ORIGINAL PAPER
International environmental law as a complex
adaptive system
Rakhyun E. Kim
•
Brendan Mackey
Accepted: 11 September 2013
Springer Science+Business Media Dordrecht 2013
Abstract Complex adaptive systems are a special kind of self-organizing system with
emergent properties and adaptive capacity in response to changing external conditions. In
this article, we investigate the proposition that international environmental law, as a net-
work of treaties and institutions, exhibits some key characteristics of a complex adaptive
system. This proposition is premised on the scientific understanding that the Earth system
displays properties of a complex adaptive system. If so, international environmental law, as
a control system, may benefit from the insights gained and from being modelled in ways
more appropriately aligned with the functioning of the Earth system itself. In this
exploratory review, we found evidence suggesting that international environmental law is a
complex system where treaties and institutions self-organize and exhibit emergent prop-
erties. Furthermore, we contend that international environmental law as a whole is adapting
to exogenous changes through an institutional process akin to natural selection in bio-
logical evolution. However, the adequacy of the direction and rate of adaptation for the
purpose of safeguarding the integrity of Earth’s life-support system is questioned. This
paper concludes with an emphasis on the need for system-level interventions to steer the
direction of self-organization while maintaining institutional diversity. This recommen-
dation stands in contrast to the reductionist approach to institutional fragmentation and
aims at embracing the existing complexity in international environmental law.
R. E. Kim ( &)
Fenner School of Environment and Society, The Australian National University, Canberra, Australia
e-mail: rakhyunkim@gmail.com
R. E. Kim
United Nations University Institute of Advanced Studies (UNU-IAS), Yokohama, Japan
B. Mackey
Griffith School of Environment, Griffith University, Gold Coast, Australia
e-mail: b.mackey@griffith.edu.au
123
Int Environ Agreements
DOI 10.1007/s10784-013-9225-2
Keywords International environmental law Complex adaptive systems
Self-organization Emergence Complexity Adaptive governance Earth
system
1 Introduction
Complex adaptive systems (CASs) are everywhere. A CAS by definition is ‘‘a system in
which large networks of components with no central control and simple rules of operation
give rise to complex collective behavior, sophisticated information processing, and
adaptation via learning or evolution’’ (Mitchell 2009, p. 13). Examples where CAS
thinking has been useful include ecosystems (Levin 1999; Gross et al. 2006), the Earth
system (Lenton and van Oijen 2002), natural resource management regimes (Rammel et al.
2007; Booher and Innes 2010), environmental law (Ruhl 1997), policy (Emison 1996;
Folke et al. 2002) and governance (Duit and Galaz 2008; Cherp et al. 2011), and inter-
national investment law (Pauwelyn 2013). Despite obvious differences between these
social and ecological systems, complexity theory has provided a common conceptual
framework that bridges the gap between scientific understandings of the two.
In this paper, we investigate the proposition that international environmental law (IEL),
as a set of treaties and institutions directed at reducing human impacts on the environment,
exhibits some key characteristics of a CAS. There are two key justifications for under-
standing IEL in toto as a CAS. First, the subject matters of IEL at all scales, from species to
Earth’s subsystems (e.g., the climate system), display CAS-like properties. Ecosystem
responses to human impacts, for example, are nonlinear, uncertain, and unpredictable
(Levin 1999). Here, the traditional top-down, command-and-control approach is of limited
effectiveness as it is premised on a false assumption of ecological equilibrium (Holling and
Meffe 1996; Folke et al. 2002). Governance of CASs rather requires their control systems to
behave like a CAS in order to be effective (Ashby 1956; Dooley 1997; Ostrom 1999; Ruhl
2008; Ahmed and Hegazi 2009; Duit et al. 2010). What has been proposed as an alternative
model is adaptive or polycentric governance, which is considered to be best suited for
enhancing institutional ‘‘fit’’ (Young 2002; Galaz et al. 2008) with the complex dynamics of
Earth’s social-ecological systems (Holling 1978; Berkes et al. 2003; Walker et al. 2004;
Folke et al. 2005; Folke 2006; Olsson et al. 2006; Ostrom 2010
). This emerging governance
model is ‘‘ecological’’ and draws heavily from complexity theory (Ostrom
1999; Folke et al.
2002; Gunderson and Holling 2002; Duit and Galaz 2008; Duit et al. 2010).
Second, there are good reasons to believe that IEL is already some kind of CAS
(regardless of its effectiveness); hence, it is logical to approach IEL through the lens of
complexity theory. Empirical research has advanced considerably since scholars first began
pondering whether a distinctive system of IEL emerged, not just more random norms about
environmental protection (e.g., Birnie 1977; Kiss and Shelton 1986; Freestone 1994; Boyle
and Freestone 1999; Najam et al. 2004; Bodansky 2006). For example, it has been
observed that multilateral regimes evolve (Bodansky and Diringer 2010; Young 2010) and
to some extent mutually adjust (Galaz et al. 2012b; Kim 2012). Furthermore, Kim (2013)
showed that 747 of multilateral environmental agreements have self-organized into a
complex network, which is far from random. Although it is difficult to prove IEL is a CAS,
it should be useful to draw on the existing analyses and further assess IEL against some key
CAS characteristics and suggest how IEL can be understood as a CAS.
R. E. Kim, B. Mackey
123
Methodologically, the ‘‘IEL as a CAS’’ approach aims to understand how IEL in toto
works and influences the planetary environment (c.f., Decleris 2000;Jo
´
hannsdo
´
ttir et al.
2010). The key unit of analysis is not individual treaties or institutions, but the links that
hold the system together. By filtering details and amplifying macroscopic patterns, we
describe and explain emergent properties that are not reducible to the properties of indi-
vidual components (Gallagher and Appenzeller 1999). This non-reductionist understanding
can be used for developing system-level interventions that would enhance the alignment of
the ‘‘maze’’ of international environmental agreements with the dynamics of the Earth
system as a whole (United Nations Environment Programme 2012). The ultimate purpose of
this exercise is to contribute to ‘‘adaptively managing the complex adaptive legal system to
adaptively manage other complex adaptive natural and social systems’’ (Ruhl 2012, p. 1).
In particular, a CAS perspective holds the key to understanding the relationship between
architecture and adaptiveness, which constitute major analytical problems for Earth system
governance (Biermann 2007). Complexity scientists explain that certain system architec-
tures, in which the components differ and where incomplete connectivity causes modu-
larity, tend to have adaptive capacity (Scheffer et al. 2012). Examples include human
brains, which are optimized for information transmission and rapid adaptation to exoge-
nous perturbations (Sporns et al. 2004; Bullmore and Sporns 2009; Stam and van Straaten
2012). The CAS approach to IEL, therefore, has the power to contribute a theoretical
explanation as to why ‘‘loose couplings’’ of governing institutions are desirable over other
forms (Keohane and Victor 2011; Young 2011; Galaz et al. 2012b; Zelli and van Asselt
2013; Orsini et al. 2013; see also Orton and Weick 1990). In this sense, the CAS lens
allows us to choose appropriate responses to fragmentation of IEL or, more broadly,
institutional complexity (Oberthu
¨
r and Stokke 2011; Zelli and van Asselt 2013).
In what follows, we discuss key features of a CAS and briefly review the scientific
explanation of the Earth system as a CAS. We then consider in some detail how IEL can be
understood as a system of treaties and institutions, which is complex and adaptive. We
conclude by discussing what these imply for the future of IEL.
2 Complex adaptive systems
2.1 What is a complex adaptive system?
According to Meadows (2008, p. 2), a system is ‘‘a set of things … interconnected in such a
way that [they] produce their own pattern of behavior over time.’’ In a system, one can
identify parts, the parts affect each other through flows of energy or information, and the
parts together produce an effect that is different from the effect of each part on its own
(Meadows 2008). It follows that a system must consist of three kinds of things: elements,
interconnections, and a function or purpose.
If ‘‘the collective behavior of [the] parts together is more than the sum of their indi-
vidual behaviors’’ (Newman 2011, p. 800), the system might be complex. If not, the system
is merely complicated (Ottino 2003). Underlying all agent interactions of a complex
system is often simple, deterministic rules. What makes the interactions complex is how
these rules, when set in motion among the diverse components a system, produce nonlinear
relationships including reinforcing and stabilizing feedbacks. Because of the nonlinearity,
local interactions give rise to larger-scale behavior that is not implicit in the parts of the
system. This property of complex systems is called emergence. An example is the stability
of characteristics of the atmosphere (Petit et al. 1999). This appearance of emergent
International environmental law
123
features happens in the absence of an external planner or controller. In other words, no one
designed the system to operate in a particular way, yet it maintains its system identity. This
second defining property of complex systems is called self-organization.
CASs are special cases of complex systems, although the line between them and
complex systems is not clear. For the purpose of this analysis, we define CASs as complex
systems with the ability to adapt to changes in the external environment as a result of
experience via conditional action and anticipation (Holland 1995; Kauffman 1995; Bak
1996; Levin 1999). Adaptation occurs through an autonomous process that uses the out-
comes of local interactions among diverse system components to select a subset of those
components for replication or enhancement (Levin 1998, 2002). Natural selection of
biological evolution is the prototypical example of such an autonomous process. Through
this process, CASs constantly evolve and unfold over time in relationship to the larger
environment in which they operate (Arthur 1999).
CASs are dynamic but exhibit coherence under change (Holland 1995). This critical state
of stable disequilibrium is a hallmark of CASs (Bak 1996). The region where CASs operate or
strive toward is called the ‘‘edge of chaos,’’ a critical transition point between order and
randomness (Lewin 1992; Waldrop 1992; Kauffman 1993; Bak 1996). It is the balance point
‘‘where life has enough stability to sustain itself and enough creativity to deserve the name of
life’’ (Waldrop 1992, p. 12). In the context of global environmental governance, the edge of
chaos essentially is where institutional stability and flexibility or resilience and efficiency
maintain the right balance for effective and adaptive governance (Walker and Salt 2006;
Saunier and Meganck 2007; Duit and Galaz 2008). In terms of system architecture, this point
is the frontier between regular lattices and random networks (Watts and Strogatz 1998 ),
where incomplete connectivity among institutions causes modularity or clustering.
2.2 Earth as a complex adaptive system
Earth as a whole can be considered as a complex system, comprised of many interwoven parts or
subsystems, nonlinear feedbacks with delays, whose dynamics are characterized by critical
thresholds and abrupt changes (Steffen et al. 2004). The Earth system displays emergent
properties that are not fully explained by an understanding of the parts (Lenton and van Oijen
2002). For example, the relationships between greenhouse gases in the atmosphere and the
temperature are not a simple cause–effect relationship, but rather a complex coupling involving
several global-scale feedback loops between the atmosphere, land, ocean, and geosphere
(Steffen et al. 2004). Earth’s climate, therefore, is an emergent property of the Earth system.
In what sense might the Earth system be adaptive? Earth can be understood as com-
prising component ecosystems, each of which is an adaptive system (Holland 1995; Levin
1998). Ecosystems are assembled from biological parts (populations of species) that have
evolved over long time and broad spatial scales (Levin 1998). The collective experiences
of populations of species across a range of ecosystems over time shape the collection of
parts from which the ecological community’s assembly occurs (Levin 1998). But what
about the Earth system as a whole: can it be considered a CAS?
Vernadsky (1998) defined the biosphere in terms of the role the biota plays in modifying
the chemical composition of the atmosphere, ocean, land surface, soil, and substrate.
Consistent with Vernadsky’s early empirically based studies, it is now well established that
the biota play a significant role in Earth’s biogeochemical processes (Steffen et al. 2004).
The Gaia hypothesis (Lovelock and Margulis 1974) proposed that the biota play the critical
role in regulating Earth’s physical environmental conditions and maintaining them in a
condition fit for life. Strong evidence of planetary self-regulation comes from the 420,000-
R. E. Kim, B. Mackey
123
year isotope record contained in the Vostok ice core (Petit et al. 1999), which shows the
regular pattern of inferred atmospheric carbon dioxide, methane concentrations, and
temperature through multiple glacial–interglacial cycles. The tightly constrained upper and
lower bounds of all these variables are a typical feature of a CAS.
Lenton and van Oijen (2002) argued that the biotic dimension of the Earth system
fulfills the CAS criteria of Levin (1998) as it contains sustained diversity and individuality
of components (populations of organisms), localized interaction among these components
(food webs), and at least one autonomous selection process (natural selection). The bio-
sphere (sensu Vernadsky 1998) can be understood as an emergent property of the Earth
system in toto as it represents the consequence of interactions between life and the physical
environment. The Earth system therefore shares the generic CAS properties identified by
Arthur et al. (1997) including dispersed interaction, the absence of a global controller,
cross-cutting hierarchical organization, continual adaptation, perpetual novelty, and far-
from-equilibrium dynamics (Lenton and van Oijen 2002).
Scientific debate continues as to the extent to which biota and ecosystems regulate
versus influence Earth’s environmental conditions, and the relative strength of biological
processes compared with the other physical components of the Earth system, including
those processes that involve exchanges of energy and matter among the Earth’s subsys-
tems. However, the extraordinary extent to which over geological time periods the biota
and Earth’s chemistry have coevolved (Williams 2007) supports the proposition that the
Earth system is complex and, in many ways, adaptive. Irrespective of the precise mech-
anisms by which the Earth system exhibits at least apparent self-regulation, the facts are
that Earth has kept within the general boundaries supportive of life since the onset of life,
the biota has both adapted to and altered Earth’s chemistry, energy balance and climate
subsystem, and our species, Homo sapiens, have evolved and flourished within an even
narrower set of planetary environmental conditions—called ‘‘planetary boundaries’’ or
‘‘safe operating space’’ by Rockstro
¨
m et al. (2009).
With the rise in technology and population growth, humans are now a major forcing
factor on the Earth system (Steffen et al. 2007). The Earth system has been altered by
human societies to the extent that global environmental degradation is evident and plan-
etary boundaries are being exceeded or threatened (Steffen et al. 2004; Millennium Eco-
system Assessment 2005; Intergovernmental Panel on Climate Change 2007; Rockstro
¨
m
et al. 2009). As there is a limit to the resilience of any CAS, if pushed hard or persistently
enough, the Earth system may undergo a phase transition through which a radically new
system architecture is installed, which will then be locked in through a path-dependency
effect. In fact, scientists argue this is indeed what has happened at the planetary scale since
the Industrial Revolution (Steffen et al. 2004). Human actions triggered a regime shift in
the Earth system from the climatically stable Holocene to a new and largely unknown
geological epoch named the Anthropocene (Crutzen 2002; Steffen et al. 2007, 2011).
The view of the Earth system as a CAS has significant implications for the future of IEL
(c.f., Duit and Galaz 2008). Global environmental changes are inherently unpredictable;
hence, our governing institutions need to be sufficiently flexible and able to rapidly adapt
when necessary to, for example, nonlinear changes. At the same time, institutions must be
stable and rigid enough to ensure that humanity stays within the ‘‘safe operating space.’’
The right balance between these contrasting properties is achieved and maintained in a
CAS, and hence the proposition IEL should be designed as one. The way forward nec-
essarily involves gaining a detailed understanding of the existing institutional network as a
system. Without this, well-intended reforms could backfire and undermine current attempts
to create a new, adaptive form of IEL (Galaz et al. 2012a).
International environmental law
123