202 Self-Regulatory Perspectives on Personality
For example, two mutually inhibitory nodes cannot both be
highly active at the same time. Thus they constrain one an-
other. Constraints among multiple nodes are settled out dur-
ing the repeated updating of activation levels.
This idea of multiple constraint satisfaction is now having
a substantial impact on how people in social psychology think
about a variety of topics (Kunda & Thagard, 1996; Read,
Vanman, & Miller, 1997; Schultz & Lepper, 1996). It is an
idea that has a great deal of intuitive appeal. It captures well
the introspective sense that people come to conclusions and
decisions not by weighing the evidence, exactly, but rather by
letting the evidence sort itself until it reaches a degree of in-
ternal consistency. The conclusion then pops into mind.
Another term that goes along with this picture is self-
organization(e.g., Prigogine & Stengers, 1984). The idea be-
hind this label is that multiple causal forces which have no
intrinsic relation to each other can cause the spontaneous
emergence of some property of the system as a whole that
does not otherwise exist. The term is used to describe emer-
gent qualities in a variety of scientific disciplines. A number
of people have begun to invoke it as a basis for emergent
properties in dynamic systems (Nowak & Vallacher, 1998;
Prigogine & Stengers, 1984).
Self-Organization and Self-Regulation
Some would argue that models of self-organization in dy-
namic systems represent a serious challenge to the viability of
the type of self-regulatory model with which we began. That
is, it might be asserted that behavior onlyseemsto be self-
regulated—that behavior instead self-organizes from among
surrounding forces, like foam appearing on roiling surf.
Do feedback processes actually reflect self-organization—
a haphazard falling together of disparate forces? Or are there
structures in the nervous system (and elsewhere) in living sys-
tems that carry out true feedback functions? In considering the
relation between the two sets of ideas, it is of interest that
MacKay (1956) anticipated the principle of self-organization
many years ago when he described a system of feedback
processes that could evolve its own goals (see also Beer, 1995;
Maes & Brooks, 1990). Thus, MacKay found the principle of
self-organization to be useful, but he found it useful explicitly
within the framework of a self-regulatory model.
Our view is, similarly, that the concepts of attractors and
trajectories within phase space complement the idea that be-
havior is guided by feedback processes but do not replace it
(Carver & Scheier, in press). There do appear to be times and
circumstances in which forces converge—unplanned—and
induce acts to occur that were not intended beforehand. How-
ever, there also seem to be clear instances of intentionality in
behavior and its management.
It is of interest in this regard that contemporary cogni-
tive psychologists often assume the existence of both bot-
tom-up organizational tendencies and top-down directive
tendencies (see, e.g., Holyoak & Spellman, 1993; Shastri &
Ajjanagadde, 1993; Sloman, 1996; Smolensky, 1988). That
view would seem to fit a picture in which self-organization
of actioncanoccur, but where actions can also be planned
and executed systematically, from the top down. Similar
two-mode models of regulation have also appeared in sev-
eral literatures in personality-social psychology (Chaiken &
Trope, 1999). In short, there seems to be some degree of con-
sensus that human experience is part self-organization and
part self-regulation.
Even when the focus is on planful behavior, the two kinds
of models seem to complement each other in a different way.
The feedback model provides a mechanism through which
goal-directed action is managed, which the phase-space
model lacks. The phase-space model suggests ways of think-
ing about how multiple goals exist and how people shift
among those multiple goals over time, an issue that is not
dealt with as easily in terms of feedback processes.
That is, think of the landscape of chaotic attractors, but with
many different basins rather than just two or three. This seems
to capture rather well the sense of human behavior. No basin in
this system ever becomes a point attractor. Behavior tends to-
ward one goal and then another, never being completely cap-
tured by any goal. The person does one thing for a while, then
something else. The goals are all predictable—in the sense
that they all influence the person—and the influence is highly
predictable when aggregated across time. But the shifts from
one to another occur unpredictably (thus being chaotic).CATASTROPHE THEORYAnother set of ideas that has been around for some time but
may be reemerging in influence is catastrophe theory, a math-
ematical model that bears on the creation of discontinuities,
bifurcations, or splittings (Brown, 1995; Saunders, 1980;
Stewart & Peregoy, 1983; van der Maas & Molenaar, 1992;
Woodcock & Davis, 1978; Zeeman, 1977). A catastrophe oc-
curs when a small change in one variable produces an abrupt
(and usually large) change in another variable.
An abrupt change implies nonlinearity. This focus on non-
linearity is one of several themes that catastrophe theory
shares with dynamic systems theory, though the two bodies of
thought have different origins (and are seen by some as quite
different from each other—see Kelso, 1995, chap. 2). The sim-
ilarity is nicely expressed in the statement that the discontinu-
ity in catastrophe theory reflects “the sudden disappearance of
one attractor and its basin, combined with the dominant
emergence of another attractor” (Brown, 1995, p. 51).