Next: 6.3 Combining Cognitive Science and the Situative View Up: 6. Discussion Previous: 6.1 Everyday Reasoning

6.2 Where and When Do Conventions Exist?

As modeled in MOVERS-WORLD, convention is the subject matter of memory. Memory is organized around points of coordination within the progress of a joint activity (see Section 4). This principal of organization assumes that individual actors are trying to converge on an ideal structure for the activity. The analysis of the data shows that with the addition of this expectation (not its completion), reductions in the number of primitive actions taken, communication costs, and plan effort are all achieved as the overall performance of the community continues to improve. Section 5 shows that, given this principal of organization, the machinery of individual memory can convert raw common experiences into improvements in the performance of joint behavior. The analysis of the experimental data of MW does not, however, provide evidence that there ever exists a unique or ideal point of convergence in the mind of each of the participants (see Section 5.6). Each memory provides an orientation, a set of expectations, how the current activity will proceed. Some recollections may be better than others, but none of the individual memories have any special status. In other words, improvements in coordination for practiced joint activities are achieved, even though no ideal or unique script for a conventional behavior is determined.

Perhaps a unique idealization of a conventional behavior exists in an external representation. For example, in the usage of an air telephone there exists, in the form of instructions, an external representation of the conventional behaviors the individual actor must produce in order to use the device. Considerable work is necessary to achieve the behavior dictated by the instructions [70,3]. An icon of a credit card informs the air telephone user of the method of payment, but to achieve the conventional behavior of inserting the credit card so it can be read necessitates a great deal of work [6].

Reconsider the case of Dick and Jane coordinating their efforts in order to achieve joint passage through a doorway. What the model suggests is that as Dick and Jane continue to practice their joint activity, they both have the expectation that an ideal and unique script for the convention will eventually emerge. Because the recall of prior joint activities orients future behaviors, the expectation of conventional behavior shapes performance. Nevertheless, as independent actors, Dick and Jane do not converge on an ideal script; effective performance will continue to require online adjustments and cooperation in order to keep on-track and in-synch. Even if the wheelchair came with a set of instructions that describe a conventional behavior for one person to aid a second person in a wheelchair at passing through a closed doorway, it would still require considerable work on Dick and Jane's part to achieve their common goal.

To recapitulate, there is no ideal internal script that represents the convention. An external representation of the conventional behavior requires work, cooperation, and coordination of effort. Whatever structure for the conventional behavior that exists prior to the joint activity at best approximates a behavior of convention. But then where, or perhaps when, is the convention?

Research studies in ethnomethnodology [22] and situated activity [70,48,1] report that the structure for activity is, in part, a product of the activity. Perhaps this is also the case for behaviors that are conventionalized, i.e., the structure of convention is at least partially a product of activity. In the case of Dick and Jane, there is no analysis that will anticipate the coincidence of details that confront them as they wend their way through the activity. By means of a program of combining expectations with the specifics of the situation that confronts them, Dick and Jane are able to coordinate their behavior. Although there are prior structures for the activity that informs participant behavior, additional structure arises during the course of the activity.

An example of a structure that could emerge during a joint activity is a shared plan [34]. From the perspective of collaboration, each occasion of joint activity is managed through the ongoing construction of a shared plan [32,33]. Shared plans may be partial and can be revised as the activity continues. For a group of collaborators to have a full shared plan there must be a mutual belief that they are committed to the success of the action and a mutual belief of the need to accomplish each of the subactions. A key idea is that even when they only have a partial shared plan to achieve a given subaction -- collaborators only partially know what they are going to do -- they always have a full shared plan to work out the details of the partial shared plan. A unique shared plan is constructed on each occasion of joint activity.

If the structure of a conventional behavior partially emerges as a product of joint activity, then convention bears some relation to a shared plan. When Ed and Joan start living together they have not worked out a convention for cleaning house before company comes over, but over time they might. If there is a shared plan for the nth time they perform this activity, there was also one for the first time they did it. This points to the critical difference between a shared plan and a convention. A shared plan is a construct that is tied to a single episode of joint activity. A convention is a construct that is tied to the history of related joint activities. In other words, if the unit of measure is a single episode of joint activity, the shared plan is a synchronic analysis of the behavior of the participants and conventions are a diachronic one. Convention is measured by a reduction in the amount of work needed by participants in a joint activity to achieve a common goal. Over time, through practice at working at a recurring problem of coordination, certain points of coordination become expected, a design for the activity develops, and a convention becomes a part of the initial common ground. The emergence of convention coincides with the development over time of a home task environment for coordinating behaviors. A shared plan realizes a conventional behavior only after a community of actors have worked together, over many episodes of joint activity, to achieve a given goal in a given context. The shared plan reflects the work participants do in performing a behavioral convention on a given occasion, but the same convention may be invoked by different shared plans on different occasions.

To summarize: When do conventions exist? They exist after a community of actors have had the opportunity to work together for awhile: they develop with the home task environment. Where do conventions exist? They exist in the predisposition of individual actors that develops from their shared practice and is contingent on historically conditioned constraints and affordances of their shared task environment.


Next: 6.3 Combining Cognitive Science and the Situative View Up: 6. Discussion Previous: 6.1 Everyday Reasoning
Last Update: March 10, 1999 by Andy Garland