Complexity & Org Science LO5778

Philip C. Anderson (Philip.C.Anderson@Dartmouth.EDU)
21 Feb 96 14:49:49 EST


The INFORMS (formerly ORSA/TIMS) College on Organization Science
will be holding a "conference within a conference" at the society's
national meeting this fall, addressing applications of complexity theory
to organization science. The conference theme is also the subject of a
forthcoming special issue of Organization Science. The College is now
inviting scholars to submit abstracts of theoretical and/or empirical
papers to be presented at the national meetings.

"Complex organizations" have been an important arena in
organizational studies for decades. Historically, organizational scholars
have examined vertical complexity (the number of levels in a hierarchy),
horizontal complexity (the number of differentiated departments), and
spatial complexity (the geographic dispersion of organizational subunits).
Organizational environments have also been characterized as more or less
complex depending on how heterogeneous and dispersed resources are within

However, a different view of complexity is emerging that may have
important implications for organizational scholarship. Within the past
decade, interest in the "sciences of complexity" has increased
dramatically. The study of complex system dynamics has perhaps progressed
farthest in the natural sciences, but it is also beginning to penetrate
the social sciences. This interdisciplinary field of study is still
pre-paradigmatic, and it embraces a wide variety of approaches. Although
it is not yet clear whether a genuine science of complexity will emerge,
it does seem clear that scholars in a variety of fields are viewing
complexity in a different way than organizational scholars traditionally

A number of findings now seem fairly well-established, including
the following:
* Many dynamic systems do not reach an equilibrium (either a fixed point
or a cyclical equilibrium).
* Processes that appear to be random may actually be chaotic, in other
words may revolve around identifiable types of "attractors." Tests exist
that can detect whether apparently random processes are in fact chaotic.
* Two entities with very similar initial states can follow radically
divergent paths over time. The behavior of complex processes can be quite
sensitive to small differences in initial conditions. This can lead to
highly path-dependent behavior, and historical accidents may "tip"
outcomes strongly in a particular direction.
* Very complex patterns can arise from the interaction of agents
following relatively simple rules. These patterns are "emergent" in the
sense that new properties appear at each level in a hierarchy.
* Complex systems may resist reductionist analyses. In other words, it
may not be possible to describe some systems simply by holding some of
their subsystems constant in order to study other subsystems.
* Time series that appear to be random walks may actually be fractals
with self-reinforcing trends. In such cases we may observe a "hand of the
past" in operation.
* Complex systems may tend to exhibit "self-organizing" behavior.
Starting in a random state, they may naturally evolve toward order instead
of disorder.

A number of government agencies have begun to take an interest in
serious academic studies of complex systems. One consequence is that
scholars who are familiar with techniques for analyzing complex systems
but unfamiliar with organizational science have begun to develop models of
complex systems including organizations, and models of organizations as
complex systems. The organization science community has a great deal to
contribute to such endeavors, and stands to gain important new insights
about organizing in return.

The INFORMS College on Organizational Science has decided to hold
a "conference within a conference" on complexity theory at the society's
national meeting this fall. We invite scholars who are interested in
presenting papers to submit four-page abstracts to the fall program chair
by May 15. 15-20 papers will be selected from the pool of submissions.
Papers based on the accepted abstracts will be presented in five
back-to-back sessions at the Atlanta Hilton on November 4, 1996.

Additionally, Organization Science will soon announce a call for
papers for a special issue of the journal devoted to this topic. The
special issue will be guest edited by Alan Meyer (University of Oregon),
Kathy Eisenhardt (Stanford University), Andrew Pettigrew (Warwick
University), Kathleen Carley (Carnegie-Mellon University), and Philip
Anderson (Dartmouth College). Papers presented at the Atlanta meetings
need not be submitted for the special issue, nor is there any guarantee
that papers presented at the conference will be accepted for the special
issue. However, the linkage between the conference topic and the special
issue topic is deliberate. Because complexity theory is a novel topic in
organization science, we hope to accelerate progress by exposing people
interested in this subject to each other's ideas at the conference. We
anticipate that the interchange between people interested in this area at
the conference will stimulate richer, more interesting papers for the
special issue, which will have a submission deadline in early 1997. We
encourage scholars who do not wish to present a paper but who are
interested in the topic to join the learning community we will construct
in Atlanta.

Although studies of complex systems in other disciplines are often
very sophisticated technically, we do not view the conference or the
special issue as a methods forum. The most interesting research into
complex systems sheds fresh light on nonlinear dynamics, which usually
evolve from interactions among agents. Organizational scholars seldom
come to grips with nonlinear phenomena. Instead, we tend to model
phenomena as if they were linear in order to make them tractable, and we
tend to model aggregate behavior as if it is produced by individual
entities which all exhibit average behavior. This conferences focuses on
research that examines complex, nonlinear, interactive behavior within and
between organizations. At this juncture, organizational researchers have
few templates that suggest to them how to hypothesize about or model such
behavior. It is difficult to know how to draw a conceptual model and how
to report the results of empirical inquiries into complex organizational
phenomena. Both the conference and the special issue aim to provide
scholars with useful templates to follow when analyzing complex processes
that involve organizations.

We wish to discourage submissions which are purely methodological;
appropriate manuscripts will examine some aspect of complexity as it
applies to organizations. We also are not receptive to papers that simply
assert that a particular phenomenon is complex or nonlinear, or which
merely call for scholars to take a different view of the world. We
consider the stylized facts listed above to be well-established, and are
not looking for papers that merely suggest these ideas apply to
organizations as well. The question of interest is how ideas that arise
from studying complex systems can add to what we know about organizations.
Examples of appropriate topics might include, but certainly are not
limited to, the following: * Research that specifies plausible sources of
hidden order in apparently random processes that occur within or among
organizations. Can we illuminate how that which appears random is
actually ordered but in complex ways? Such insights are particularly
interesting if they generate testable implications. * Research that
explains how simple organizational processes become complex ones. At what
point do behaviors that are individually well-understood interact in ways
that create difficult-to-understand aggregate outcomes? What are the
consequences of rising complexity in this sense? * Research that compares
several plausible rule sets for a group of interacting agents, and shows
that behavior we observe in organizations can be produced by one model of
interactions but not by others.

A list of useful books and articles addressing complexity theory
may be found at the INFORMS College on Organization Science's Word Wide
Web site: Abstracts should be sent to
the fall program chair at the following address, to arrive by May 15,

Professor Philip Anderson
Amos Tuck School, Dartmouth College
Hanover NH 03755-9000

Abstracts should emphasize why the work is interesting to
organizational scholars, instead of focusing on extensive citations or
methodological descriptions. Authors whose abstracts are accepted for
inclusion in the conference program will be notified by June 1, 1996.


Philip.C.Anderson@Dartmouth.EDU (Philip C. Anderson)

Learning-org -- An Internet Dialog on Learning Organizations For info: <> -or- <>