Chapter 5 - Poietic Planning Systems

In the preceding chapter I described planning as a tool which may enable us to act - or more appropriately - to plan to act sustainably. Given the systems concepts discussed in the earlier chapters (2 and 3), there are two key questions to consider. First, do we recognize systems that are a subject of planning as poietic systems? Second, do we recognize planning systems themselves as poietic systems? The intent of this chapter is to respond to these two questions.

Following the preceding discussion, I consider these questions within the context of sustainability. I continue to use the same examples that have been described earlier.

• In the first section, I consider planning for emergent systems. After briefly discussing the issues, I describe criteria for identification and evaluation of both the systems of concern and the planning systems. I use small groups, science, and Point Pelee National Park as examples.
• In the second section, I consider planning systems themselves. By relating the system concepts to various planning models I generate a typology. The conventional rational-comprehensive model matches autopoietic system characteristics, and many of the emerging models match sympoietic system characteristics. The typology is comparable to the one illustrated earlier distinguishing between normal and post-normal science. When coupled with a conceptual map of system types, this heuristic has the potential to aid planners in choosing appropriate approaches.

An advantage of systems thinking, and of the concepts introduced in this thesis, is that they have the potential to suggest possible directions as well as to provide a framework for critique. In this second section, I consider some of these possibilities. I believe many of the planning approaches that continue to emerge in response to the inadequacies of the rational-comprehensive approach carry sympoietic characteristics and have some ability to address the issues of concern when planning for sustainability. I primarily use examples from National Parks to illustrate some of the points and possibilities.

5.1 Planning For Emergent Systems

In this section, I discuss application of the self-organizing and poietic systems concepts to the systems of concern to planning. There are two key questions. First, do we currently use the autopoietic and sympoietic lenses for interpreting and understanding the systems that are the subject of planning? Second, what are the implications of doing so? I believe the different characteristics of the two poietic system types suggest two different answers to the first question. I believe an autopoietic lens is used, although implicitly, to understand most complex systems. In consequence, the prevalent tendency to misinterpret sympoietic systems as autopoietic, especially in planning with respect to natural systems, means that planning approaches, the subject of Section 5.2, are often inappropriate.

Although I believe it is more appropriate, the sympoietic lens has not been applied for identifying, conceptualizing or understanding the systems of concern. This lack arises from two fundamental human tendencies. First, is the perceptual tendency noted earlier: we identify systems by separating them from their environment - by drawing boundaries. This means we have a propensity toward identifying poietic systems as autopoietic. In opposition to this emphasis, I argue that we also identify systems by recognizing relations rather than boundaries. This is especially true for systems larger than us, such as ecosystems and social systems. Standing within a forest, I recognize it as a system because I perceive the linkages between the different components, not because I can separate the system from its background. Similarly with social systems - I recognize myself as part of a particular culture because of linkages, not because of boundaries. In coping with planning issues, however - and here I consider particularly natural resource issues - it is simpler to define the systems by drawing a line on a map than by articulating relations. As I have explained, the reason for conceptualizing sympoietic systems through recognition of their self-organizing factors is to provide an alternate heuristic. This can make system identification possible without recourse to drawing boundaries.

The second issue that discourages recognition of sympoietic systems is the desire for predictability - another fundamental human concern. "The desire to anticipate, shape, and control the future is a fundamental condition of human existence for individuals as well as collectivities" (Bolan 1996, p 497). This preference encourages a tendency toward identifying systems as autopoietic. As described regarding science - our models influence our interpretations. This points to the second reason for conceptualizing sympoietic systems: to emphasize the uncertainty and lack of predictability that are critical characteristics of so many natural and social systems, yet not recognized when they are interpreted as autopoietic.

In response to the second question - what are the implications - there are again, two answers. For autopoietic systems, with their self-produced boundaries, and their tendencies toward centralized information and control, homeostatic balance, and predictable trajectories, planning in the conventional manner is a useful and adequate tool. In contrast, for sympoietic systems with their lack of clearly defined boundaries, and their tendencies toward distributed information and control, their evolutionary orientation, and their unpredictable trajectories, planning is more problematic. The uncertainties inherent in sympoietic systems make the appropriate link between knowledge and action also uncertain.

5.1.1 Criteria for Evaluation

In another paper (Dempster 1997) I have developed the basics of a methodology for conceptualizing emergent systems that addresses many of the issues and system characteristics described in this thesis. A brief summary of this methodology is presented in Appendix 2. In this section, I take a more general approach, simply listing some of the basic criteria that should be considered with respect to four areas. I describe four sets of criteria, regarding system identification, characteristics to be considered when planning, characteristics of planning systems, and complications that arise from complex human characteristics.

System Identification

The characteristics of sympoietic systems described in Chapter 2 make many of the conventional approaches to system identification problematic. In particular, drawing boundaries, even according to valid, clearly articulated criteria, excludes potentially important considerations relevant to understanding systems. Although this may be the most 'natural' way to perceive things, and although it is quite functional for autopoietic systems, boundaries are not the most appropriate heuristic for conceptualizing sympoietic systems. The complexity of these systems defies bounding. Even on an intuitive level, however, we cannot grasp everything. By relinquishing boundaries we are forced to conceptualize systems in another manner. The alternative articulated here uses self-organizing factors and poietic characteristics of a system instead.

The scope of consideration used to identify and define the system of concern in any planning situation should:

• begin by establishing approximate scale and domain rather than considering boundaries;
• focus on relations and linkages;
• include the self-organizing elements of the system: global-local influences, feedback and recursion, points of tension, and complexity characteristics;
• recognize the emergent patterns and attractors resulting from these constitutive elements;
• recognize the different autopoietic-sympoietic characteristics, especially regarding boundaries, information, and predictability;
• establish system pattern of organization, structure, degree of organizational closure, and the nature of structural coupling;
• recognize the nesting and overlapping holarchies involved, and their self-organizing and poietic elements and consequences; and,
• recognize our limited ability to fully understand any of the above factors.

System Characteristics

As argued above, the characteristics of autopoietic systems tend to be of a type that can be accommodated by conventional approaches applied to understanding and managing situations, especially with respect to natural systems. In contrast, the characteristics of sympoietic systems are prevalent and critical, yet seldom recognized and acknowledged. When deciding how to cope with such systems, that is when developing plans and policies, planners must recognize both the restrictions and potentials of sympoietic characteristics, including those on the following list. Most of these are simultaneously advantages and disadvantages, and some appear contradictory. Balance is critical.

• Systems are constrained onto particular attractors by self-organizing interactions. Attractors represent possible manifestations of the rule set, but it is essential to recognize that not everything is possible.
• Sympoietic systems maintain a dynamic, rather than homeostatic, balance. Change - possibly even dramatic and surprising change - is an essential part of system maintenance and evolution.
• Sympoietic systems have a limited amount of predictability.
• Small changes can lead to big changes due to the reinforcing effect of positive feedback.
• Since the pattern of organization and information content depends on past relations, history has momentum, setting restrictions on a system's future trajectory.
• Different types of information are distributed among components so collaborative/cooperative (rather than central/hierarchical) control is an integral part of system function.
• Since the systems are organizationally ajar, environmental uncertainty provides a critical advantage by increasing a system's adaptive potential.
• The systems are organizationally ajar, not open, so incorporation of information into a system will be restricted by internal structural coupling.
• Systems are structurally coupled to their environment/context. It is essential to understand how coupling occurs and how fixed and strong it may be.

Planning Systems

The two preceding sets of criteria are targeted toward conceptualizing and understanding the systems of concern. I have noted that planning systems can also be recognized as poietic systems. Although these will be covered in the next section, I note that both of the above sets of criteria should be considered in the identification, characterization and potential behaviour of the planning systems themselves. In addition, for planning systems it is also essential to:

• recognize the degree of connection between the planning system and the system of concern;
• recognize the learning potential in both; and
• recognize the degree of connection between the planning system and other disciplines with knowledge relevant to the system of concern and to the planning process.

Human Complexities

In addition to the above factors, the increased complexities of human systems and the subsequent problems and potentials provide additional concerns that are relevant to both the systems of concern and planning systems. These arise from the characteristics described in Section 3.2.1.

• The potential for interpretation - recognizing structures without direct contact - also involves the potential for misinterpretation. Biased perceptions lead to inappropriate understanding, yet are inescapable.
• Enhanced adaptability from the potential for learning, especially from the experience of others or from symbolic representations, is hampered by the possibility for learning to be based upon inaccurate information.
• Learning also provides the potential to develop complex knowledge systems but these may allow detachment from the environment.
• Creativity allows us to design objects or systems that have never existed before, but it may also allow us to believe in, or attempt to create, impossible things.
• Self-awareness provides the potential to recognize the inherent difficulties and uncertainties involved in these various situations, but it is not always used. Even when it is used, it does not eliminate the uncertainty.
• Self-awareness may make us want to control that which perhaps should not be controlled.
• Our apparent ability to comprehend systems simpler than our individual selves, or the social systems we are components of, does not suggest that our social systems are the most complex systems in existence. There is no logical or rational argument against the possibility of infinitely more complex systems, beyond our ken. These system likely influence ours by providing directions and constraints.
• The intentionality possible in human systems, does not negate the inherent restraints and potentials of the self-organizing process and the consequent uncertain, unpredictable, yet often recognizable, patterns.
• As planners we are enmeshed in the paradox of future causality. By guiding intentional action we create directions and constraints, generating consequences, both planned and unplanned, that shape the systems of concern for the next cycle of planning.

5.1.2 Examples

To consider the implications of the systems concepts for planning further, I return to consideration of the social system examples used in the chapter on social systems. These three quite different systems allow for drawing attention to different considerations. I will not explicitly consider application of the evaluation criteria here, especially the first set, since I have characterized these systems as poietic in the foregoing discussion. Instead I will consider various other factors involved in attempting to plan with respect to such systems.

Designing Small Groups

Much of the research regarding small groups has been aimed at improving their functioning in a variety of situations. Empirical observation of characteristics such as the matching of leadership style and level of group maturity illustrate aspects that can be used for designing functional groups. When considering such a possibility, the complex human characteristics identified above become important. In this case, they are not only relevant regarding behaviour of small group members, but regarding small group designers - although the latter are often also members. I will concentrate on implications of the human characteristics regarding design.

I follow a somewhat cyclic process which begins with the existence of a particular small group. Cognition allows observation and interpretation of the factors involved in generating and maintaining a small group. Imagination allows creation of new possibilities based on ideas about what is desired. Intentional action manifests the ideas through changes to the group or formation of new groups. Self-awareness allows reflection on successes. Finally, learning allows integration of new experience into subsequent small groups.

If this cycle involved in developing small groups was so straightforward, however, we would likely have many more successful small groups. The difficulty is that, as noted, these human capabilities involve disadvantages as well as advantages. To illustrate the complications, I use the poietic concepts of pattern of organization and structure. By making a distinction between actual, interpreted, and designed, patterns of organization, and recognizing that actions must be mediated through structure, some of the difficulties can be pointed out. This interpretation is illustrated in Figure 5.1.

By the 'actual' pattern of organization I refer to the 'natural,' or undesigned set of relations that defines a small group. It includes all of the various aspects described as well as other possible factors that are unknown or unrecognized. This pattern of organization will be manifest in a particular structure: a small group. It is important to recognize that this structure is the only aspect of the small group available to our senses - we can only infer the pattern of organization. By observing these structures, then, we develop an interpreted pattern of organization - our 'best guess' as to what the actual one is. In essence, this is what research regarding small groups is about, and reflects the purpose of research for other systems also. From our interpretation we design a pattern of organization that we believe would better achieve our desired results. Note that this must include design considerations as well as the articulation of values regarding what is desired. We then manipulate the structure, attempting to manifest the desired effect or system, either by altering components and relations or by altering self-organizing influences. For example, we bring in a new leader, alter the amount of time spent on developing interpersonal relations, or encourage a focus on defining norms internal to the group.

Whether structural manipulations are successful depends on at least three considerations. First, it depends on our success at manipulating the structure: do our actions have the desired effect? For example, does the new leader help? Second, continued success also depends on whether the structure created can produce an actual pattern of organization that is able to self-produce. Are the newly established interpersonal relations of such a different character that they alter the pattern of organization? Success of these will depend on the self-organizing factors involved, and the degree and type of recursion occurring. If the new structure/organization is not self-producing, continued intervention may be required. Third, success depends on the nature of the structural coupling: do we create a structure that still couples to the context? Can new information requirements be met by the paradigm of the context? Are the newly created norms so different that the new system can no longer interact with its context?

Altering the self-organizing influences can also have an effect on the small groups structure that arises, holding the potential to set the system onto a new and different attractor. In addition, since the systems are recursive, changing the structure may also change the self-organizing factors and the pattern of organization. Even in small groups the point of tension is difficult to find and maintain, considering all the factors involved (Figure 3.6). In more complex systems it is even more difficult as will be considered in the next example.

Science

In Chapter 3, I discussed Kuhn's (1970) interpretation of the evolution of science, using the distinction between autopoietic and sympoietic systems to parallel normal and revolutionary science, respectively. Opposed to the application of normal science in situations involving high risks and uncertainty regarding values and/or regarding system complexity, Funtowicz and Ravetz (e.g. 1992, 1993) advocate the development of post-normal science. I noted that this is also a sympoietic type of approach to science - and also a contrast to normal science. The distinction between the sympoietic systems in the two different situations - revolution and post-normal - highlights an important, though perhaps tenuous, difference. It relates to the nature of the self-organizing process involved in their generation and also the human potential for intentional action. Revolutionary science - a sympoietic mode of inquiry - arises 'naturally,' without intentional design, from the self-organizing factors generating science described above. It is a key stage in the evolution of science as described by Kuhn. In contrast, post-normal science - also a sympoietic mode of inquiry - is being advocated intentionally since it has not arisen 'naturally' from the process of inquiry. In effect, generation of such a system of inquiry would be a planned or designed process. The question, then, is how can such a process be encouraged? Can we apply the systems concepts described in this thesis to achieve such an end?

I use this example as a contrast to the preceding one in which the potential for effecting change seemed relatively straightforward, relating to alteration of the system structure. In this case, I believe different tactics are required. These systems are recursive - there is no single point of entry - change must occur on many different levels. In particular, I believe three key aspects of the system must be considered. First, the self-organizing factors, especially the global-local influences, offer a potential leverage point. Second, the sympoietic nature of the system advocated allows for change on an organizational level. This allows for change, at least from the theoretical perspective, through opening boundaries and through developing linkages to different knowledge interests and communities. The reality of this possibility is limited by the third point: structural coupling of the system to its context must be recognized. Normal science is an integral and respected system within western industrial society. Changing the role, and nature, of science, requires changing the linkage and expectations of society as well.

Funtowicz and Ravetz concentrate primarily on the second consideration by emphasizing the need to expand the peer community, bringing interests and concerns typically external to the validation process into consideration. Although I will not discuss these issues here, I suggest that approaches relevant to planning articulated in the last section would be of value in understanding how this may be done. With respect to the first consideration, it is necessary to reconsider self-organizing factors noted in Figure 3.9. The most likely candidates for making alterations are ontological and epistemological influences and methods. I believe some sense of change is already occurring as illustrated by increasing recognition and attempts to incorporate different types of knowledge and by post-modernist critiques. The key point to emphasize in this example is that change will be most effective when it simultaneously occurs at many different points.

Point Pelee National Park

In the preceding chapter, I used the systems of concern for Point Pelee National Park, in southern Ontario, as an example to explain the systems concepts in reference to both the geophysical and social systems. Here I integrate the examples, commenting on the implications for planning with respect to the Park and the issue of shoreline hazard management. Before considering this particular park, however, I draw attention to three problematic issues that are of continuing concern for park and protected area planning in general, to indicate that in many cases sympoietic systems have been misinterpreted as autopoietic. With each issue, misinterpretation has led to inadequate management of park resources. This is also true for other natural resources ranging from commercial forests to fisheries.

• Cross-Boundary Issues There are many examples, in Canada and around the world, of parks and protected areas being isolated as islands in significantly altered landscapes: wildspace encircled by development. Point Pelee National Park, with a barrier constructed to its northern end, is a classic example. Riding Mountain National Park in Manitoba, which appears on aerial photos as a dark rectangular patch of forest in a wide expanse of agricultural land, faces the same difficulties. Another example is Pacific Rim National Park Reserve, a thin strip along the west-coast of British Columbia which has clearcut logging to its boundaries and which is also influenced by the ocean, an unprotected area.

• Biodiversity The notion of sympoiesis helps emphasize recognition of biodiversity as a source of adaptive potential. This points to the danger of reification: considering biodiversity as a thing (Galindo and Bunnell 1995) or as an amount of something, such as species. Biodiversity represents the variety of information - the differences among biological systems (Wood 1994), including systems of different types at different scales. It is this variety that provides the balance between uncertainty and redundancy that is critical to the adaptability of sympoietic systems.

• Integrity and Health Recognizing the distinction between autopoiesis and sympoiesis is of particular importance for understanding ecological integrity. The prevalent comparison of ecosystem 'health' with organism health (e.g. Rapport 1989, Costanza 1992) is an example of misinterpreting sympoietic systems as autopoietic. The difference between homeostatic and evolutionary dynamic balance illustrates the inadequacy and potential danger of such a comparison, supporting the notion that it is problematic (Ryder 1990, Suter 1993, Wicklum and Davies 1995). The requirements for being 'healthy' involve different mechanisms in the two system types: 'restrict change' for autopoietic systems and 'let change happen' for sympoietic systems. Using different terms - health and integrity, for autopoietic and sympoietic systems respectively - can help to ensure the difference is recognized and appreciated.*1

These three issues are critical issues for parks and protected areas and must be considered in their planning and management. With respect to the discussion of Point Pelee, I focus primarily on the first issue through consideration of the geophysical system rather than biological systems. The former is one of the most basic sustainability issues for the park since it relates to the persistence of the peninsula-marsh structure itself. At an earlier time, the approach taken involved significant intervention such as creation of barriers to protect the shoreline from erosion. Recent approaches have reduced the amount of intervention, recognizing the need to let 'nature take its course' (Parks Canada 1991).

The description of Point Pelee in Chapter 3 illustrated that there are factors well beyond the park boundaries that are necessary considerations for appropriate planning. Figure 5.2 illustrates factors similar to those in Figures 3.13 and 3.14, using a different conceptualization in order to emphasize self-organizing aspects. Figure 5.3 is redrawn from Figure 3.16. The pair of Figures (5.2 and 5.3) indicate some of the factors involved in generating the geophysical and social systems relevant to Point Pelee National Park. Figure 5.4 illustrates the autopoietic and sympoietic characteristics of these systems. The points made regarding cross-boundary issues are critical for Point Pelee since the park itself obviously does not cover, spatially, all of the factors noted as relevant.

Understanding these social influences and the self-organizing factors governing social systems is essential for making informed management decisions.

Table 5.1 lists some of these elements and actions of the management system that could possibly influence social elements which impact the geophysical system. Although many social elements impact natural systems, those listed here include only a few illustrative examples related to shoreline erodability. While manipulating structural factors may have short-term immediate impacts, to effect real change it is essential to also alter the underlying factors. We often 'manage' natural resources by altering structural components in a manner similar to that illustrated with the example of small groups. However, the critical question, especially with respect to large, complex natural systems, is how we influence their ability to continue their self-organization or self-production - to be sustainable. With this question in mind, however, it is imperative to recognize the normative issues at stake and the potential for systems to flip onto a different attractor. The question is not whether the system can continue its self-organization or poiesis, but whether it continues in a manner according to our preferences in any particular situation. In this situation, given the presence of a national park system with a dual mandate for maintaining ecological integrity and providing recreational opportunities, continued existence of the peninsula-marsh structure is the stated preference (Parks Canada 1991).

In consideration of this preference then, I emphasize the need to 'manage' the factors influencing the system, in particular those governing the self-organizing process in social systems. In this example, the potential for the park planning and management system to influence the self-organizing factors relevant to shoreline hardening must be considered. Figure 5.3 noted the point of tension among these factors. Figure 5.5 is a re-illustration to indicate the potential effect of social changes on this point of tension.

Shifts back toward the natural condition could be effected by increasing the cost of intervention, creating appropriate regulations, and encouraging an environmental ethic. Alleviating fear, decreasing the emphasis on personal property rights, and altering the orientation toward progress could also have an influence. Indicated on the diagram are changes that could be promoted by actions of the park management system primarily through education. Beyond this, it is again essential to recognize the role of structural coupling. For example structural changes such as creating and documenting new regulations will not be effective without the social or political will for adhering to and enforcing these regulations.

These points emphasize that the 'knowledge' relevant to planning for a National Park such as Point Pelee must incorporate many aspects that are not necessarily considered as critical to planning and management of an ecological area. Further consideration of these aspects and the potential to plan for them will be covered in the next section on planning systems, where I will reconsider Point Pelee as an example. The argument will be made that development of sympoietic planning systems can help to cover this wide territory.

5.2 Emergent Planning Systems

The final aspect of planning to cover is the central aspect: process and procedure. This is the area most typically associated with planning 'theory,' in part because it is the area of research interest that relates almost exclusively to planning. The various planning models discussed in the literature (rational-comprehensive, transactive, participatory, radical, and others) refer to planning procedures. It is with respect to this aspect of planning that the tension between theory and practice, and between description and prescription, is perhaps the strongest and most contentious. Confusion regarding the latter was noted in the preceding chapter. While some claim that how planning is done indicates how it should be done (e.g. Scott and Roweis 1977), others claim that there are gaps between descriptions and prescriptions wide enough to question the usefulness of the former and justification of the latter (McClendon 1995). I believe the confusion arises from the fact that both 'sides' are the purview of 'planning,' from lack of attention to the is-value-ought connection, and from lack of concern for the emphasis on learning.

I begin this section by considering the rational-comprehensive model, noting some of the difficulties involved in its application. I then describe some of the alternatives, including participatory and adaptive models. A conceptual map, linking systems types and planning models, illustrated in Section 5.2.2, provides a heuristic for choosing appropriate approaches. I close the section by using a range of examples to illustrate the issues.

Rational-Comprehensive Planning Model

The rational-comprehensive model (e.g. see Alexander 1992, Patton and Sawacki 1993) was developed to be precisely what its name suggests: rational and comprehensive. Both are attractive qualities for dealing with the intricacies of social issues, and the natural-social system interface. As noted in the preceding chapter, the approach arose from the control oriented, engineering paradigm, which reflected the social context of the period (Friedmann 1987, 1996, Wolfe 1989). This prescriptive model is presumed to enable the objective application of knowledge to determine the best solution in any particular planning situation. Such theoretical abstraction regarding desirability of application, however, is no guarantee that the procedure will - or even can - occur in practice (Alexander 1992). Wolfe (1989, p71) notes that "even as the rational model was being formalized and incorporated into planners' daily lives it was being questioned, modified, amended, and has come to be seen as an idealistic guide to action which is not attainable but nevertheless useful."

Despite its repeated failure to generate expected outcomes, the rational-comprehensive approach maintains a central role in planning (Wolfe 1979, Alexander 1992, Friedmann 1995). Many authors (e.g. Hudson 1979, Wolfe 1989, Alexander 1992) point to the simple logic of the approach, the fundamental nature of the steps, and the attractiveness of its touted qualities, as the key reasons for its continued desirability. Despite the many alternate models, it is difficult to argue against at least trying to follow the dictates of the rational-comprehensive approach. Based on the systems concepts discussed in this thesis, however, I believe such an assumption can be questioned, giving rise to four issues.

• The first issue relates to the self-organizing factors governing social systems. The steps of the rational-comprehensive model impose a direction on the planning process that acts as a global influence. Through interaction with local influences, potentially unexpected and unwanted structures emerge. The process is neither innocent, nor neutral - outcomes arise because of its application, not despite its application. Campbell's point regarding a system's capacity to fulfill future visions is a relevant concern. A difficulty with the rational-comprehensive model is that a discrepancy between the capacity, and belief in this capacity exists. I will discuss this concern below.
• The second issue relates to the autopoietic characteristics of the model. These characteristics make such a planning system less adaptable, offering little potential for coping with the many uncertain situations that planners and planning faces.
• Third, the model tends to advocate a linear, problem-solving approach, yet the systems of concern tend to be complex, non-linear, recursive systems. Multiple points of entry are not only possible, but - considering the effect of structural coupling - are recommended.
• Finally, the recursive nature of poietic systems and the importance of structural coupling suggests that once we begin using a particular approach, it may become entrenched and difficult to change - even if such change is recognized as desirable. I believe the latter reflects our current situation. As Bolan (1996, p 498) states, "Difficulties stemming from deeply flawed institutional arrangements cannot be cured by a change in leadership or an improved analytic technology." Yet we are faced with "problems that are so deeply grounded in institutional arrangements that it is impossible to 'solve' them without redesigning those arrangements" (ibid. p 498).

A planning system acting according to the rational-comprehensive model, presuming autonomy, organizational closure, homeostatic balance, and predictability, exemplifies autopoietic characteristics. It applies a command and control logic, highlighting experts and assuming that accurate and complete information can be applied in a rational manner to determine the best outcomes. The difficulties in realizing the model as designed arise because these assumptions are inaccurate except in the simplest situations. The systems of concern typically carry sympoietic characteristics. Critical information is hidden spatially (among many components), temporally (in the future as well as in the past), and in the increasing and decreasing levels of complexity. Ignorance and indeterminacy (Wynn 1992) are inescapable factors since the systems are evolutionary and unpredictable. Logic, rationality, and centralized and hierarchical control, are not functional tools for coping with these systems. Their application cannot be expected to generate predictable outcomes, even under the most 'ideal' circumstances. Lindblom's (1959) observations, first described in a paper titled "The Science of Muddling Through," illustrates the actual outcome from application of the rational-comprehensive model.

It is instructive here to consider the definitions of system pattern of organization and structure, and the problems pointed out when considering the design of small groups. As with social systems, I consider planning systems to be composed of socially defined components: actors (planners, clients, subjects), knowledge (about actors, situations, systems of concern, planning processes), institutions (bureaucracies, regulations, conventions, norms), processes, and other components as well as the relations among them. I believe it is critical to understand that we do not have access to the actual pattern of organization of any systems, but that we are always working with interpretations and designs. In addition, we can only intervene through structures and, in some cases, through self-organizing factors. The actual pattern of organization, and subsequently the structures that eventuate and the success of the poietic process (i.e. the sustainability of the system) are beyond our direct control. This does not mean that they are beyond our influence or impact as indicated by the sustainability concerns discussed throughout.

Figure 5.6 illustrates the difficulties involved in application of the rational-comprehensive model using poietic system elements. Part a) is a simplistic representation of the recursive planning process from the conventional perspective. Part b) incorporates the complexities of the poietic process. (Compare to Figure 5.1 of small groups.) To follow a generic sequence, take the initial position as a particular planning process - a structure emerging from some undetermined self-organizing factors expressed by a particular pattern of organization. Observing this process, theorists/researchers perceive a disjointed, uncontrolled planning process: their interpretation of the pattern of organization believed to give rise to the perceived planning structure. By imagining other possibilities and preferring more ordered, controlled, better informed approaches, a new process is designed: the rational-comprehensive model. By manipulating structures, an attempt is made to alter the original process or to generate a new one. Due to complications introduced by structural and self-organizing influences, however, the resulting process does not match the targeted rational-comprehensive model. Instead, it fits the description of incremental planning.

Due to the potential for misinterpreting sympoietic systems as autopoietic, belief that the process does reflect rational-comprehensive characteristics can be sustained. Admittedly, belief that it is not sustained is just my interpretation! However, as noted above regarding other situations, one of the advantages of distinguishing between the two poietic systems types is the potential for drawing comparisons. Even though ideal systems - either end of the system continuum - may not exist, the characterizations provide a heuristic for recognizing differences. Without an alternative lens to look through, the outcome will be interpreted through the autopoietic lens, and an autopoietic system (though possibly a hard to fit representation of one) is all that it will be possible to perceive.

Our ability to act intentionally allows us to decide according to beliefs, even if these rely on inaccurate representations. The consequence here is incrementalism - the 'muddling through' described by Lindblom (1959) - the outcome of repeatedly applying an autopoietic approach when a sympoietic one may have been more suitable.

Two further points need to be made here. The first, relates to the notion of structural coupling. As illustrated in Figure 5.6, both the interpreted and designed patterns of organization are structurally coupled to their conceptual context. As indicated by the discussion of science in Chapter 3, paradigms carry significant weight in formation of perceptions and interpretations and also in the creation of new ideas. The social influences, described in the same chapter, emphasize the potential for such paradigms to be shared and entrenched. The conceptual context will be reflective of particular paradigms. The interpretations and designs will be coupled to, and hence, reflective of, this context. In this case, and as noted, the rational-comprehensive model is coupled to the social context which prioritizes experts (especially scientists) and rationality. The second point, is to recognize the self-organizing factors influencing the structures that eventuate. Decisions made through application of the rational-comprehensive process, by setting directions, will generate patterns through their interaction with social constraints and positive feedbacks. These patterns, in turn, create particular, potentially surprising, and eventually entrenched, outcomes. It is essential to recognize that these outcomes arise because of our intentions, not despite them.

The impacts that emerge from these interactions can be not only surprising and undesirable, but subtle and long lasting. For example, difficulties surrounding the townsite of Banff situated in a National Park, arise from an initial vision of developing the area to accommodate public use. Wilder (1993) cites creation of the three-mile limit in the United States by Washington and Jefferson, in 1793, as an incremental decision that continues to have a problematic influence on marine policy today. Planning efforts with respect to the automobile provide another prevalent, ubiquitous, and problematic example. Development of freeways, suburbs, and shopping malls planned through rational-comprehensive approaches, that (with hindsight) can be seen as neither rational nor comprehensive, have created landscape patterns, personal habits, and dependencies that are difficult to alter (see e.g. Richmond 1995).

In each of these situations, decisions were made based on images of the future by a planning system with the capacity to influence the systems of concern. Based on an original decision that reflected a minor change from the status quo of the time, a series of similar, incremental decisions have acted as positive feedback. These and other self-organizing factors, have led to 'unplanned' outcomes. The systems are future causal, but not necessarily successful at attaining the prerepresentation envisioned.

The third example noted leads me to question: what freeways are we building now? Consider the following.

• Whittaker (1996, p 16) notes that "electronic communication promises to replace some of the need for travel for business, entertainment, shopping, and education, giving businesses greater latitude in choosing locations." What patterns and social problems will eventuate from 'rationally' and 'comprehensively' planning with respect to these possibilities?
• Richmond (1995, p 306) is concerned that the "fascination with computers is becoming ever greater, as is the tendency to formulate planning problems around what a personal computer can do rather than based on the essence of the problems themselves... Planning problems, in turn, are modeled to give the appearance of being soluble by such devices. The need for "computer literacy" is being stressed while too many students are illiterate in their own language and unable to use it effectively to structure thought." What are the implications of these notions given the recursive nature of planning - especially if there is continued belief that such problems can be 'solved' using the rational-comprehensive model which lends itself to problems 'solvable' by computer?
• Roszell (1996, p 33-4, and noted above) reassures us that, with respect to the management of National Parks, "business is not our business but the means to our ends." However, McNamee (1993, p 17) observes that "evolution of the national parks system... has been influenced more by the nation's focus on economic development and prevailing social values and less by the need to preserve wilderness." What is the potential for negatively influencing ecological integrity through small incremental changes that go unnoticed since they are set within a process that is presumed to be 'rational' and 'comprehensive'? What effect does adding another self-organizing global influence - cost recovery - have on the planning system emerging from these influences? What potential is there to change a national parks bureaucracy when it becomes more tightly structurally coupled, through a commonly held business culture, to political and economic systems providing a context?

Implications of the significant and pervasive influence of these types of directions as self-organizing factors cannot be predicted, but should not be underestimated. In addition, the recursive nature and the ratchet effect common in social systems should not be ignored. As Bolan notes:

Social structure refer[s] to the patterning of interaction plus the continuity of the interaction or, in short, the reproduction of the patterning... Rules themselves have a dialectical quality: they are simultaneously the medium for the production and reproduction of social practices and the outcome of those practices (Bolan 1996, p 504).

More accurate interpretation and understanding of system characteristics and of their self-organizing potentials is critical for considering the possible impacts of the planning directions we pursue and for considering what approaches may be most appropriate.

Other Planning Models

As a contrast to the rational-comprehensive approach, a variety of other planning approaches have arisen. Some - advocacy, transactive, participatory models - have been developed as a challenge to the inadequacies of rational-comprehensive planning as a prescriptive model (Hudson 1979, Friedmann 1987). Others - satisficing, muddling through, disjointed incrementalism - are primarily descriptions of "how decisions are made in the real world" (Wolfe 1989, p 71). Although I believe application of these systems ideas would be beneficial for understanding the latter decision-making processes, my interest lies with prescriptive models, so I will not include any further discussion of descriptive approaches.

Taking my lead from Funtowicz and Ravetz's typology of scientific approaches (Section 3.2.4), Figure 5.7 is a parallel diagram illustrating three zones of prescriptive planning approaches. The axes represent different characteristics of planning situations that arise. They focus on two different types of uncertainty - one reflecting values and the other reflecting knowledge. The resulting zones must be recognized as broad, loosely defined categories. In addition, it is critical to note that outer zones incorporate inner zones. Adaptive approaches rely on participatory models, and neither of these preclude a reliance on some of the basic stages involved in rational-comprehensive planning. This typology will be useful for considering appropriate choice of models relevant to particular situations. I will return to these considerations in the following section, and will also note similar methods of choice such as mixed-scanning.

• Zone 1: Rational-comprehensive model Due to the characteristics outlined above, this model is most applicable where certainty regarding knowledge and technology relevant to the planning situation is greatest and complexity regarding values and goals are lowest.
• Zone 2: Participatory models This zone includes a variety of approaches which place emphasis on the inclusion, into the planning process, of actors relevant to the planning situation. Contrasting with the rational-comprehensive model, which tends to rely on experts, these approaches recognize plural values, encouraging involvement of stakeholders in decision making. By including more diverse interests and participants, they have a greater ability to cope with uncertainties. The most well known and long standing models in this zone are transactive planning (e.g. Friedmann 1973, 1987) and advocacy planning (e.g. Davidoff 1965). The former model promotes a view of planning as social learning. Advocacy planning is based on a concern for publics and interests under-represented by the traditional distribution of power and approaches to planning. It promotes the idea that planners should emphasize these disadvantaged groups.
  There is also a wide, and continually expanding, variety of participatory models being developed on a somewhat ad hoc basis. Found in many areas of planning, from social to environmental issues, they primarily arise as a response to the public demand for more inclusive processes. Examples include BCRTEE (1991), Dorfman (1991), CORE (1992), Law and Hartig (1993), Deimer and Alvarez (1995), Johnson and Duinker (1993), and Brown (1996).
• Zone 3: Adaptive models "Adaptive management" has been written about, promoted, and - at least to some extent in some cases - applied for many years (see, for example, Holling 1977, Walters 1986, Lee 1993, Hennessey 1994, Gunderson et al. 1995). "Adaptive planning," however, is a concept that has yet to be clearly articulated. At this level, however, the distinction between planning and management, which is already somewhat fuzzy, becomes more so. One of the basic ideas behind adaptive management is to generate learning opportunities by incorporating experimentation, monitoring, and re-evaluation into management at the design stage. Such an approach must necessarily involve planning. In a sense, adaptive approaches reflect a merger of planning and management. I articulate the notion of adaptive planning in order to emphasize the importance of incorporating flexibility and responsiveness at the decision-making and design stage. Being adaptive only at the 'management' stage is inadequate for coping with the complexities of the situations we face, since management in more often associated with implementation than with planning. This is especially the case with respect to the rational-comprehensive approach, a comment I return to below. The adaptive model encourages a more integrated approach.

Since the objective of such approaches is to build in the potential for adapting as necessary, they are the approaches most suited to situations involving high levels of uncertainty and a diversity of values - the outer edge of the typology illustrated in Figure 5.7. This make such approaches relevant to the conditions prevalent in sympoietic, and social systems. I re-emphasize that, as with Funtowicz and Ravetz's typology, the outer zones incorporate the inner zones. Such approaches will still rely on participatory, and possibly even rational-comprehensive, approaches.

There are some models described in the literature that illustrate characteristics with potential to be adaptive. Nelson's civics model for planning (e.g. Nelson 1993b, Nelson and Serafin 1993, 1996), and strategic land use planning models applied in British Columbia (e.g. CORE 1994, Brown 1996) provide two key examples that will be discussed in further detail below. Gunderson et al. (1995, especially Westley 1995) provide useful descriptions that lean toward prescriptive considerations. Others describe responsive and adaptive characteristics of individual planners, rather than of planning systems or models (e.g. Friedmann 1993, Various 1995). Consideration of these characteristics may also be useful. Teubner et al. (1994) provide an edited volume that brings together discussions focused on encouraging the generation of self-organizing systems in legal and auditing systems, an approach that parallels some of the discussion below. The literature on strategic planning and organizational design also suggests possibilities (e.g. Barrett 1995, Crossan et al. 1995, Jick 1995). The entrepreneurial mind-set critical to success in the business arena lends itself well towards adaptive approaches. Planning could learn from many of these ideas, without incorporating the business mind-set. The strong emphasis being placed on monitoring (e.g. Skibicki et al. 1994, Woodley 1995), is also indicative of the need to learn from our actions, a consideration integral to adaptive approaches.

5.2.2 Choosing Appropriate Planning Approaches

All of these approaches are valid and appropriate some of the time under some circumstances... [It is] perhaps time for us to abandon the search for a single, overarching, discipline-defining paradigm. Instead we should celebrate our diversity, our flexibility and our continued survival. (Brooks 1993, p143)

Choosing planning approaches according to the characteristics of individual situations is a notion that has been advocated for many years and that is gaining greater currency. Etzioni (1967) developed the concept of mixed-scanning. Hudson (1979) constructed a classification scheme for allowing comparison and parallel application of different models. Alexander (1984), in a paper titled "After Rationality, What?" and again in 1996, recommended a contingency approach. Christensen (1985, p 63) suggested that "by matching planning processes to problem characteristics, planning offers a chance to overcome, or a least reduce, uncertainty. In this way planning processes can be understood not as predetermined but, instead, as tools whose use must vary depending on circumstances." The distinction between autopoiesis and sympoiesis also offers a vehicle for choosing suitable matches and for suggesting alterations to planning systems already in place.

As noted a few times above, I do not suggest that the autopoietic-sympoietic distinction should be the only criteria applied to gain understanding. I do, however, believe it provides a useful heuristic. The complexities and indeterminacies of the situations we face require multiple perspectives - plural interpretations provide an advantage.

Figure 5.8 is a conceptual map that integrates a conceptual map of system types with that of the prescriptive planning models just outlined and the scientific approaches outlined in Chapter 3. To be more comprehensive, I have also included a typology of assessment approaches discussed by Nelson and Serafin (1993). Correspondence between the conceptual maps allows their use as a matching tool. Suitability of the approaches can be determined by comparing respective locations on the integrated conceptual map. I suggest that it is essential to use approaches from zones that correspond to system types. Consider an autopoietic system, for example, with its predictable, homeostatic, growth-oriented, reproducible characteristics. As illustrated by the conceptual map, such a system should be matched with rational-comprehensive planning approaches, normal science, and analytical assessments. Autopoietic systems are suited to the reductionist tendencies, use of quantifiable information, reliance on predictable outcomes, and limited level of involvement that are characteristic of rational-comprehensive planning. Systems which exhibit some sympoietic characteristics and/or which exist in the biological domain, however, involve more uncertainty and complexity. They therefore require more flexible approaches. Participatory planning models, professional consultancy, and interpretive assessments, as listed in Zone 2, are more suitable. Even these approaches, however, are inadequate for dealing with the complex, uncertain, and unpredictable sympoietic systems, especially in the social and cultural regions. These systems call for the flexibility and plurality promoted by adaptive planning, post-normal science, and adaptive assessments.

This heuristic is only a rough guide and should not be used on anything but a very general level. My key point is to encourage recognition of sympoietic systems and the subsequent need for more flexible, adaptive approaches when planning for sustainability. The rudimentary matching suggested here may benefit by incorporating techniques of authors noted above, especially of Etzioni (1967).

It is important to note that approaches farther from centre can be used for less complex systems, but the reverse is not the case. As noted, the rational-comprehensive approach - an autopoietic approach - is suitable for autopoietic systems. Adaptive approaches, with their emphases on monitoring and learning, are more suitable for sympoietic systems, yet they are also possible for autopoietic systems. The complex nature of the social and environmental issues currently facing our species, and the imperative need to address them, indicates the need for developing very different planning approaches. It is here that I differ from Christensen (1985), quoted above.*2 Where she advocates reducing uncertainty, I suggest embracing it through the formation of sympoietic planning systems, which I will discuss in the last section.

This notion suggests another advantage that arises from recognizing the correlations illustrated in the integrated conceptual map on a different level. Above I considered the systems portion of the integrated map as identifying the systems of concern to planning. It can also be recognized as reflecting the characteristics of the planning approaches that are represented in each of the three zones. As noted above, rational-comprehensive planning carries autopoietic characteristics, a characterization reflected by the diagram. In contrast, the description of adaptive planning suggests sympoietic characteristics. This match suggests a means for improving planning approaches that may already be in use in any particular situation. In essence, any planning system and its system of concern should have matching characteristics. By conceptualizing and typing both the systems of concern and the planning system, the presence or lack of such a match can be ascertained. Following this, discrepancies between the autopoietic-sympoietic characteristics that exist, and those that are preferred, can be used to determine changes to make to the planning systems.

Recognizing this link between approaches and characteristics, I return to the comment above regarding the distinction between planning and management. In Zone 1, at the rational-comprehensive level, people believe planning and management to be separate. The purpose of planning is to create a plan - the purpose of managing, to implement it. The two are perceived and treated as connected, but different, autopoietic systems. At the adaptive level, planning and management must merge into a more cohesive and seamless process - an integrated approach that carries sympoietic characteristics.

To consider application of these systems concepts to planning systems and their subsequent implications, examples are again useful.

5.2.3 Examples: Point Pelee and Other National Parks

I have stated my concerns regarding sustainability, especially the importance of ecological processes, and the potential for planning to play an important role regarding these concerns. In consideration of these concerns and potentials, then, I apply the systems concepts to planning for parks and protected areas. In particular, I discuss the planning and management system for Canadian National Parks, with a particular focus on changes at Point Pelee National Park and recent developments at Georgian Bay Islands National Park. As with examples in the previous discussions, system conceptualization, and the subsequent discussion of implications will be cursory. The intent is to illustrate application of the concepts rather than to provide a comprehensive view of the planning and management system. I take a historical view, drawing comparisons across different parks and over different periods, noting changes in documentation. In this manner, application of the rational-comprehensive approach - but also a shift away from this approach toward incorporation of more sympoietic characteristics - can be demonstrated.

Figure 5.9 approximates the pattern of organization of a planning system for a national park. As would be expected for a self-producing system, the components have a high degree of interconnectedness. Some of these components are described in park documentation, including zoning descriptions, management plans, and staff roles. The latter refer to defined positions identified in job descriptions. 'Personalities' refer to the individual characteristics of the people holding these positions, including their world views, attitudes, and personal interaction skills. The term 'perception of the site' is used to emphasize that social components, rather than the actual biophysical structures are the focus of consideration. It is the human perceptions and interpretations of the biophysical structures that are relevant here. Knowledge refers to specific information contained in reports and carried by individuals. Available knowledge, however, will be limited to that which is accessible and is of a type legitimized by the system. Growing recognition of different types of knowledge such as traditional ecological knowledge (e.g. Inglis 1993) points to the importance of this component. Concerns regarding issues of knowledge availability will be noted below.

The structure of such a system will be the manifestation of this pattern of organization in some particular park planning system, for example, the actual components of the planning system at Point Pelee National Park. There are three considerations of particular importance that I draw attention to. Many components are non-trivial systems. Individual personalities, for example, will be influenced by the complex human characteristics discussed in Chapter 3. Second, the system as a whole will be structurally coupled to its environment/context, which will include other institutions (government bureaucracies, policy statements), public expectations, and the natural environment. Third, due to the complexities of social systems, particular components will be structurally coupled to particular contexts. These will each be sub-systems of the social system. For example, knowledge will likely be based upon, and extracted from a particular knowledge system - a particular type of knowledge represented by a particular world view or paradigm. Some implications of such coupling can be drawn from the discussion of science (Section 3.3) and of learning (Section 4.3.1) above.

Typing Planning Systems

At a simplistic level, identifying the self-organizing, sympoietic, or autopoietic nature of the planning system can be considered by answering the question: is the system more like a river, an ecosystem, or an organism? (See Section 2.4.1) In the case of Point Pelee, the planning system appears to be more like an organism - an autopoietic system with relatively self-defined boundaries that produces its own components. This can be recognized by considering similarities to the descriptions of autopoietic organisms given in Section 2.1. The planning system receives a 'package' of organizational information from national policies that enables it to produce components, just as a parental autopoietic organism provides its offspring with the organizational information required for the offspring to create itself. Based on the information received, the planning system generates goals, objectives, staff roles, management plans, management zones, public input, and other components that are essential for its own development and continuation. In this sense, it is self-producing. In addition, it will generate boundaries by determining the scope of management influences.

As autopoietic systems, the planning systems for individual parks have transmitted self-organizing potential, which is reflected in the hierarchical and historical nature of the National Parks system described in the preceding chapter. National guidelines and policies have been developed over time with consequent development and adjustment of particular institutional structures. These structures, in turn, create further guidelines (McNamee 1993) illustrating the self-producing nature of the park management system. This transmitted self-organizing potential reflects incremental planning descriptions.

After the early years, in which planning for each park was done on a rather ad hoc basis, the first official national parks policy was adopted by Federal Cabinet in 1964 (Lothian 1987). With development of national policies, park planning systems were gradually created that reflect a transmission of national planning processes and plans to individual park levels. For example there is a series of National Park Provisional Plans created in the 1970s, which reflect the national template despite differences in parks. Although local situations varied, the processes applied and final documents are almost surprising in their similarity - right down to the cover design - excepting colour! As examples, comparison of the plans for Fundy, Forillon, and Prince Albert in 1970 with plans for the same parks developed in the 1980s illustrate the changing nature of planning. (See National and Historic Parks Branch 1970a, 1970b, 1970c, Canadian Parks Service 1991, Parks Canada 1995a, 1995b) By the 1980s, greater differences are apparent in plan documents. Each park planning system is gradually gaining its own identity. Although the template is received from the national level, local differences in the components noted above (especially personalities, interpretations, and available knowledge) will produce different systems in the different parks.

With the recent turn towards more collaborative approaches, including increased public involvement (Parks Canada 1994), co-management (e.g. Gwaii Hanaas, Parks Canada 1998), and Ecosystem Conservation Plans (e.g. Georgian Bay Islands National Park, Nelson and Skibicki 1997), distinctions between parks are becoming ever greater. These differences reflect a shift toward more sympoietic characteristics, such as creative self-organizing tendencies, involved in the generation of the newer management plans.

To illustrate the changes I use examples from Point Pelee National Park over the past twenty years. To illustrate newer approaches I draw from two examples: development of Georgian Bay Islands National Park Ecosystem Conservation Plan (GBINP ECP) using Nelson's civics model, and aspects of the collaborative land use planning process applied by the Commission On Environment and Resources (CORE) in British Columbia.

Planning Models: Rational-Comprehensive to Adaptive

A key indicator of the type of poietic planning system in place in any particular situation is the type of planning model applied.

Point Pelee National Park: For Point Pelee National Park, the stages outlined in the "Resource Management Analysis: Proposals and Summary" of 1978 (Parks Canada 1978) illustrate a classic application of the rational-comprehensive planning model for determining management objectives with respect to the natural features of the park. Under the section titled, "Methods" is the following list:

The project went through six major stages:

Definition of the project objectives and influencing factors

The establishment of general approach and strategies

Data review, synthesis and description

Data analysis and evaluation

Establishment of resource management concerns and proposals

Resource management objectives. (ibid., p 14)

There are also indications, however, that wider influences on the process are acknowledged. The section immediately preceding the description of methods lists "Influencing Factors." The list includes the National Parks Act, Parks Canada policy, Parks Canada zoning system, national role of the Park, bio-geographic and socio-economic factors, and operational constraints. Each is briefly described, noting a mixture of constraints on the process and on the resulting objectives. In a sense, acknowledgment of these factors suggests implicit recognition of some self-organizing factors involved in the planning process. As I have indicated above, however, available knowledge, perception of the site, and staff personalities are also critical factors that need to be recognized. Although these are not articulated, a sense of the former two factors can be gathered from reading the document. There is a tone that definitely establishes reliance on scientific knowledge and a problem-solution focus that are typical of rational-comprehensive planning. Examples include the following excerpts.

Based on the review of existing data, present and anticipated future resource management problems were extracted and described. These aspects are then discussed and evaluated in terms of the course of action that Parks Canada should follow to solve or minimize these problems. (Parks Canada 1978, p 17)

Large scale passive management. All special management needs (such as disturbance, endangeredness and wildlife requirements) have been handled and integrated at the lower scales. It remains only to see that the majority of the vegetation cover regenerates itself or proceeds along normal predictable successional pathways. (ibid. p 21)

In contrast, a more recent Park Conservation Plan for Point Pelee National Park, has a somewhat different tone. An appendix describes the methods followed for plan development. Accepted knowledge still seems to be based on predictive science:

The goal of natural resources management must be predictable since predictability leads to more of the best kind of National Parks resources management - confidence that no management (just monitoring of the decision not to manage) at all is needed. So called "problems" should be eliminated before they arise. (Parks Canada 1991, p 151)

However, recognition is given to general concerns from a more holistic perspective:

The use of the term "problem" is one of convenience only. It is not used in the traditional sense but, is used as a substitute for the phrase Park "problem, issue or concern" found in the N.R.M.P guidelines for preparing Park Conservation Plans. In its narrowest definition the N.R.M.P. consider problems, issues and concerns to be those for which Resource Management Plans are needed and in turn Resource Management Plans are only needed when active or manipulative resource management is warranted. This leads to focusing in on those parts of the problem which involve manipulative action. The result is distortion of the true problem and neglect of the ecosystem as a whole as well as the cumulative incremental impacts of Parks operations. (ibid. p 150-51)

In addition, some of the broader, self-organizing factors are acknowledged in this document also - although on a more specific basis. In order to guide management, each of the identified concerns is ranked according to a set of criteria to determine their respective priorities. The ranking is recognized as "highly subjective," and "not linear," to the end that "no rigorous absolute scale was used" (ibid. p 155). Table 5.1 lists the ranking criteria, which represent many of the self-organizing factors that are of a concern to the park and to the planning process, and which should be incorporated into planning and management.

The differences between the 1978 and 1991 documents - both related to establishing resource management concerns - illustrate a shift from the more autopoietic, rational-comprehensive approach toward a system that incorporates more sympoietic characteristics.

• CORE model: As noted above, one of the attractions of the rational-comprehensive model, is the clear, step-wise process that provides the sense of both rationality and comprehensiveness - two preferred characteristics. Also noted, however, were the difficulties in achieving such steps. To illustrate the possibilities for encouraging sympoietic characteristics, I draw on examples from the CORE process and Nelson's civics model. The following stages are recommended by the CORE model (Brown 1996).

1. Getting ready
2. Designing the process
3. Developing plan goals
4. Building a toolkit
5. Developing land use plan scenarios
6. Agreeing to a comprehensive strategic land use plan
7. Plan implementation, monitoring, review and amendment (see ibid., p 29)

• Civics model: This model is also quite distinct from the rational-comprehensive model since it offers a set of "basic processes" to be covered in an interactive, adaptive manner, rather than a series of steps to be followed in a particular order. Box 5.2 lists the basic processes, illustrating their loose connections. In this case, 'planning' is only one element of the whole process, fitting with the recognition noted above that at the adaptive level, planning and management merge. In this model, the processes "would all be at work in different sectors and situations at the same time, interacting on a reinforcing and continuous basis. In definition and in practice, the processes overlap and interact with one another as they are practiced by different governments, citizen groups, businesses and individuals alike" (Nelson and Serafin 1996, p 15). This description also reflects sympoietic, rather than autopoietic characteristics.

Degree of Organizational Closure

An aspect to consider regarding the poietic nature of planning systems is their degree of organizational closure. As noted above, I consider the pattern of organization of a planning system to include the relations among actors, knowledge, objectives, and other components. To be organizationally ajar, such systems must be able to alter these components and/or their relations - for example, through addition of new actors, or new knowledge. In the following discussion I will focus on two key approaches relevant to altering the degree of organizational closure. The first approach is to make the planning system more sympoietic, by increasing public involvement, for example. The second, is to develop a wider sympoietic system of which the planning system is a component, for example, by encouraging collaborative approaches. The former is likely more useful for mitigating internal threats such as high levels of use, or for establishing management objectives. The collaborative approach will likely be necessary for threats arising external to the park, such as from neighbouring land uses. I noted both possibilities when considering the potential for Point Pelee National Park's planning system to influence the social systems impacting the park, suggesting the latter as necessary (Section 4.5.1). To be most effective, I believe the two approaches should go hand-in-hand, however, recognizing the differences can be useful. The example of Georgian Bay Islands National Park discussed below illustrates their integration.

One of the most obvious methods for achieving organizationally ajar planning systems is through public involvement. This factor is especially important to planning systems responsible for public resources. Figure 5.10 illustrates varying degrees of involvement that are possible, interpreting Arnstein's (1969) ladder of participation according to autopoietic-sympoietic characteristics. Drawing from examples, I illustrate the differences.

• Point Pelee National Park: The process for development of the Point Pelee Master Plan in the late 1970s was described as a consultation program, and had six objectives:
• to provide the public with information on the aims of Parks Canada in the development of a management plan
• to provide the public with information on the problems caused by extensive visitor use of the small environmentally sensitive areas
• to promote awareness of the problems faced in the management of the Park
• to provide the public with opportunities to express their views on the development and management of the Park
• to allow Parks Canada to modify the development proposals during the preparation of the Management Plan in order to meet the public's needs and expectations
• to provide the public with a response to their comments indicating the Parks Canada position on the issue (Parks Canada 1978, Appendix III)

These objectives carry connotations that may, in actuality, place it even further down on Arnstein's ladder.

The more recent National Parks Management Planning Process Manual (1987) uses language that carries similar connotations.

The purpose of public participation is to allow the public to express their views. Citizens' ideas and comments can provide valuable input to the planning process. Public input is an integral part of this process and results in better decisions. Through such participation, there can be orientation of the public to Parks Canada's policy and objectives... (Parks Canada 1987, p 37).

Opening system boundaries by increasing public involvement in the decision-making process is a suggestion that has received approval (Wells and Brandon 1993). Broader, more inclusive approaches toward incorporating the public as a means of reducing organizational closure can be illustrated by examples.

• Civics Model - Georgian Bay Islands National Park: One of the early stages involved in application of the civics model is development of a communications strategy aimed at identifying and gathering information from stakeholders, including their issues and values. Such a strategy was applied in the creation of the Ecosystem Conservation Plan for Georgian Bay Islands National Park (Dakin 1996). Dakin (p 7) noted "the ongoing nature of the stakeholder identification process, as new groups are continually emerging, often in response to newly identified issues." The process acknowledged the sympoietic nature of the system in place by drawing information from diverse sources. Guided by both stakeholders and circumstances, a snowballing technique was used to identify relevant groups and individuals. In addition, the communication strategy capitalized on associations already in place, such as the Georgian Bay Association. These factors illustrate that the strategy, a key part of the planning process, was organizationally ajar. This interpretation is supported by a description of the Communication Strategy found in the Ecosystem Conservation Plan:

• CORE process: The rules of procedure for visitors to the Anaheim Round Table, an initiative of the CORE process, provides another example of a planning system that is organizationally ajar.

• Gwaii Hanaas National Park: Another innovative approach to public involvement regarding planning in National Parks is illustrated by Gwaii Hanaas National Park. In this case, the Internet is being used to obtain feedback on a backcountry proposal (Parks Canada 1998). Although the document and process are otherwise relatively standard (except that greater recognition is given to aboriginal issues due to co-management practices), using the Internet allows inclusion of many more people. The latter point is especially relevant for parks with a national interest. In addition, use of the Internet allows potential participants to decide on their participation in the process, although only people with access to the Internet can participate in this manner.

Although these examples reflect sympoietic characteristics, they do not necessarily equate to sympoietic systems. Being organizationally ajar means that a system - on an organizational level - is neither open, nor completely closed. It must be able to accept, but limit, alterations to its pattern of organization.

For comparison, consider what occurs in an ecosystem. Since these systems do not have boundaries, new species can enter at any time. Whether or not such species will successfully take hold in any particular ecosystem will depend on the respective system structures. If the ecosystem is dry and the new species does not have a structure that can cope with such conditions, it will not be incorporated. In consequence, the information it carries will not become part of, or available to, the ecosystem. In a similar manner, interested individuals and groups will determine their incorporation and participation as components of a sympoietic planning and management system according to their respective structures. For example, national park management systems have tended to rely on 'scientific' knowledge for decision making. Such knowledge tends to be quantitative and specific. Incorporating traditional ecological knowledge, which is often anecdotal and holistic (Inglis 1993), may be difficult even if the planning system articulates a desire for its incorporation. 'Site perceptions' and staff 'personalities' which are suited to, and comfortable with, facts and figures will not necessarily be able to accommodate alternate types of information. Management plans structured to list achievable, quantifiable objectives may be incapable of incorporating information that cannot be listed in such a format.

Note that this is a structural problem - a system with the same pattern of organization, but with different site perceptions and personalities may be able to incorporate the new information. The non-trivial nature of the various components is critical, since they may provide new structural opportunities (e.g. staff member who is comfortable with alternate knowledge despite science background). Changes in pattern of organization advocating public involvement will not necessarily allow such participation. Changes in the structure of the planning systems are also essential. A system which retains an exclusively 'scientific' orientation to knowledge will not be capable of full public participation since certain types of knowledge will not be accepted. Accepting a different type of knowledge, then, may require altering structural components of the planning system. The self-producing nature of such a system illustrates that organizational and structural change must go hand-in-hand. Recognition of the role played by paradigms discussed in Chapter 3 illustrates and emphasizes some of the difficulties involved in changing systems on this level. The notions discussed regarding post-normal science are also relevant.

Interpreting Arnstein's ladder along an autopoietic-sympoietic continuum (Figure 5.10), leads to recognition of two consequent and parallel continua that have heuristic quality and can aid in understanding these points (Figure 5.11). First, there is a gradation from organizationally closed to organizationally ajar systems. Second, there is a shift regarding the role of structural coupling from primarily external to primarily internal. To allow full public involvement a system's organization must be ajar and the systems structure must be conducive to coupling with new components. For example, use of the Internet by Gwaii Hanaas may increase input, but there is no certainty that responses will be incorporated just because they are submitted. The procedural rules of the Anaheim Round Table, however, advocate development of a structure that will allow coupling of diverse interests.

At the middling positions on Arnstein's ladder - degrees of tokenism - a planning system may claim to be organizationally ajar by allowing public participation, however, on a structural level, it may not be able to accept new components that are offered. It is not sufficient for a system to be organizationally open - the poietic process is cyclic, involving both organization and structure.

These concerns are also relevant for development of collaborative approaches discussed next.

The second approach for altering the degree of organizational closure in a planning situation is to develop a wider sympoietic system of which the planning system is a component.

• Point Pelee National Park: In Chapter 3, the impacts of shoreline management on the geophysical system relevant to Point Pelee National Park were discussed. The boundaries of the park include a very small spatial portion of the geophysical system forming the marsh and peninsula. Since the jurisdiction of the National Park planning system is restricted to this area, it cannot, through an autopoietic planning approach, influence many factors that are critical to the ecological integrity of the park. Aside from the spatial issue, several social elements impact the geophysical system, including progress orientation and environmental ethic. The potential, however, for significantly altering such basic social influences through park management is limited. Due to these factors, the advantage of generating a wider sympoietic system becomes apparent. Such an approach is advocated by encouraging development of a regional shoreline management program as suggested in 1978 (Parks Canada 1978), and by other, especially educational, programs (see also Parks Canada 1991). Current working relations with regional partners on issues such as shoreline guidelines, restoration, education, and other aspects indicate that the management system is already moving in this direction (B. Stephenson 1997, pers. comm.).

• Georgian Bay Islands National Park: Spatially, Georgian Bay Islands National Park faces similar concerns as Point Pelee National Park. The park covers a very small land area and is influenced by social impacts both internal and external to the park. As stated in the park plan:

In recognition of the need to incorporate concern for the greater park ecosystem an Ecosystem Conservation Plan was developed by Nelson and Skibicki (1997). As noted in the above discussion, the civics model, including development of a communication strategy was used to create the plan. The resulting document provides a good example of a mixed approach. The authors recognize the need for both autopoietic and sympoietic characteristics in order to cope with the complexities involved in the situation.

Spatially, the plan recommends two interconnected areas: a Core Area, and an Area of Cooperation and Communication. The latter includes a Near-Core Area which represents a transition zone. Organizationally, the plan recommends that the National Park planning system already in place should be augmented by a Consultative Committee and a Greater Park Ecosystem Forum. As illustrated in Table 5.2, the latter two are responsible for the areas beyond the park boundaries.

The Core is described as having the "most significant assemblage of natural and cultural values in the GBINP area" (ibid. p 33). The authors recommend an autopoietic approach: "Since it is the Core Area, Parks Canada should exercise its full mandate and a high level of control over land and resource use activities" (ibid. p 34).

In practice, however, such controlled application of policies and procedure is typically limited - a concern discussed by Nelson and Skibicki. They list several other government and social aspects that have some level of jurisdiction or influence over resources of immediate concern to the park such as fish, waterfowl, snowmobile trails, and water quality. Recognizing the limited potential for park planning and management to have such a high degree of control emphasizes an important point. While I recommend generating a wider sympoietic system for planning with respect to the park, it is crucial to recognize that the park planning system is already part of a wider social sympoietic system. The problem, however, is that it is not a sympoietic system with the conservation and maintenance of the park's ecological integrity as a central concern. In order to meet the park's mandate, then, the park planning system must encourage the generation of a sympoietic social system that does have ecological integrity as a key focus.

By linking spatial and organizational aspects of the plan, the potential for maintaining control in the Core, and developing collaborative approaches in the Area of Communication and Cooperation is illustrated. The following quotes demonstrate the essence of the arrangements.

The principle means of collaboration should be through a Consultative Committee which would work in association with a Greater Park Ecosystem Forum. The Consultative Committee would focus as a matter of ecological priority on the Near-core Area. Such a focus is of fundamental importance to the ecological integrity of the GBINP, the Core Area and the environmental health and sustainable long term use of the Greater Park Ecosystem as a whole. (Nelson and Skibicki 1997, p 37)

The Consultative Committee should be convened by Parks Canada and should play a communicating and cooperating role in linking programs and resources in the common interest. (ibid. p 37)

Greater Park Ecosystem Forum... would meet on a regular basis - at least annually - to exchange information and learn from each other's experiences. The Forum would be open to all concerned agencies, groups and individuals. The Forum would provide the broad framework for discussion of land and water uses, their effects, and planning and management responses in the Greater Park Ecosystem or Area of Cooperation and Communication. (ibid. p 38)

While the authors of this plan do not cover details about the processes to be used in the creation or functioning of these organizational arrangements, they do indicate that the park planning system must play an active role. They suggest that "an officer of National Park staff should be assigned on at least a half time basis to work on the planning and implementation of the ECP and specifically on the development of the Consultative Committee and the Forum" (ibid. p 52). Such a commitment may alleviate the structural problems noted above. If an individual with appropriate skills held such a position, they could encourage incorporation of knowledge and interests typically excluded from consideration. In effect, this is a means of altering the planning system's structure to enable structural coupling of new and different components. As noted above, such potential is necessary for the system to be collaborative, rather than just offering token participation.

Poietic System Considerations in National Park Planning Context

Interpreting these examples as autopoietic and sympoietic systems illustrates some of the advantages and disadvantages of the two system types. The most appropriate planning system in any particular situation will depend on the circumstances involved. For national parks, the implications of shifting from autopoietic to sympoietic systems and the implications of remaining autopoietic should be given careful consideration. For example, what advantages and disadvantages does acting as an autopoietic system confer on park planning and management? What advantages and disadvantages would acting as a sympoietic system confer? What are the consequences of making a change, including the process involved in changing? As national parks, both Point Pelee and Georgian Bay Islands have to follow the national mandate. Is it possible (or desirable) for an individual park planning and management system to encourage a sympoietic approach? What other systems should be included in movement toward a sympoietic system? What are the implications for national policy and for other parks? Although public involvement has been discussed to illustrate the potential advantages of developing the sympoietic nature of planning and management, other components and systems should also be considered. This could include other government departments, institutions, organizations, and economic sectors.

The following points outline some considerations relevant to understanding the advantages and disadvantages of the two system types.

• Autopoietic vs. Sympoietic Characteristics: Since the systems of concern are not bounded, it is essential to incorporate planning and management issues beyond park boundaries. This suggests a need for sympoietic planning. In many cases, however, especially for such small parks as Point Pelee and Georgian Bay Islands, a firm degree of control within the park may be of merit, in order to influence internal threats. The potential for centralized control - a characteristic of autopoietic but not sympoietic systems - should not be relinquished lightly. Yet, the disadvantages of autopoietic systems must be recognized in order to ameliorate their effects. The problems identified when discussing groupthink (Section 3.3) exemplify possible difficulties. Our capacity for self-awareness, which I stress below, is important in this respect. It is critical to be aware of the potential for misinterpretation with respect to the systems of concern, but also with respect to the planning system itself.

• Transmitted Self-Organization: I have described the self-organization within national park planning as transmitted, due to similarities with the transfer of information between biological autopoietic systems. Our complex human capacities, however, mean that there are also key differences. In particular, transmission of organizational information is not necessarily a one-time occurrence, but may be a continuing process. In National Parks the latter is the case. Although each park planning system has some degree of autonomy, changes in national policy will influence them. This characteristic is more sympoietic in nature, providing a potentially greater capacity for adaptation, even in autopoietic planning systems. The difficulty, however, is that the national system itself has autopoietic characteristics, and subsequently holds particular advantages and disadvantages. Noting the political nature of park planning in the previous chapter and the move toward emphasis on business plans - an influence being advocated from a national level and coupled to the political context emphasizing finances - illustrates potential disadvantages. The potential for individual parks to be overcome by regional or local concerns rather than representing national concerns, also illustrates disadvantages - but in this case of not having strong transmitted self-organization.
• Recursion: The self-producing qualities of both systems types indicate that multiple points of entry are not only possible, but necessary. For autopoietic systems, it is essential to recognize that changes to system structures leave the systems vulnerable unless the environment/context also changes. Sympoietic systems change in a more evolutionary manner. In both systems, history is critical in determining future potentials.
• Structural Coupling: The importance and potential influence of structural coupling must also be recognized. Both types of system rely on structural coupling, however, autopoietic systems require more predictable inputs and have less ability to adapt to changing structural contexts. In discussing the role of park planning within the wider social systems, I noted potential difficulties regarding structural coupling to the economic context, with the current shift toward 'business plans' as a concern. Planning systems will also be coupled to particular conceptual contexts, especially knowledge systems such as normal science. Dependence on these contexts will be greater in autopoietic systems, which will have various advantages and disadvantages.
• Available Knowledge: One of the values of sympoietic systems is that they contain a wide variety of information distributed among their components. Incorporating the public has the potential to increase the information content of the system. This does not simply mean distributing Parks Canada information among the public. It means that the wealth of knowledge carried by interested individuals becomes available to the planning and management system. As explained by the distinction between the role of information in the two system types (Section 2.2), this factor could increase the adaptive potential of the planning and management system. Understanding the concept of structural coupling indicates that making such a change may not be straightforward. Despite even the best intentions, a system which advocates public involvement but which does not present a structure conducive to such involvement will not enable its actual occurrence.
• Non-Trivial Components: Another factor to be aware of is that many components of these planning systems - in particular the individuals involved - are non-trivial complex systems. As such they will be responding and behaving according to their own needs and interests in potentially unpredictable manners. Issues that could be of concern include the ability of different individuals to accept different types of knowledge, to prioritize personal advancement within the organization, and to use different interpretive lenses for understanding the systems of concern.
• Self-Awareness: I believe that a key advantage provided by our complex human characteristics is our potential for self-awareness. We have the ability to perceive our actions and to consider the possible consequences. This is not the same as predicting what will occur, but simply illustrates an ability to reflect on possibilities. In regards to either poietic approach to planning and any of their respective advantages and disadvantages, ensuring such self-awareness is critical. Argyris and Schon's (e.g. 1996) advocacy for being "reflexive practitioners" argues this position. It is essential, however, to encourage such self-awareness at a level of the planning system itself, not just in the individuals acting within the system. The potential for developing such system-self-awareness is a consideration I must leave to future study. In addition, the advantages and disadvantages for enabling this potential in the two system types should be considered.

*1 Although I am most comfortable with the terms used in the manner noted, it is essential to understand that the problem arises from drawing correlations between autopoietic and sympoietic characteristics, not from the terms themselves. 'Integrity,' if it is used to represent homeostatic balance, does not alleviate the problem.

*2 Christensen also develops a biaxial typology with similar axes, creating a four box matrix. Although she advocates a different approach, her discussion provides some useful insights.