PARENT SESSION
WA7 - Scale, Structure, and Process in Complex Ecosystems
Chair: McCormick, Ron1, 1 1112 Alexander Avenue, Tacoma, Washington, USA
Co-chair: Johnson, Alan2, Linkov, Igor3, 2 509 Westinghouse Rd, PO Box 709, Pendleton, South Carolina, USA3 Arthur D. Little, Inc., Cambridge, MA
8:00 AM to 12:00 PM - Wednesday, 20 November 2002
Room Ballroom A
(545) For simple systems we can use models, but complex systems must have narratives.
Allen, Timothy*,1, 1 Botany, University of Wisconsin, Madison, WI, USA
ABSTRACT- Most scientists, including eco-toxicologists, have got as far as realizing they use models. Models are the domain of systems that may be complicated, but are still in fact simple by virtue of working at one level. Calibrated models for simple systems are commonly mistaken as objective statements, but they in fact appear solid and clean only because there is normative agreement as to the level of analysis. Scale change causes emergence in what is seen and reported. Models cannot tolerate new levels, so emergent levels in complex systems cannot be modeled. Complexity is not a material matter. Pebbles are simple for kicking, but their rock is complex for geologists studying development of crystalline structure; complexity depends on who is looking and asking. Complex systems have no majority opinion to fix the level, so we must each decide what is structural new not merely behavioral. Also, what is significant not incidental, continuous not discrete, and rate-dependent not rate- independent? These decisions are the hallmark of complexity. Complex systems require a narrator to link levels who accepts responsibility for the spin of the story. When toxicologists put their calibrated models in context, they too become narrators more than modelers. We need some terms here. Levels may be scalar, and relate to levels of observation through observational grain and extent. Do not confuse them with levels of organization whose relationships are prescribed by definition. Entities at organizational levels appear because of observational criteria that put them in the foreground and concomitantly assign everything else to the background. Parts of things are recognized and then linked to give the whole that appears on those aggregation criteria to be the thing in the foreground. Existence at a level assigns things to an equivalence class. Scalar levels of observation are linked to levels of organization by levels of analysis. To the astonishment of social scientist and the humanities, scientists often do not know that they have a level of analysis, let alone what theirs in particular is. New levels may appear over time through emergence, or as a matter of more inclusive or detailed observation. The new level appears as a constraint that orders its lower levels so as to give reliable observables, such as ecotoxicological species or syndromes.
Key words: scale, complexity, emergence, narrative