people with Asbergers Syndrome, Autism or even sociopathy.

In my previous posts, I argued that we need the ability to move from egocentric to allocentric awareness, using what I called intersubjective collective intelligence systems. Returning to our earlier question, we can see how such a perspective could be useful for detecting and understanding threats and opportunities in diffuse, dynamic environments like those we often find ourselves in.

How might such systems work? In my next (and last) post, I will draw on lessons from Dave and others to outline how such intersubjective collective intelligence systems might be designed in order to produce similar kinds of insights, if not similar experiences, to that provided by allocentric awareness.

For now, I leave you with an example of how such approaches are being used in a related field, that of architecture and urban design.

The image above shows two different aerial views of a typical urban square. The one on the left illustrates the area visible form a single point, i.e., the red dot. This is what you would see if you were standing on that dot in the square looking around you. Everything within the black shape represents what would be visible from that point of view.

The image on the right shows what is visible from all points of view. The diagram uses space syntax analysis to calculate what parts of the square are the most co-visible, i.e., which parts can see the most and can also be the most seen. Areas in red are the most co-visible, whereas areas in blue are the least. This differentiates between a corner, which may be able see a large portion of the square, but can only be seen from certain perspectives, and a centre, which can both see and be seen by a wide variety of perspectives.

Why does this matter? As it turns out, allocentric analysis is a much better predictor of social activity in urban space than egocentric ones. Hundreds of studies show that people are far more likely to walk and relax in the red areas, while thieves, teenage children and homeless people are far more likely to be found in least visible areas. This impacts which shops are likely to be the most successful, who can charge higher rents, where people are likely to feel more or less safe, and where it is best to put public amenities such bus stops and rubbish bins, etc.

Most architecture and design is practiced from the first perspective, i.e., what a building will look like and how it will be experienced from a single point of view. We’ve all been in buildings like this; they look great from the front but are uncomfortable and inefficient from the inside. The comparison to online collective intelligence systems is like the difference between mass media broadcast, which is tuned to look good from a narrow point-of-view, to Web 2.0 and social media, many-to-many discussion.

In my next post, I explore some practical principles about how such systems may work, tying it back to the original questions at the beginning of this series of posts.