Maturana and Varela's construct of 'domain' circumscribes a particular instantiation of a system's status and prospective trajectory of interactions. Where the given system is operating as an observer, its cognitive domain circumscribes the set of cognitive interactions into which it can enter. As Maturana and Varela (1980) point out, the indexical / explanatory 'connections' drawn by an observer among phenomena are contingent upon the manner in which the observer's cognitive domain intersects the domain(s) in which those phenomena are engaged.
In studying group task interactions, it is often apparent that participants are operating with respect to different such 'domains' of reference. This has been noted by Robinson (1989;1991), who delineates the distinctions between 'cultural' and 'formal' levels of language in task-oriented group interactions. Indexical mis- or disorientation among conversants is often associated with flipover, which Robinson and Bannon (1991, p. 225) describe as the phenomenon in which "...the object of an interpretation and the interpretation itself change places." They provide multiple examples characterizing flipover in terms of a mutually created interpretive unity (e.g., a position, a document) becoming a fixed object subject to later interpretation. To use Robinson's (1989) terminology, flipover occurs when interactors, having modified or interpreted the formal level representation through cultural level dialogue, then confront the results of their cultural level interplay as a new state of the formal level. Flipover is thus characterized as an unavoidable problem in interaction (therefore in group work). On the other hand, as a "...dialectical movement between an object of interpretation and interpretation itself..." (Robinson and Bannon, 1991, p. 223), flipover imparts utility and power to collaboration.
In my studies of individual and group decision making tasks, it had been readily apparent that flipover was an issue. There were two basic types of flipover events which could be generally discerned:
Today's problem solving / decision making activities increasingly involve collaboration among members of a work team. The nascent means for collaboration, problem framing, and even decision-related inference is information technology (IT). Issues are typically framed with respect to 'systems' of diverse character, scope, and degree of complexity, and this framing is increasingly done in terms of depictive devices (e.g., knowledge schemata, flowcharts). The two classes of flipover events noted above are therefore integral (as well as relevant) to today's problem solving / decision making processes.
As observers (in the autopoietic sense) whose cognitive domains intersect both the task and depictive venues, decision makers are vulnerable to disjunctions or mismatches between actual task scenarios and the representational devices which are their primary interface with the target setting. Task / depictive flipover is problematical when it causes such disjunctions, and these in turn induce blind spots or distortions which mislead decision makers.
One example of such a problem is depictional lock-on . This is the label for the phenomenon in which meeting participants 'lock onto' the depictional device(s) employed as the sole structure for framing their interactions and the accretion of a knowledge artifact within a given session (Whitaker et al. , 1995). In numerous knowledge elicitation sessions, I've noted participants adapting their (e.g.) referential patterns, physical orientation, and inferential tactics to conform to the affordances of the depictive tools at hand. On the one hand, this is a desired effect, to the degree that it connotes a sessional facility with the given tool(s). On the other hand, it can be negative when discussion of the task venue is subordinated to the instrumental mechanics of the depictions.
Such subordination has been observed consistently in knowledge elicitation sessions employing concept mapping -- a simplistic idea sketching procedure in which key concepts are transcribed into an unrestricted semantic network format of labelled nodes and links. Attentional (and even postural) 'lock-on' to the concept map occurs within 5-10 minutes after the session opens. At that point or soon thereafter, the participants (knowledge elicitees) begin framing their contributions with respect to (e.g.) 'adding nodes and/or links' here or there (as opposed to freely relating aspects of the subject matter). In other words, past the lock-on point, the processual course is navigated with respect to the depictional mechanics rather than topical priorities. Typically, such 'locked-on' sessions are closed by the elicitees' commenting that the depiction is sufficient, rather than any claim that the topic has been adequately explored.
Constitutional level flipover is problematical when decision makers cannot agree upon or co-orient to a shared mapping among relevant micro / macro phenomena. The illustration I conventionally use is that of a meeting in which selected company managers are discussing changes in a factory facility. For the sake of illustration, consider the change to involve reconfiguration of work units / elements (and, accordingly, shifts in these units' interrelationships). Comprehensive planning would require a similarly comprehensive 'map' of the units and their interrelations under both current and prospective conditions. Now consider the unit / relationship mappings which constitute the primary 'task domains' for selected managers:
The above scenario is simplistic, but it should suffice to illustrate the distinct personal 'maps' with which each of the listed managers must navigate their way through the proposed changes. Confronted with a specific need (e.g., optimizing profits), potential course of action (e.g., downsizing the workforce) may be straightforward for the accounting and shop managers, but equally problematical for the health manager and the union steward. Such conflicts of interest are the rule rather than the exception, and their attempted resolution accounts for much of the time spent conceptualizing, planning, and implementing enterprise changes. Imperfect resolution largely explains the negative consequences (foreseen or not) which mar many attempts at such innovation.
With regard to this discussion, the problem can be seen as a conflict in merging each participant's conceptualization of his/her local / particular task domain into a composite spanning the entirety of the affected enterprise. To be sure, each of the participants may assume that his/her specific task domain does span the entire enterprise. Indeed, the listings above highlight the manner in which each of them addresses the facility's workers. The point is this: The fact that each of the participants' 'task domain maps' includes workers as key elements does not mandate that in each case the worker-element is equivalently 'critical', nor that it is delineated uniformly with respect to the web of interrelationships constituting each participant's task domain 'model', nor that all the participants' such 'models' are mappable onto each other (e.g., in terms of scope or complexity).
The prevalence of the above-described flipovers in actual problem-solving / decision-making exercises serves as an important clue for how to profitably delineate a model for addressing them as application areas for IT interventions. I have employed these clues to develop a cyclical process schema model for problem-solving / decision-making as illustrated in Figure 1.
Figure 1: Process Schema for Problem-Solving / Decision-Making
(After Whitaker: 1992; 1994)
The processual flow (moving from left to right in Figure 1) involves noting a problem (or issue motivating concern), adopting a perspective on its nature, deciding upon a course of action, and then taking said action. These steps are plotted at the top of the figure, employing Boyd's (1987) OODA Loop (Observe-Orient-Decide-Act) construct. The figure is laid out for the simplest case -- one in which the decision maker is immersed within the task domain in which the problem issue is identified and within which intervention is enacted. In this uniform-perspective case, it follows that the decision maker is operating via a syn-referential vantage in the task venue, but abstracting to a sys-referential vantage when operating in the depictive venue.
During the Orientation phase of this process, the issues identified within the task venue are mapped onto a depictive vehicle (formally or informally) which serves as the basis for subsequent analysis and decision formulation. This is an instance of the task / depictive flipover discussed earlier. The processual waypoint at which this flipover occurs is termed the depiction bottleneck . In many cases, the depiction bottleneck is also the waypoint at which mappings between syn- and sys-referential vantages (on the identified problem issue) become reified. It must be pointed out that the task/depictive and constitutional-level flipovers are distinct, and that they are not necessarily accomplished in tandem (either in terms of temporal, procedural, or conceptual coordination). Because the referential transformation(s) implemented at the depiction bottleneck ground the remainder of the process, mistakes or insufficiencies deriving from this waypoint will adversely affect the subsequent decision and action(s).
During the Action phase of the process, the decision framed within the depictive venue is translated into instrumental interventions back in the task venue. This, too, is an instance of task / depictive flipover. The processual waypoint at which this flipover occurs is termed the enactment bottleneck . It is at this waypoint that (all too frequently) errors or deficiencies in an intervention become apparent. Barring those cases where such errors are directly attributable to flawed decision making under adequate depiction, the problems that arise here are typically traceable back to inadequacies in negotiating the earlier depiction bottleneck.
The process schema model sketched above has been employed to demarcate critical phases of activities ranging from individual problem solving exercises to entire enterprise (re-)design projects. As a framework for doing 'autopsies' on flawed or failed efforts, the process schema has proven useful in dissecting where things went wrong -- particularly in pinpointing the depictive or referential choices which facilitated failure. By functionally linking the two critical bottlenecks, the model has proven useful for promoting compatibility between the depictional / referential choices made at the depiction bottleneck and the task-oriented factors which govern their acceptance at the enactment bottleneck. With adequate indexing of actors and their task / depictive domains, the process schema can be recursively applied to all levels or aspects of a complex problem-solving initiative. Because it is explicitly cyclical, the process schema has also proven useful in mapping out ongoing chains of such activity in enterprise operations.
These successful applications of the overall process schema are, in my estimation, secondary to its conceptual lesson that the depiction bottleneck is typically the most critical waypoint in such processes. This lesson emphasizes the risks in utilizing IT support for enterprise management. As a result, I have concentrated on analyzing the manner in which IT support tools can constructively facilitate awareness of such risks and provide methods for minimizing them at this critical juncture. The process schema provides a structured argument for the old adage that '90% of a correct decision derives from correctly framing the problem'. The specific line of research I've undertaken with respect to this point -- facilitating presentational techniques for contextualizing problems -- is the subject of other documentation accessible at this site.
