Scriptapedia/Boundary Objects and Failure Modes in GMB

Boundary Objects and Failure Modes (Hovmand, 2014)
In system dynamics and group model building, a boundary object is defined as: ''"A tangible representation of dependencies across disciplinary, organizational, social or cultural lines that all participants can modify. It can effectively advance shared understanding when participants can transform the representation to show more clearly their understanding of the dependencies among them and the implications for each participant’s resources, operations and goals." (Black and Andersen, 2012,195)''

In system dynamics, boundary objects have three essential characteristics (Black and Andersen, 2012). To this list, Black reminds us that boundary objects are social constructions and there are several different ways that visual representations can fail to be boundary objects (Black, 2013 ). In particular, Black has identified three general failure modes for boundary objects:
 * First, they are tangible two- or three-dimensional word sparse representations (e.g., diagrams with few words).
 * Second, these representations are sufficient for participants to show key concepts, actions, and the relationships between them.
 * Third, they are accessible and modifiable by all participants.
 * The visual representation is owned by one stakeholder group or knowledge domain with others deferring to this representation. This can happen when a model is no longer seen as the group’s model, but the system dynamics expert’s model and participants automatically defer to the system dynamics expert’s representation of their issue. Or, the visual representation is dominated by detail from one stakeholder or knowledge domain.
 * Each stakeholder group or knowledge domain develops their own independent visual representation to the exclusion of the others. This occurs, for example, when rather than trying to negotiate and resolve the semantic differences and priorities between different stakeholder perspectives, the group fragments and each decides to build their own visual representation while ignoring other viewpoints.
 * The visual representation covers all stakeholder groups and knowledge domains without being selective in identifying key concepts, actions, and relationships. Similar to the second example, rather than resolving the semantic differences, the group ends up with a visual representation that essentially includes everything.