Representation

A representation is a surrogate for reality. The surrogate can be physical or abstract. Physical representations are miniature models of reality, such as a physical 3D model of an amusement park. Physical representations are instrumental in learning about and understanding reality. They provide observers with an overview and enable manipulation of reality [Page 369]that otherwise could not be perceived directly and comprehended holistically. Likewise, abstract representations capture the essence of reality to facilitate analysis, modeling, and understanding. Statistical models summarize parameters and highlight significant characteristics to allow observers to grasp an otherwise seemingly chaotic reality. Mathematical models, for example, are numerical and computational representations to generalize and predict reality. Data models represent reality by formalizing data elements and their relationships. Furthermore, knowledge representation schemes extract logical or procedural rules to guide reasoning and decision making about real-world problems. A robust and efficient representation can stimulate novel approaches to investigation and produce innovative, reliable solutions.

The scale and complexity of reality prohibits direct observation and manipulation. Most scientific and humanistic research relies upon how reality is represented and how the representation is relevant to the issues on hand. Without direct observation of reality in its entirety, geographic representations determine what questions can be answered and how rigorous the answers can be. Fundamentally, any entity, relationship, or idea that cannot be captured by a representation is considered nonexistent and therefore is preexcluded from potential solutions. For example, many GIS packages lack the capacity for representation of time or temporal features. In these nontemporal GIS, temporal concepts, temporal objects, and temporal relationships are nonexistent; hence, there is no information about change, movement, or history in them. Moreover, a nontemporal GIS assumes that the world is static; there is no concept of “before” or “after,” nor is there a concept for “start” or “end.”

Representation is fundamental to understanding and problem solving, since a representation defines the entities and relationships that are recognized to formulate questions and drive solutions. Therefore, determinations of “what to represent” and “how to represent” are critical. The determinations depend upon sequential decisions about ontological, conceptual, logical, and analytical issues. The following discussion outlines the key issues in these considerations. Since all of these issues are scale dependent, the reader is reminded of the need to address representations at multiple geographic scales to capture reality at multiple levels of granularity.