Sunday, November 1, 2009

Distributed Cognition, Red Balloons, & Bioluminecent Squids (oh my)

Hutchins et al originally developed the theory of distributed cognition studying the workings of a Navy ship crew. In Distributed Cognition in an Airline Cockpit (1996) Hutchins & Klausen update this work by exploring cognitive theories about the workings of an airline cockpit by taking the unit of analysis to be the functional group and their organized processes and interactions rather than their individual minds. [1]

The ideas of distributed cognition have philosophical implications that bring to mind far-out science fiction ideas and visions of a future with a technology-mediated hive minded consciousness. Since cognition generally refers an information processing model, the actual approach of the theory is much more utilitarian than philosophical. Hutchins’ early work in shaping disturbed cognition theory arouse out of the need to apply cognitive theory to the workings of a ship crew and an airplane cockpit, and other highly organized distributed groups.

So, despite the interesting philosophical implications that challenge traditional models of the mind, distributed cognition is a surprisingly simple and elegant theory that works very well as a model.

Distributed cognition theory simply extends the level of analysis beyond the individual. The difference between someone who remembers something by writing it down and someone who relies on their short term memory is simply the difference between a cognitive system using an external process (pen, paper, and written language) or an internal process (short term memory). This idea of embodied cognition has many implications for HCI particularly with recent advances in ubiquitous mobile computing and emerging augmented reality technology making the separation between the individual and the “computer” thinner and thinner.

In Distributed Cognition: Towards a New Foundation for Human-Computer Interaction Research, Hollan, Hutchins, & Kirsh list three tenets of distributed cognition:

1. Distributed cognition is social
Hollan et al. present social organization as form of cognitive architecture. They argue that cognitive processes involve trajectories of information and the patterns of these information trajectories reflect some underlying cognitive architecture. Social aspects of distributed cognition have been widely studied and examples include the behavior of juries and the stock market.

2. Distributed cognition is embodied
In the embodied view of distributed cognition, minds are more than passive engines that render internal models of external phenomena. The organization of mind in development and in operation is an emergent property of interactions among internal and external resources.

In this view, work materials are more than mere stimuli for a disembodied cognitive system. From time to time they become elements of the cognitive system itself. Other examples such as a blind person’s cane or corrective eye glasses become a central part of the way some individuals perceive the world. Well designed work materials, such as a calendar alert system on a mobile device, become integrated into the way people think, see, and control activities, part of the distributed system of cognitive control.

3. Distributed cognition is culturally embodied
This view, includes the complex cultural environments in which we all live and work. Culture shapes the cognitive process of systems that transcend the boundaries of individuals. [2]

The environment is viewed as a reservoir of resources for learning, problem solving, and reasoning. Culture is a process that accumulates partial solutions to frequently encountered problems. Culture can be viewed as a residue of previous activity, that individuals can utilize to solve problems and build on the success of others.

Recent examples of distributed cognition

As technological innovations create new opportunities for distributed individuals to work together with others on a common task, distributed cognition may be becoming more and more useful as a model.

In 2008, Clay Shirky documented many of these new developments in Here Comes Everybody: The Power of Organizing Without Organizations [3]. Shirky, like many others, argues that our culture of media consumption is transitioning to include participation and interactivity. In one of Shirky’s most powerful arguments considers the potential “cognitive surplus” of individuals who begin to use social computing to create and organize. Shirky sees a trend of Americans spending less time passively engaging with media and more time actively participating and interacting with media. He calculates that if Americans applied just 1% of the time they currently passively participating in media (i.e. watching television) and used that time to engage & interact more actively with media this shift would represent the time and effort required to create and maintain 1,000 wikipedia sized projects each year. Clay Shirky’s talk can be viewed here:

a written transcript is here:

Just last week DARPA launched a contest to celebrate the 40th anniversary of the internet. This challenge indicates an interest in exploring the workings and emergent capacities of social-computational systems of people and computers. The Darpa Network Challenge places 10 red weather balloons in prominent locations across the country for four hours on a single day. The first individual or team to provide the exact latitude and longitude of all of the 10 balloons wins a $40,000 prize. Darpa describes the challenge as follows:
“To mark the 40th anniversary of the Internet, DARPA has announced the DARPA Network Challenge, a competition that will explore the role the Internet and social networking plays in the timely communication, wide area team-building and urgent mobilization required to solve broad scope, time-critical problems.

The challenge is to be the first to submit the locations of ten moored, 8 foot, red weather balloons located at ten fixed locations in the continental United States. Balloons will be in readily accessible locations and visible from nearby roadways.”
The nature of the task seems to require a complicated series of interactions with thousands of others, yet according to the rules of the contest only one individual may win the prize. Because of this element of competition and the relatively short time to organize, it’s likely groups will organize a number of different, complex coordinated and competing processes. These processes will likely emerge from the behavior and ideas of many, competing individuals and groups and will not be as engineered or planned from the top as Hutchins’ airplane cockpit.

It seems likely that many of the individuals who first happen upon the balloons in this four hour period will not be aware of the contest. Therefore, the information will need to travel through a complex path via communications using social computing tools like facebook and twitter, until it reaches an informed agent or group. This example is analagous to the perception of individual photons by an single first order neurons that are communicated to higher level visual processing structures of the brain. There are other complex neural processes such as those that enhance the signal to noise ratio. Similarly, it’s easy to imagine the contest will have many false positives and bits of misinformation winding their way through twitter, facebook, and other networks during the four hour period that contribute to a signal to noise problem. It may become necessary to send individuals to verify rumored locations or develop an algorithm to estimate the reliability of each report or otherwise attenuate the signals from the chatter.

By dangling a $40,000 carrot, it seems DARPA is pushing social networks and other web platforms to discover what structures of distributed cognition emerge. This problem space is very different from Hutchins’ early work in distributed cognition theory. Hutchin et al needed a model to represent the processes involved in the workings of a ship crew and an airplane cockpit. These processes were engineered, improvised, evaluated, and tested. In the DARPA experiment the scale will be to large to be engineered and determined from the top, many of the structures of distributed cognition will emerge in unpredictable ways.

Flipping the Script - The mind of an individual as a distributed system

Hollan et al introduce the idea that concepts and models of social groups can be used to describe what is happening within an individual’s mind. They site Minsky’s Society of Mind in suggest that cognition of an individual may also be distributed.

This idea of distributed components of an individual is similar to a talk that outlined new research in molecular biology that, like distributed cognition theory, shifts the level of analysis beyond the individual bacteria to study the complex processes of highly coordinated groups of bacteria. Biologist’s Bonnie Bassler April 2009 Ted Talk: Discovering Bacteria’s amazing communication system is a summary of ideas that outline cognitive-like processes carried out by well-coordinated bacteria. Further extending the level of analysis, she discusses a species of squid that utilize a highly organized community of bioluminescent bacteria to precisely control an anti-predation behavior and outlines new research in bacteria quorum sensing and other group processes.

This 15 minute talk is definitely worth a look:

Implications of Distributed Cognition of founding assumptions of HCI

These readings on distributed cognition advocated for a new foundation of Human Computer Interaction Research, nearly 10 years ago. Since then, there have been incredible instances of distributed individuals using innovations in technology to work together in new and powerful ways.

Is distributed cognition theory as it now stands, useful in explaining new ways people are using technology to work together? Does the theory need to be updated and re-evaluate? Does the theory need to be combined with other theories of group behavior such as activity or game theory? Do we need a more developed theoretical frame to better understand and conceptualize the many new ways people are working together?

As people are becoming more comfortable working as a group via technology they are utilizing and repurposing powerful technology for diverse uses. Apparent social computing has advanced to the point that DARPA needs to study security implications. As designers how can we stay ahead of things and design software that allows for flexible and emergent use, so that people can work together in new and expected ways.

Human Computer Interaction as a field has typically focused on a single user interacting with a computer to complete tasks. These days, computing is becoming more and more about rapid advances in complex systems of humans and computers working together in new and interesting ways. Human Computer Interaction must develop new theories to be useful and relevant. Rather than using theory to drive innovations in social computing, it may become increasingly important for the field of HCI to study new and interesting uses of existing technology and develop new theories to understand and shape these advances.


1. Hutchins, E. & Klausen, T. (1996). Distributed cognition in an airline cockpit. In Engestrom, Y. and Middleton, D. (eds). Cognition and Communication at Work. Cambridge University Press, Cambridge.

2. Hollan, J., Hutchins, E., & Kirsh, D. (2000). Distributed cognition: Toward a new foundation for human-computer interaction research. ACM Transactions on Human-Computer Interaction, 7(2), 174-196. [pdf]

3. Shirky, Clay (2008) Here Comes Everybody: The Power of Organizing Without Organizations The Penguin Press HC

Links to Video of talks

Clay Shirky's April 2008 talk at Web 2.0 Expo in San Francisco

Bonnie Bassler April 2009 Ted Talk: Discovering Bacteria’s amazing communication system

No comments: