Abstract: Holism, nonlinear dynamics, sensitive dependence on initial conditions, and self-
organization are examined in an effort to determine whether these concepts have applicability
to research and theory on crime, criminals, and the criminal justice system. It is concluded that
these concepts, all part of a growing science of chaos, show promise of clarifying conflicting
empirical findings and resolving important theoretical issues within the general field of crimi-
nology. The topics covered in this article include empiricism and the problem of prediction, the
reconciliation of polar opposites, reaffirming choice, and learning and socialization. A case
example is used to illustrate how a chaotic interpretation of individual criminal behavior and
change differs from the traditional positivist, classical, and rehabilitative perspectives.
Nonlinear dynamical systems theory, better known as chaos, has been used to
explain such elusive and diverse phenomena as human brain waves (Basar, 1990),
cardiac arrhythmias (May, 1989), ecosystems (Bak & Chen, 1991), international
relations (Grossmann & Mayer-Kress, 1989), and the New York stock exchange
(Corcoran, 1991). Insomuch as chaos appears to have clarified the behavior of
inorganic systems, the next logical step would seem to be an examination of its
ability to elucidate the behavior of organic, in this case, criminal, systems. Being a
systems theory, chaos focuses on relationships between variables and seeks to
explain how these variables interact to form larger systems of influence. This article will consider whether principles borrowed from chaos theory can help clarify
several long-standing problems in criminology. First, however, crime must be
defined and key chaotic concepts described.
DEFINING CRIMINAL BEHAVIOR
There is probably no definition of crime that will satisfy all criminologists. The
present definition is therefore offered with the understanding that it suffers from
obvious limitations. Crime will be defined in this article as a quantifiable behavioral act that violates an established societal norm and is of sufficient severity
and/or frequency to elicit a corrective response from the governing structure of
that particular society. Violations of societal norms that do not come to the atten-
tion of criminal justice authorities are consequently not covered under this defini-
tion. It should be noted that this is a practical, rather than moral, definition, and
one that reflects the fluidity and relativity of human behavior. Laws change, as do
those in authority. Revolutionaries are often classified as criminals by those cur-
rently in power because they threaten the existing power structure. However, if
successful in overthrowing the existing government, the revolutionaries become
the lawmakers and the outgoing leaders become the criminals. Prior to the com-
munist takeover in Cuba, the supporters of Castro were portrayed as criminals by
the sitting Cuban government. However, once Castro assumed power, the sup-
porters of Batista suddenly became the criminals.
The practical definition of crime employed in this article, although it may be
viewed with skepticism by some, is actually in keeping with a chaotic view of the
universe. Chaos theory contends that people are in continual interaction with their
environment. As such, human behavior can only be understood with respect to
this environment. Conclusions based on an analysis of the person independent of
his or her current physical, social, or political environment are artificial and mis-
leading. Criminologists must consequently consider the behavior of the victim
and offender, the actions of the criminal justice system, and the general cultural
context of the interaction before crime can be understood. In the final analysis, no
single definition of crime is capable of capturing the complexity of the act for
which it was intended. Be this as it may, there is still a need to define crime in as
precise terms as possible. Mindful of the fact that any definition of crime will be
incomplete and less than fully satisfactory, the present article defines crime as a
behavioral act in violation of a societal rule, which is of sufficient severity and/or
frequency, that if known to the authorities would, in most cases, be followed by a
negative or corrective response from the societal body responsible for dispensing
criminal “justice.”
THE CHAOTIC PERSPECTIVE
In proposing chaos as a paradigm for criminology, it is important to keep in
mind that chaos is actually a compilation of assorted beliefs, ideas, and principles
connected by a common interest in holism, universality, and the behavior of sys-
tems. Accordingly, it is uncertain whether chaos has achieved the level of concep-
tual elegance and integration to be called a theory. Furthermore, because this
model asserts that order frequently arises out of disorder, the term chaos is some-
what misleading. Whether we decide on the term chaos, complexity, or unpredict-
ability, or call the ideas developed by Lorenz, Feigenbaum, Mandelbröt, and oth-
ers, a theory, a model, or a set of loosely connected ideas, chaotic interpretations
of the universe have had a major impact on the physical sciences. Chaos is actually
the third revolution in physics to take place in the 20th century, having been pre-
ceded by relativity and quantum mechanics (Gleick, 1987), and has all but
replaced the Newtonian view of an ordered universe. Unfortunately, the social sci-
ences have lagged behind the physical sciences in appreciating the nature of cha-
otic phenomena. In fact, many of the more popular models of human behavior still
rely on reductionistic principles borrowed from Newtonian physics. The goal of
this article is to demonstrate how chaotic principles like holism, nonlinearity, sen-
sitive dependence on initial conditions, bifurcation, and self-organization can be
applied to criminal systems, and in so doing address problems that have long
plagued the field of criminology.
HOLISM
Chaos is concerned with the behavior of systems. As such, it holds that the
actions of any single part of the system can only be understood with reference to
the entire system. Owing to the fact that every variable in a system is potentially
capable of influencing every other variable, chaos provides researchers with a
mechanism for evaluating variable interaction, without having to resort to the
cumbersome task of measuring each potentially relevant variable. Proponents of
chaos would argue that the universe is both ordered and disordered, simple and
complex, predictable and unpredictable. Chaos, it would seem, is a science of
contradiction and the reconciliation of seemingly antithetical phenomena and
processes. Whereas Aristotle and his followers were drawn to the separation of
opposites, Newton to the mutual neutralization of opposites, and Freud to the con-
flict between opposites, chaos is concerned with synthesizing opposites. It is the
complementary nature and potential union of opposing forces, according to
Sabelli and Carlson-Sabelli (1989), that stimulate growth and move the organism
toward increased levels of differentiation and adaptability. This dynamic interac-
tion of variables gives rise to what advocates of this approach refer to as system
flow, a process that requires both sensitivity and flexibility (Weiss, 1973). Sensi-
tivity encompasses a system’s overall awareness of the need for change, whereas
flexibility is an estimate of the system’s ability to implement such change.
NONLINEAR DYNAMICS
Linear models of mathematics assume that solutions are contained in multi-
variable equations. Unfortunately, open systems may not operate along linear
lines and so are frequently unamenable to simple linear analysis. Research on
open systems—an open system being defined as one that is influenced by and/or
interacts with other systems—demands a working knowledge of nonlinear
mathematics. Owing to the fact that nonlinear equations are nonadditive, they
often yield more than one solution (Barton, 1994). However, because nonlinear
equations frequently do a better job of modeling the behavior of open systems,
theorists operating out of a chaotic framework generally prefer nonlinear models to the more simplistic linear procedures employed by most behavioral scientists.
The principle of nonlinearity suggests that there is no single cause of an event and
no simple way to conceptualize the behavior of systems.
SENSITIVE DEPENDENCE ON INITIAL CONDITIONS
It has been speculated that a butterfly flapping its wings in Brazil could poten-
tially set off a tornado in Texas a month later (Lorenz, 1979). This seemingly
absurd scenario, although admittedly improbable, illustrates the principle of sen-
sitive dependence on initial conditions. Succinctly stated, this means that seem-
ingly insignificant events can sometimes have a major impact on the long-term
behavior of systems. This is because small differences can become amplified over
time and are therefore potentially capable of radically altering the movement of
the entire system. Sensitive dependence on initial conditions is most commonly
observed in unstable systems like the stock market, the weather, and human
behavior. Also known as the “Butterfly Effect,” in honor of Lorenz’s original flap-
ping butterfly wings analogy, sensitive dependence on initial conditions makes it
exceedingly difficult to predict the long-term behavior of open, relatively unstable
systems.
BIFURCATION
System growth depends on a process known as bifurcation. Around the middle
of the last century, P. F. Verhulst discovered that a population grows in linear fash-
ion until it reaches a certain critical level. The “Verhulst Process,” as it is now
called, occurs when a change in an important parameter drives the system to even
greater rates of growth despite the presence of certain limiting external conditions
(Peitgen & Richter, 1986). As a consequence of this process, population growth
tends to fluctuate between higher and lower values on successive years, a process
known as bifurcation or period doubling. Bifurcation involves the doubling of
points and contributes to the development and transformation of a system’s out-
come field (Feigenbaum, 1978).
SELF-ORGANIZATION
Self-organization is the capacity of a dynamic system to generate new forms. A
system will become unstable when infused with energy sufficient to bring about
changes in time, temperature, or space. However, this instability often gives rise to
system growth owing to the fact that systems possess the ability to self-organize
(Prigogine & Stengers, 1984). Self-organization requires communication
between individual cells for the purpose of establishing a common solution to a
problem, and the creation of a new and more complex pattern of organization.
Systems that vacillate between different values are in a better position to process
information, shift into an alternative mode of activity and adapt, than systems that
remain fixed on a specific value, action, or idea (Packard, 1988). That unstable
systems enjoy greater opportunities for adaptation than stable ones is underscored
by the fact that epileptics exhibit less flexibility and variety in their EEG patterns
than nonepileptics, and heart patients manifest less variation in EKG than healthy
adults (Freedman, 1995).
THE CHAOTIC PARAMETERS
OF CRIMINOLOGIC PHENOMENA
With a rapidly expanding base of information, it would stand to reason that
criminology should be in a position to answer many of the questions that have
plagued the field since its inception. However, with only a few exceptions (e.g.,
Gottfredson & Hirschi, 1990), the theories that have been generated by this prolif-
eration of criminologic data have been of the mini-model variety in which circum-
scribed features or aspects of crime or criminal justice are scrutinized. Although
these mini-models have furnished researchers with new insights and have encour-
aged further data collection, they have also added to the growing sense of theoreti-
cal fragmentation that appears to characterize the field of criminology. Williams
(1984), in fact, takes the field to task for what he sees as its lack of theoretical
imagination. Arguing that criminology has sacrificed innovation and intuition for
methodological rigor and statistical precision, Williams calls for renewed interest
in creative theorizing as a means of understanding crime, criminals, and the crimi-
nal justice system. It is proposed here that chaos may be capable of supplying
criminology with an organizing framework from whence disparate theoretical tra-
ditions and empirical findings might be integrated.
EMPIRICISM AND THE PROBLEM OF PREDICTION
In an effort to determine the “causes of crime,” positivists have taken the
empirical road to knowledge acquisition. Scholars adhering to a positivist
philosophy of criminological investigation assume that absolute or flawless pre-
diction is possible with the proper combination of variables, save the variance
attributable to measurement error. Identifying the variables that predict who will
become delinquent is therefore one of the primary goals of empirical criminology.
Exploring such issues as parental deviance, social class, genetics, intelligence,
and peers, positivists have been able to identify, with modest success, those
individuals who eventually become delinquent. However, as Rutter and Giller
(1984) point out, a substantial number of juveniles classified by investigators as
“high risk” on the basis of their background and behavior never become delin-
quent, and a fair number of youth classified as “low risk” become seriously delin-
quent despite predictions to the contrary. Thus, although certain conditions and
characteristics may distinguish between delinquents who do and do not go on to
become adult offenders, the percent of variance these predictors account for is
