Basic HTML Version

(1999). The contemporary characterization of emergence requires that emergent properties
have three features: they are supposed to be novel, to be unpredictable, and to display
downward causation.
While the philosophical community has, historically, viewed the existence of this kind
of emergent property as unlikely, contemporary science has re-embraced the notion of
emergence in ways that are really quite striking. You can track this on Google n-grams and
also by a quick Google search. It is my contention that the contemporary scientific
understanding of “emergence” is as a legitimate explanatory category that shares many, but
not all, of the features of the traditional account.
How are scientists describing emergence? Emergence is sometimes characterized as
the “arising of novel and coherent structures, patterns and properties during the process of
self-organization of complex systems.” Goldstein (1999). Thus, self-organization is one of
the mechanisms or modes in which higher orders or properties can be generated. As Karsenti
(2008) said, “Dynamic organization emerges from the collective behavior of agents, the
individual properties of which cannot account for the properties of the final dynamic pattern.”
There is thus a great deal of interest in these higher-level properties. For example,
how do the division of labor and the partitioning of individuals in a social insect colony
work? Why are there so many individuals who are attending their brood and so many who are
foraging? Among the foragers, why are there individuals foraging for nectar and others
foraging for pollen? How do these various features of the structure emerge, how are they
stabilized, and how do we explain them?
The patterns that count as emergent are often those of biological aggregations, such as
schooling fish and flocking birds. They display features that are best characterized by chaotic
dynamics. So these higher-level structural properties invite us to re-examine what we think of
as a legitimate mode of explanation. These higher-level emergent properties have impelled us
to give a different characterization of how explanation works in certain cases. Two examples
will illustrate the changes in the mode of explanation that are required.
E. Coli and emergence
There are many examples in the literature that provide a flavor of what is generating this new
interest in explanatory character. One is chemotaxis and
E. coli
E. coli
is a bacterium that
tracks nutrients in the medium in which it swims. It has a tumbling behavior and a non-
tumbling behavior, and adapts to the chemical gradients of its environment through
chemotaxis. The change of a chemical stimulant induces a rapid change in its tumbling
frequency, which is one of these plastic responses that complex systems are able to produce.
What is interesting about this behavior is that it is possible to change almost all of the
internal components of the bacterium and still get the behavior – there is little change in the
metabolic function of chemotaxis, even if you change many of the genes that code for the
components that make up the bacterium – so it is not a dedicated behavior that could be
explained by the presence or absence of a control mechanism. The presence or absence of a
component does not generate this higher level of behavior; rather, the behavior is a