Navigating amid uncertainty in spatial planning

Abstract

In view of the need to adapt to uncertain climate change through spatial interventions, this article explores how spatial planners might navigate amid uncertainty. To draw out insights for planning, we examine planning frameworks which explicitly recognise uncertainty and uncertainty descriptions from studies in environmental risk and climate uncertainty. We build our case by addressing the implications of different characteristics of uncertainty and describe how planners can handle uncertainty based on the nature, level and location of uncertainty. We argue that a plural–unequivocal characterisation of uncertainty helps planners in their search for adequate and warranted interventions amid uncertainty.

Keywords

ambiguity climate change long-term consequences moral responsibility spatial planning uncertainty

Introduction

Spatial planners contribute to the organisation of the environment by preparing planning decisions and taking deliberate actions to develop places (Balducci et al., 2011; Christensen, 1985). When taking decisions or preparing possible interventions, planners will always be confronted with uncertainty (Albrechts, 2004; Balducci et al., 2011; Christensen, 1985; Lau, 2015; Rauws et al., 2014). Uncertainty is a particularly important consideration when assessing the potential long-term consequences of interventions in the environment (Connell, 2009; Salet et al., 2013; Van der Vlist et al., 2015; Walsh et al., 2015). Planners are also required to assess the impacts of climate change in order to design spatial interventions for climate adaptation, even though climate change can alter local environments in unpredictable ways. Moreover, these interventions, such as the construction and adaptation of infrastructure, can influence the location and layout of development for many decades, with possible unforeseen consequences for other spatial interventions or society (Graham and Marvin, 2001). Altogether, this demands a coherent understanding of uncertainty to inform adequate planning interventions.

Contemporary planning theory considers uncertainty as part of complexity thinking (e.g. Batty, 2013; De Roo and Silva, 2010; Innes and Booher, 2010). However, planning theorists have described uncertainty in many different ways (e.g. Abbott, 2009, 2012; Balducci et al., 2011; Bertolini, 2010; Gunder, 2008; Hillier, 2010, 2013; Lau, 2015; Rauws et al., 2014; Salet et al., 2013), which allows for a variety of interpretations. These differences between interpretations give rise to three issues. First, where planners only partially understand uncertainty their interventions may be redundant or deficient. Second, if planners act with a poor understanding of uncertainty, decisions and interventions may turn out to be maladaptive. Third, the normative implications of structuring development for decades under conditions of uncertainty include a moral responsibility: decisions and interventions in response to uncertainty may be unjust, or lead to injustice. These issues can be illustrated by the example of building a dike for protection against flooding in the face of uncertain climate change. The dike may be built too high or low, in the wrong place or way, or with unjust consequences for flood safety.

We argue that to deal adequately with specific uncertainties, spatial planners should understand the differences between uncertainties. The aim of this article is to explore perspectives for handling uncertainty in spatial planning. To gain insight into the adequacy of planning interventions when confronted with uncertainty, we set out to answer the question, ‘What differentiation of uncertainty offers insight into adequate and warranted planning interventions?’ We structure our article into three parts: first, how uncertainties can be differentiated; second, how these relate to planning; and third, what insights can be derived from such a differentiation for handling uncertainty in spatial planning, for both adequate and warranted interventions. We aim to contribute to the subject of uncertainty in planning and to the development of a theory on the adequacy of interventions in relation to uncertainty. Moreover, we investigate some of the ethical considerations of handling uncertainty.

In the next section, we examine four theoretical frameworks in which uncertainty has an important role. These frameworks were developed by Abbott (2005, 2009), Christensen (1985), Friend and Hickling (2005), and Islam and Susskind (2013). In the following section, we describe three characteristics of uncertainty based on recent conceptualisations of uncertainty in studies pertaining to climate change uncertainty and environmental risk (building on e.g. Brugnach et al., 2008; Kwakkel et al., 2010; Skinner et al., 2013; Van den Hoek et al., 2014; Walker et al., 2003). As these studies propose characterisations of uncertainty for the purpose of better handling long-term but as yet unknown environmental changes, they may offer insights into how to handle uncertainty in spatial planning. Then, in the main body of the article, we connect planning with uncertainty conceptualisations, first by articulating what we mean by ‘planning’ and then by building on the three characteristics of uncertainty (nature, level and location). We explore the characteristics of uncertainty to gain insights into handling perspectives for planning. Finally, we synthesise these insights to determine what theoretical footholds the three characteristics of uncertainty offer to help planners navigate amid uncertainty.

Uncertainty frameworks in planning

Uncertainty frameworks in planning theory have evolved over several decades. A widely used framework is the one developed by Christensen (1985), which was also used by Balducci et al. (2011) and referred to by many others (e.g. Abbott, 2005; Alfasi and Portugali, 2004; Gunder, 2008). A similar framework is the one used by Stacey (2007), and Islam and Susskind (2013). Another influential framework was proposed to assess uncertainty in the strategic choice approach (Friend, 1993; Friend and Hickling, 2005; Friend and Jessop, 1969). This framework was further developed by Abbott (2005, 2009, 2012), who added some components to the characterisation of uncertainty in the strategic choice approach. We base our discussion of these four frameworks on a review of the literature on uncertainty and planning and on citations and alterations of the discussed works.

Christensen (1985) links the variables ‘knowledge’ and ‘agreement’ to two types of uncertainty, uncertainty about means and uncertainty about ends, in a matrix to obtain four planning situations. Planning problems can be mapped according to these four situations. There is an implied value judgement that the best situation is achieved when there is agreement and sufficient knowledge to make a decision (Bertolini, 2010). Regarding uncertainty about means and uncertainty about ends in planning, Christensen (1985) states, ‘By matching planning processes to problem characteristics, planning offers a chance to overcome, or at least reduce, uncertainty’ (p. 63). Uncertainty about means can be overcome by gaining more knowledge or initiating a learning process; agreement is necessary to overcome uncertainty about ends.

In the context of the strategic choice approach, Friend and Jessop (1969), Friend (1993), and Friend and Hickling (2005) describe three types, or ‘areas’, of uncertainty and link these to three ‘structuring principles’ for planning. Uncertainties relate to a planning problem and the planning process must be informed by ex-ante assessment of prevalent areas of uncertainty. Friend (1993) reviews the three areas of uncertainty and discusses how these areas influence planning. The first area is ‘uncertainty in the working environment’, which calls for deeper investigation. The second area, ‘uncertainty about related choices’, demands wider collaboration. The third area is ‘uncertainty about guiding values’, which indicates that objectives are not yet clear enough. Based on these areas, the planner must learn ‘to manage uncertainty in a strategic way. This means considering carefully which areas of uncertainty are most significant in any particular planning context and what possible forms of actions might be initiated in response’ (Friend, 1993: 1).

Both frameworks are used and adapted by others, most elaborately by Abbott (2005, 2009, 2012). Abbott (2005) synthesised the frameworks of Christensen (1985) and Friend (1993) into one framework with environmental uncertainty and social or planning-process uncertainty as two overlapping dimensions and five distinct ‘natures’ of uncertainty (chance, external uncertainty, causal uncertainty, organisational uncertainty and value uncertainty) as subdivisions (Figure 1). Although Abbott acknowledged that distinct uncertainties have different implications for planning, he did not further address these, except for stating that ‘Different planning theories can thus provide guidance on how to understand and manage different dimensions of uncertainty’ (Abbott, 2005: 248).

Figure 1.

Abbott’s (2005) framework: two overlapping dimensions with five ‘natures’ of uncertainty.

The framework used by Islam and Susskind (2013), adapted from Stacey (2007), bears strong similarities to the framework of Christensen (1985), but differs in three main ways. First, the variables have been changed from binary variables (either knowledge or no knowledge; either agreement or not) to ratio-scaled, gradual variables. Second, the framework has been altered from a means–end dichotomy to an agreement–uncertainty dichotomy, thereby overcoming the difficult relationship between means and ends (e.g. Simon, 1961). Third, uncertainty is connected to the complexity of a planning situation. Adapting the framework in such a way allows the complexity of the planning problem to be characterised according to the amount of uncertainty and disagreement (cf. means and ends in the matrix of Christensen (1985) and facts and values in Simon (1961)). More disagreement and more uncertainty indicate greater complexity. As Islam and Susskind (2013) show, a planning problem or situation can be positioned on a scale from simple situations with little disagreement and/or uncertainty via complicated situations to complex planning situations with much disagreement, large uncertainty or both (see Figure 2).

Figure 2.

Islam and Susskind’s (2013) framework: consensus and uncertainty add up to complex and disorderly planning situations (Stacey, 2007).

In the four frameworks, the uncertainty debate revolves around how to deal with uncertainty (Christensen, 1985; Islam and Susskind, 2013) or uncertainties (Abbott, 2005; Friend, 1993) in planning. One insight from the conceptualisation of uncertainties in the frameworks is that the coupling between what uncertainty is and what uncertainty entails for planning is coloured by the planning paradigm within which uncertainty is defined. For example, Christensen (1985) uses the differences between uncertainty about means and ends to classify prototypical planning processes, such as rational planning (when everything is clear) and a bargaining process (when only ends are uncertain). Each of the frameworks makes assumptions about what planning is, how this affects their ontological position and how uncertainty is used to assess the adequacy and justification of planning interventions. We argue that a detailed understanding of uncertainty could surpass differences between rational–comprehensive, incremental and communicative, relational planning (Christensen, 1985), but reject the ontology-free ideal in which planners can choose the uncertainty and planning definitions as they like. We now turn to how we can better understand uncertainty in planning by examining how it is characterised in risk- and climate studies.

Characteristics of uncertainty in risk- and climate studies

In risk and climate studies, uncertainty is categorised in many different ways (e.g. Kwakkel et al., 2010; Skinner et al., 2013; Van Asselt, 2000; Van Asselt and Rotmans, 1997; Van den Hoek et al., 2014; Walker et al., 2003). These studies show a rapid evolution in the conceptualisation of uncertainty over the past two decades. Skinner et al. (2013) reviewed characterisations of uncertainty in environmental risk studies and concluded that even in one scientific domain, terminology is inconsistent and sometimes contradictory. Of the many characterisations identified, nature, level and location are used by a range of reviews and studies to understand uncertainty (Brugnach et al., 2008; Kwakkel et al., 2010; Skinner et al., 2013; Van der Keur et al., 2008, 2014; Walker et al., 2003). We examine these characteristics to identify what uncertainty might entail for planning.

The nature of uncertainty

The nature of uncertainty relates to the question of why a phenomenon is uncertain. Kwakkel et al. (2010) distinguish three natures of uncertainty: ontic, epistemic and ambiguous uncertainty. Ontic uncertainty is also referred to as variability, stochastic uncertainty, aleatory uncertainty, random uncertainty, fundamental uncertainty or chance (Rauws, 2015; Skinner et al., 2013; Walker et al., 2003). Ontic uncertainty arises from variability in a phenomenon and is inherently unpredictable and irreducible, even if systems could be better understood over time. The origins of ontic uncertainty are the unpredictable and chaotic dynamics of physical, economic, political and cultural phenomena and human behaviour (Walker et al., 2003). An example is the future discharge of a river, which can never be predicted a month ahead due to the inherent chaotic, non-linear behaviour of the water system (Milly et al., 2008).

Epistemic uncertainty is the incompleteness or imperfection of knowledge, or inexactness (error) (Hofer, 1996; Walker et al., 2003). The origin of epistemic uncertainty is a lack of data, poor quality of data, or insufficient techniques to measure parameters that may be relevant for the decision or policy at hand. Several statistical measures exist to indicate the quality of quantitative data. For example, Funtowicz and Ravetz (1990) introduced the notion of pedigree in qualitative data to ‘systematically assess the imperfection in the knowledge base, thereby providing an indication of the degree to which uncertainty may be reducible’ (cited in Walker et al., 2003: 13). The main difference between ontic and epistemic uncertainty is that ontic uncertainty cannot be reduced, while epistemic uncertainty can, namely by increasing the amount or quality of knowledge or reducing error. An example is the failure of a river dike. The mechanisms of failure are well understood and, thus, the susceptibility of the dike to failure at any location could be determined. It is, however, too expensive to measure the strength of the dike at each and every location along its length and during each possible combination of hydraulic loading and antecedent conditions, so water managers often accept a degree of epistemic uncertainty.

Ambiguity is a third and independent nature of uncertainty, as first proposed by Brugnach et al. (2008). Kwakkel et al. (2010) define ambiguity as ‘uncertainty arising from the simultaneous presence of multiple frames of reference about a certain phenomenon’ (p. 310). Ambiguity is seen as an irreducible uncertainty because of the many frames in society, which are not necessarily recognised. It is not about ‘not knowing enough’, but about ‘knowing differently’ (Van den Hoek et al., 2014). In the example of a possible failure of a dike, ambiguity can arise from the existence of different perspectives on the acceptability of such failure, and hence different perspectives on the amount of investment required to either reduce epistemic uncertainty and/or build stronger dikes to prepare for higher water levels.

The level of uncertainty

The level of uncertainty refers to the degree of certainty that can be achieved in a given situation. Different levels of uncertainty can be distinguished between full certainty and complete uncertainty or between knowing everything precisely and a total lack of knowledge (Walker et al., 2003). Kwakkel et al. (2010) proposed to redefine the levels (Table 1) to indicate the possibility of quantifying uncertainty in terms of probabilities or likelihood of facts or events (Hofer, 1996; Kwakkel et al., 2010; Skinner et al., 2013).

Table 1.

The four levels of uncertainty (Kwakkel et al., 2010: 308–309).

Level of uncertainty	Description
Level 1 – Shallow uncertainty	Being able to enumerate multiple alternatives and provide probabilities.
Level 2 – Medium uncertainty	Being able to enumerate multiple alternatives and rank them in terms of perceived likelihood without being able to compare likelihoods of alternatives (ordinal).
Level 3 – Deep uncertainty	Being able to enumerate multiple alternatives without being able to rank them in terms of likelihood or plausibility.
Level 4 – Recognised ignorance	Being unable to enumerate multiple alternatives while admitting the possibility of being surprised.

The location of uncertainty

Locating uncertainty can originate from a modeller’s perspective, in which the location is where an uncertainty manifests itself in a model (Walker et al., 2003), or from a general risk assessment perspective, in which it is ‘where the uncertainty occurs within an assessment’ (Skinner et al., 2013: 3). Based on these definitions, the location of uncertainty includes, among others, the modelled system (and its boundaries), input data, the model itself (both in representing the real world and in parameters) and the accumulated uncertainty in the outcomes of a model.

Instead of location, some use the ‘source’ of uncertainty in a similar fashion. For example, Van der Keur et al. (2008) redefined the location of uncertainty as the source of uncertainty based on three ‘subsystems’ (ecological, social and technical) and distinguished different sources (including data, the model, boundary conditions) of uncertainty for each of these subsystems (Van der Keur et al., 2008). Both source and location describe where uncertainty can be located, depending on the question, ‘what is uncertain?’

Relating planning to the characteristics of uncertainty

The characterisation of uncertainty by nature, level and location, as described above, differs from the four uncertainty frameworks in planning and offers additional insights for dealing with uncertainty. However, the possibilities for dealing with uncertainty through planning interventions are influenced by the particularities of an individual’s planning perspective (Christensen, 1985). Therefore, we first elaborate on the kind of ontological position and related perspective on planning that could match the characteristics of uncertainty.

We argue that an understanding of planning should be consistent with the ontological perspective taken. Here, we adopt a perspective in which we acknowledge that the world behaves in complex, intricate ways beyond our grasp (De Roo and Silva, 2010; DeLanda, 2006). In addition, we claim that at least some elements are knowable, without neglecting the intricate ways in which physical interactions and interpretations make reality too complex to grasp (Bhaskar, 1998; DeLanda, 2006; Harman, 2008). We assert the existence of poly-rationality and poly-epistemology (Davy, 2008), which makes knowledge both object-dependent and subject-dependent. The former implies that there is just one knowable world and that maladaptive planning is possible and possibly dangerous for human existence; the latter implies that there are multiple rationalities for interpreting the complex world and acting accordingly.

This ontological position allows for a plural–unequivocal understanding of uncertainty, which is necessary to embrace the three characteristics of uncertainty: nature, level and location. Taking this position can help us to transcend different understandings of uncertainty (Christensen, 1985) derived from the rational (e.g. Walker et al., 2003) and relational perspectives (e.g. Brugnach et al., 2008; Islam and Susskind, 2013; Van den Hoek et al., 2014) and construct a coherent frame of understanding that acknowledges both the plurality of uncertainty and its unequivocal implications for adequate and just interventions.

We build on the understanding that planning is primarily a practice (Alexander, 2015) shaped by planners who constantly navigate amid different paradoxes (Savini et al., 2015). Planners build on knowledge to navigate planning practices through multiple, connected ways of knowing (Davoudi, 2015). For our purpose, planning can thus be seen as a vehicle planners use to turn their knowledge into adequate interventions in particular situations (e.g. Alexander, 2015; Allmendinger, 2009; Van der Vlist, 1998). We see this as an activity that predominantly takes place within public institutions. We argue that, in accordance with our ontological perspective, the performance of planning and planners can be assessed against general, universal claims about uncertainty, which need to be specified for situated planning practices (Alexander, 2015; Campbell, 2006; Savini et al., 2015). In this view, we propose that distinct characteristics of uncertainty provide planners with a rationale for investigating the adequacy and justification of their interventions.

We connect uncertainty to planning via its guiding facet and its developmental facet. In the guiding facet of planning, planners produce legitimating discourses. In the developmental facet of planning, planners search for regulatory mechanisms for interventions (Davoudi, 2015; Olssen, 1999; Van der Vlist, 1998). Based on these premises, we relate the three characteristics of uncertainty to what can be known and what can be done in planning. Ontic uncertainty is an unknowable phenomenon, which implies that planning interventions necessarily take place in uncertainty. Epistemic uncertainty can be known, and thus implies that planners could reduce uncertainty by increasing or correcting current knowledge. Ambiguity arises from situations in which actors have different knowledge and perceptions, which implies that planners should search for a single way of knowing. Each of these claims builds on the idea that one can handle uncertainty in adequate ways, justified by the relevant premises concerning the type of uncertainty involved. The level and location of uncertainty can also be described for both facets of planning. All this is illustrated in Table 2.

Table 2.

A plural–unequivocal analysis of uncertainty related to the guiding and developmental facets of planning.

Planning	Nature			Level	Location
Planning	Ontic	Epistemic	Ambiguity	Level	Location
What can be known (guiding facet)	Unknowable phenomena up to a specific level	Knowable phenomena up to a specific level	Differently (un)known	Degree of certainty to which a phenomenon can be known	Knowledge about where a phenomenon can become manifest
Prescriptive to what can be done (developmental facet)	Intervening in uncertainty	Enlarging or correcting current knowledge	Search for one way of knowing	Specify what can be done regarding the measurability of phenomena	Specify what can be done regarding the possible location where uncertainty manifests itself

This analysis is still very abstract and needs to be specified in more detail to obtain an answer to our research question about what insight this differentiation of uncertainty offers for making adequate and warranted planning interventions. Such specification might be found in what each of the characteristics of uncertainty implies for the possible interventions planners make. Moreover, each of the characteristics holds some normative implications for warranted courses of action. We explore these topics in the following sections.

What are the implications of the ontic and epistemic natures of uncertainty for planning?

The nature of uncertainty has far reaching implications for planning, since it describes the possibility or impossibility of reducing uncertainty. Based on the distinction between ontic and epistemic natures, planners could ask whether or not there is room for reducing uncertainty and what type of knowledge could deliver more insight into uncertainty and what possible and warranted actions could be taken, despite the uncertainty.

Ontic uncertainty is by definition irreducible. If ontic uncertainty is encountered, for example, the unknown effect of climate change on the future discharge of a river, planners have to act under uncertainty, which implies that they must accept the risks involved. Most planning actions taken to deal with ontic uncertainty will probably be designed to make the object of planning able to withstand or adjust to how the uncertain phenomenon might unfold. In such situations, planners could opt to enhance the robust, adaptive or flexible properties of the object. For example, in response to uncertain future river discharges, the dikes could be strengthened or otherwise adapted to make them more robust (Klijn et al., 2012) or flexible (Scholtes and De Neufville, 2011).

Epistemic uncertainty can by definition be reduced by gaining more knowledge or conducting directed experiments. In this sense, epistemic uncertainty is similar to ‘uncertainties in the working environment’, which need deeper investigation, in the strategic choice approach (Friend, 1993). Planners confronted with epistemic uncertainty may respond by making plans that are adaptive or flexible (Kato and Ahern, 2008). Adaptive planning builds on the ability to alter the development path laid out in plans (Balducci et al., 2011) as new evidence arises, the predictability of a phenomenon increases or as the future unfolds (Kato and Ahern, 2008; Rauws et al., 2014). Of course, responses to uncertainty being reducible or irreducible are not mutually exclusive. As well as taking time to measure river discharges to increase knowledge with more specifications or details, the inclusion of robustness or flexibility in the object of planning might also be deemed appropriate. Investing in such robust or flexible properties to handle epistemic uncertainty can, however, lead to maladaptation or overinvestment (Van de Riet, 2003). This could be avoided by first investing in the knowledge base. Knowing the difference between ontic or epistemic uncertainty and the implications for what can be done allows judgements to be made about an adequate course of action.

There are also several normative issues pertaining to ontic and epistemic uncertainties. Being uncertain and being able or unable to reduce uncertainty defines the responsibility of planners to act (or not), and in which way. This bears a strong resemblance to the precautionary principle in environmental ethics (Gardiner et al., 2011; Munthe, 2011; O’Riordan and Cameron, 1994). If planners do not know what the effects of an action are going to be, they should not do it to avoid possible damage. However, if they do know that a harmful change or outcome is likely, they could be considered to have a moral responsibility to find out what might happen and what could be done to reduce harm, based on their personal moral principles, or those of their planning institute or society in general (Basta, 2014). Uncertainty determines what the type and extent of precautionary action might be, although the level and location of uncertainty are just as important in determining appropriate (precautionary) action.

Whether or not planners can be held accountable for inadequate or inappropriate action depends on whether or not they are able to obtain the knowledge required to act (epistemic vs. ontic uncertainty). In the face of ontic uncertainty, precaution might be the just choice. In the face of epistemic uncertainty, acting to reduce uncertainty to better guide actions might do more justice to the situation at hand. In both instances, the planner (or planning institution) is morally responsible for their actions or inaction, which can be assessed against the uncertainty at hand (Fischer and Ravizza, 1998). There is, therefore, a strong imperative to know the distinction between ontic and epistemic uncertainty for judging planning decisions and addressing planners’ accountability for the type and extent of actions taken amid uncertainty.

Frames, ambiguity and disagreement in addressing uncertainty in planning

Two issues need to be resolved regarding ambiguity in planning. The first is the relation between ambiguity and disagreement and the second is the relation between ambiguity and value, means and choice uncertainty. In addition, the nature of ambiguity itself challenges conceptualisations of uncertainty since it arises from the existence of different frames, as does uncertainty. We elaborate on these issues and address the connection between ambiguity and planning by introducing discursive uncertainty as pertinent to all discussions of uncertainty in planning.

The first issue in need of clarification is the relation between disagreement in planning frameworks and ambiguity as a separate nature of uncertainty. Ambiguity is a form of uncertainty arising from different frames in society (Brugnach et al., 2008; Dewulf et al., 2005). Disagreement is the outcome of colliding frames, values or positions among actors involved in a planning process, which can lead to uncertainty about the planning process or its outcomes (Susskind et al., 1999). This problem relates to how a frame is defined, which can be done in at least two different ways. As a time-slice based concept, the frame is defined as a ‘cognitive representation’, but when historically situated, the frame is seen as an ‘interactional co-construction’ (Brugnach et al., 2008; Dewulf et al., 2005). When disagreement is related to the perception of frames as interactional co-constructions, it can be seen as the absence of a unanimously accepted and clear understanding. Ambiguity can be seen as the mere existence of different cognitive frames pertaining to an issue. While disagreement and ambiguity result from the same multiplicity of frames (as cognitive representations) in society (Brugnach et al., 2008; Kwakkel et al., 2010), the difference is between planning as a situated practice, in which a multiplicity of frames results in disagreement, and the contextual situation, in which the same multiplicity of frames results in ambiguous uncertainty (Abbott, 2005). Ambiguous uncertainty is, by definition, irreducible from the perspective of the situated planning practice, whereas disagreement can be dealt with in a planning process (Innes and Booher, 2010; Susskind and Field, 1996). One of the possible interventions for resolving disagreement, and to some extent ambiguous uncertainty where it can be drawn into the planning situation, is to align frames in a consensus-building process or to reframe – that is, to transform the collective understanding of the planning problem, data, information or the scope of the process – through interactional co-construction or joint fact-finding (Abbott, 2005; Brugnach et al., 2008; Schenk et al., 2016; Susskind and Field, 1996).

The second issue that needs attention is the categories identified in planning frameworks: value uncertainty, means uncertainty and choice uncertainty (Abbott, 2005; Friend and Hickling, 2005; Islam and Susskind, 2013). The distinction between uncertainty about values and choices (about means) can be useful when determining the aim of a consensus-building process (Susskind and Field, 1996). The concept of guiding values from the strategic choice approach (Friend and Hickling, 2005) corresponds to differences between values that underlie frames, while choice uncertainty refers to uncertainty about the appropriate or desirable choices or means, due to different frames. Choice uncertainty can also be epistemic in nature, or partly so, if there is insufficient knowledge about the effects of a choice. So, value and means uncertainty can both be related to ambiguity, but means uncertainty can also be related to epistemic uncertainty.

An important feature of ambiguity is the existence of frames or multiple rationalities, which also pertain to what uncertainty, planning and just action are (Davy, 2008). We propose a planning perspective that enables uncertainty to be seen as a plural–unequivocal concept, but this is just one heuristic for understanding uncertainty. In planning, uncertainty is discursively constructed, often with differences between actors’ cognitive representations of reality (Dewulf et al., 2005; Hillier, 2013). There can, for example, be disagreement between actors about an uncertainty being ontic or epistemic, or being a level 2 or 3 uncertainty. This can be termed discursive uncertainty: uncertainty arising from disagreement or being uncertain about the uncertainty of phenomena. Discursive uncertainty affects actors’ perceptions of appropriate action. Moreover, the discursive practice of interactional co-construction and aligning frames can lead to agreements with wrong assumptions about the characteristics of uncertainty. When planners intend to align alternate representations of uncertainty, awareness of the possible discursive uncertainty could be helpful in elaborating different uncertainty discourses.

Ambiguity can provide insight for planning, for example on the representation of different frames and values among stakeholders in planning processes. Here, as Vanessa Watson (2003) argues, the danger of underestimating differences arises when ‘planners assume a shared rationality where it does not exist’ (p. 403). This implies several normative–procedural choices. First, who should be represented to include or avoid the exclusion of ‘frames’ in the co-construction of understanding pertaining to uncertainties. Second, if co-construction or reframing is identified as an adequate intervention, issues arise pertaining to equal access of actors to the process, fairness in processes of inclusion or exclusion, and respect for differences regarding uncertainty or other issues not pertaining to the actual planning issue. Also, for effective co-construction or reframing, discussing the appropriate level of reframing needed to cope with uncertainty can be important to do justice to multiple frames (does the issue concern reframing of values, means, actions, or of uncertainty itself?). This requires that actors accept the reduction of a multiplicity of frames to a collective perception of reality for the planning issue at hand.

Normative issues also pertain to the distributive and non-distributive effects of choices and means. Uncertainty might hamper timely choices or implementation of proper means, resulting in injustice to currently disadvantaged groups or future generations. The intergenerational aspect is one of the main moral issues debated in climate change ethics (Gardiner, 2006; Gardiner et al., 2011). While ontic and epistemic uncertainties are considered relevant to ethical considerations about climate change, ambiguity receives less attention, although it can offer a different perspective. In discussing uncertainty about choices and means, different cognitive representations of uncertainty (what we described as discursive uncertainty) about the solutions for an issue affect the adequate handling of uncertainty. Instead of discussing the effect of ontic or epistemic uncertainty on interventions, ambiguity challenges the underlying representations of reality and offers a different set of arguments and solutions related to co-construction and aligning frames.

Level of uncertainty in planning and its connection with risk

The level of uncertainty specifies possible ways of dealing with uncertainty regarding the timing and use of planning interventions. The level of uncertainty is only a descriptor for ontic and epistemic uncertainties, which may both be present in an uncertain phenomenon at a specific moment (cf. Walker et al. 2003). While ambiguity could be expressed in levels, this is probably a futile exercise as probability ranking of different frames does not make much sense. A better option may be to make a distinction between unanimous clarity and total confusion (Brugnach et al., 2008). However, ambiguity can give rise to discursive uncertainty, in which there is a debate about the level at which an uncertain phenomenon can be specified.

The level of uncertainty is strongly connected to the concept of risk. According to Knight (1921), risk is ‘measureable uncertainty’, equivalent to levels 1 and 2 (Table 1); immeasurable uncertainty is represented by levels 3 and 4 (p. 20). In planning, risk is often described in terms of the probability of an event and its impact or consequences, which can be further specified according to the exposure to the event and vulnerability to its impacts (Klijn et al., 2012; Renn, 2008). Risk management approaches that can be used to deal with uncertainty levels 1 and 2 include probabilistic and deterministic approaches and statistical-based assessment approaches, such as real options analysis (Scholtes and De Neufville, 2011).

Level 3 uncertainty can be linked to pathway approaches and scenario planning (Haasnoot et al., 2013). Level 4 can be linked to those uncertainties that we do not know (the ‘known unknowns’), in which planning needs to be open to surprises or ‘black swans’ (Kwakkel et al., 2010; Taleb, 2007). It is this level 4, ontic uncertainty that has led to demands for openness, reflexivity, responsiveness and experimental practices in planning (Ansaloni and Tedeschi, 2016; Rauws et al., 2014). The link between uncertainty and risk can become a key issue in translating knowledge about uncertainty into action, because both action and inaction can, but do not necessarily, result in risk-taking activities. The level of uncertainty determines in more detail which approaches and solutions are suitable under conditions of uncertainty.

The level is an important concept to help planners adequately handle uncertainty. The levels are conceived of as a continuum, from fully certain about a phenomenon at one end to totally uncertain at the other (Walker et al., 2003). For planners, decisions on when and how to act are made easier by knowing about the extremes and dynamics in the degree of uncertainty over time (Islam and Susskind, 2013). In addition, the size of uncertainty can be helpful in describing the behaviour of uncertainty over time and in comparing different uncertainties. The total size of uncertainty about a problem may consist of the sum of many different small uncertainties or one large uncertainty, but for the sake of clarity, when we refer to the size of uncertainty we mean the size of one uncertain phenomenon. The size of uncertainty can help to describe the behaviour of uncertainty over time, and can help to compare different uncertainties relative to one another. If the size of uncertainty is expected to decrease over time, planners would be advised to wait until there is less uncertainty before acting. If it remains the same or might increase, taking action now could be better than waiting for future change. The size of uncertainty is vital in several approaches to dealing with uncertainty, such as options analysis, scenario planning and pathways approaches (Haasnoot et al., 2013; Scholtes and De Neufville, 2011). The size of uncertainty is also often a major issue in public and scientific debate, a highly topical example being the climate change debate.

The level of uncertainty also has a normative aspect, mainly concerning the substantive issues of justice (Davy, 1997). Whereas risk always focuses on negative consequences, uncertainty introduces a positive, ‘upward’ connotation. This neutral direction of decisional relevance adds the possibility of enabling just distributions in time and space (Savini et al., 2015). Moreover, risk ethics can inform up to the second level of uncertainty (usually regarding only the probabilities of negative consequences and the effects of hazardous events) about what is the proper extent of precautionary action (or inaction) and what is the optimal way to proceed with regard to the distribution of the burden of injustice (e.g. Basta, 2014; Davy, 1997; Munthe, 2011; Van Asselt and Vos, 2006).

In addition to the information provided by the ontic nature of uncertainty on the distributional effects of acting (or not acting), the level of uncertainty could throw light on the fairness of the outcomes of possible interventions. We think an added benefit of uncertainty is that it shifts the perspective from risk-based decisions, which aim to minimise negative effects, to a substantive justice perspective in which the aim is to find the optimum balance of positive and negative effects (e.g. Davy, 1997). This presumes the existence of an optimal solution, or at least proportional differences in the fairness of interventions and predefined categories for apprehending reality to avoid ambiguity or discursive uncertainty regarding the planning issue.

Locating uncertainty

The location of uncertainty can provide further information on the source of uncertainty and where best to act. Current descriptions of the location of uncertainty are, however, not suitable for planning. While the nature and level of uncertainty are generic descriptions, the location of uncertainty is more specific to the domain of interest and is often context-dependent (Funtowicz and Ravetz, 1990; Skinner et al., 2013). Current descriptions are based on modelling studies, identifying the modelled system, its context and technical aspects within a modelling study as possible locations (Kwakkel et al., 2010; Walker et al., 2003). Van den Hoek et al. (2014) has used a systems perspective to locate uncertainty and assess the cascading of uncertainty between different subsystems.

We propose to take a simplified socio-physical system as a minimal heuristic to query the location of uncertainty relevant to spatial planning (Healey, 2007; Van der Vlist, 1998). As a minimum, uncertainty can then be located within the field of interaction between social (cultural, political, economic, etc.) and physical dimensions. Locating uncertainty is identifying where in such a system uncertainty is present, which depends on the specific planning context. In this conceptualisation, uncertainty concerning the social system relates to, among others, human activities, roles and cultural patterns. Physical uncertainty concerns the patterns and behaviour of the environmental systems. The socio-physical system forms a first subdivision to provide information for locating uncertainty and responding to uncertainty at the most adequate place. In some cases, uncertainty can cascade through causal relationships between subsystems (Van den Hoek et al., 2014). For example, the uncertain effects of climate change may induce planners to permit (legal) flexibility in land use to enable flood management interventions at a later date (Tasan-Kok, 2008), which in turn creates uncertainty for farmers if they want to take land-based investment decisions. Both the location of uncertainty and the geographical scale change, while the effects and related risks of (handling) uncertainty are transferred to other actors in society.

This example also illustrates an important normative implication of the location of uncertainty. The location of an uncertainty relates to who bears responsibility for acting (or not) on a particular uncertainty, and the possibilities for deliberately altering the externalities of an uncertain phenomenon. The location of an uncertainty is often not singular. In the example above, the uncertain effects of climate change are multifarious, which can make it difficult to determine the related responsibility to act, although locating uncertainty at least brings into the open the issue of who should probably bear responsibility and suggests who might be responsible for climate change adaptation if its impacts are not yet clear (Mees et al., 2012; Nalau et al., 2015; Thompson and Bedik-Keymer, 2012).

A second, related normative element pertains to the deliberate transfer of externalities of uncertainty to others. This transfer can be assessed for fairness. Is it fair to transfer these externalities, as in the example, from a state actor to a set of private actors? Without wanting to go into the ethical debate about distributional versus procedural assessment of this question (cf. Fischer and Ravizza, 1998), it might be worth pursuing a just transfer of externalities of acting amid uncertainty.

From the three characteristics of uncertainty and the different insights each offers for what can be known and what can be done, we can derive different types of interventions appropriate for handling uncertainty in planning and identify the normative implications of such interventions. These are summarised in Table 3. This offers a heuristic framework for navigating amid uncertainty in planning, with pointers for dealing with each of the distinct natures of uncertainty and specifications based on the level and location of uncertainty.

Table 3.

Possible interventions and normative implications for navigating amid uncertainty for the three characteristics of uncertainty.

Planning	Nature			Level	Location
Planning	Ontic	Epistemic	Ambiguity	Level	Location
What can be known (guiding facet)	Unknowable phenomena up to a specific level	Knowable phenomena up to a specific level	Differently (un)known	Degree of certainty to which a phenomenon can be known	Knowledge about where a phenomenon can become manifest
Prescriptive to what can be done (developmental facet)	Intervening in uncertainty	Enlarging or correcting current knowledge	Search for one way of knowing	Specify what can be done regarding the measurability of phenomena	Specify what can be done regarding the possible location where uncertainty manifests itself
Possible interventions	Primarily aimed at the object of planning practice (e.g. adaptive, robust or flexible objects and institutes)	Primarily aimed at the trajectory of plans and planning (e.g. adaptive, robust or flexible plans/planning)	Primarily aimed at co-construction of knowledge and frame alignment	Risk management, scenario planning, and remaining reflexive and open to self-organising and emergent properties	Pertaining to geographical or institutional scale for adequate interventions
Normative implications	● Are the effects of (not) intervening just, based on current knowledge about uncertainty? ● To what extent does not knowing determine the type and extent of precautionary actions?	● Does improved/altered knowledge lead to more just interventions? ● Could the planner have known more, and hence be responsible for not or wrongly acting?	● Is knowledge acceptable for all relevant frames in the particular planning situation? ● Are procedures for co-construction fairly inclusive?	● Does the level pertain to fair outcomes of interventions? ● Can interventions lead to fairer results over time (intergenerational)? ● Fairer over space (intra-generational)? ● Is there an optimal solution to strive for in essentially unjust situations?	● Does (non)intervention amid uncertainty about locality and scale do justice to distributive and non-distributive principles? ● Who bears responsibility for handling uncertainty in specific localities? ● Is deliberate transfer of the effect of uncertainty just?
Meta-ethical considerations	Presumption of best and fairest solution? Presumption of predefined categories to apprehend reality?

Navigating amid uncertainty in planning

There is a close connection between what is uncertain, what can be known and what can be done. The three characteristics of uncertainty inform this connection and could offer insights into how planners can navigate amid uncertainty. Existing planning frameworks do not provide a coherent enough understanding of uncertainty to inform adequate ways of handling the different characteristics of uncertainty (Abbott, 2005; Christensen, 1985; Friend, 1993; Islam and Susskind, 2013). Moreover, planning methods are tailored only to specific uncertainties (Haasnoot et al., 2013; Lempert and Groves, 2010; Walker et al., 2013) and while communicative planning tools effectively address ambiguity, they neglect other uncertainties (Innes and Booher, 2010). Effective integration of uncertainties can be hampered by a relational–constructivist vision of uncertainty, for example, about climate change, because this can become subject to negotiation and consensus-building processes, which may lead to maladaptation. Our contribution aims to amend these shortcomings. An understanding of what can be known about uncertainty can enable planners to make decisions about adequate planning interventions and take into account their normative implications. This can lead to interventions amid uncertainty and inform the normative implications for acting (or not) according to the distinct characteristics of uncertainty. In situations where planners are confronted with a world known in different ways, they can either strive to bring about a consensus on one way of knowing or accept incompatible ways of knowing, as long as this permits adequate spatial interventions. Further specifying the degree to which the world is uncertain, and where phenomena can become manifest, could aid navigation between ontic, epistemic and ambiguous uncertainties and the level and location of uncertainties.

Our conceptualisation of uncertainty in planning also raises some normative implications. Should planners strive for one way of knowing? Should they try to reduce uncertainty? This depends on the particularities of each planning situation. Contending that planning must be conceived as something specific, defined by its complex context and the behaviour of actors and planners, also makes an exploration of possibilities for handling uncertainty an interpretative and contextually embedded endeavour. With this in mind, each of the characteristics offers insight into what can be known and what adequate and just interventions might entail. Together, this offers a heuristic for navigating amid uncertainty.

Conclusion

We have presented a plural–unequivocal understanding of uncertainty which we argue offers pointers for manoeuvring planning processes through issues of uncertainty. Drawing on planning as a situated practice of navigating, we set out to re-examine the concept of uncertainty by distinguishing between different types of uncertainty and how they might inform planning as a situated practice. Although uncertainty has enjoyed a reasonable degree of attention in planning studies, we conclude that our exploration contributes to a fuller description of the possible types of planning interventions when confronted with uncertainty.

The normative implications of a plural–unequivocal understanding of uncertainty are still only highly schematic, and how planners can deal with uncertainty should be further explored. By viewing uncertainty as a plural–unequivocal concept, we have identified different characteristics of uncertainty and concomitant ways of being uncertain, and diverging ways of knowing. By investigating the implications for the guiding and developmental character of planning, we were able to draw out insights for planning interventions and summarise them. These provide pointers for discussing how to handle uncertainty in situated planning practices.

The normative questions we raised from our examination of a plurality of unequivocal uncertainties need to be discussed further; any answers will depend on the rational or moral framework within which they are examined (Davy, 2008; MacIntyre, 1988). This challenges the planner, and the role of morality in planning processes, because the unequivocal implications of uncertainty for the adequacy of actions (a dike can still fail, causing flooding and deaths) need to be accounted for. Planners and planning institutes may be held morally and legally accountable for failing to take adequate action amid uncertainty. Planners cannot circumvent this responsibility. Essentially, they need to act in proportion to uncertainty and bear the consequences of their actions. It is our hope that this exploration contributes to a better understanding of uncertainty in planning and aids the further structuring of spatial interventions to avert the consequences of climate change and other uncertain developments. The types of interventions and their normative implications discussed here, although in need of further exploration, offer pointers to planners to help them justify the proportionality of their actions and navigate amid uncertainty.

Footnotes

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

Author biographies

Mark Zandvoort is researcher in adaptive planning and water management at Wageningen University, The Netherlands.

Maarten J Van der Vlist is special associate professor at Wageningen University and principal expert adaptive delta management and land use planning at Rijkswaterstaat, The Netherlands.

Frans Klijn is professor of Adaptive Delta Planning at Delft University of Technology and senior specialist at Deltares on flood risk management, both in The Netherlands.

Adri Van den Brink is professor and Chair of Landscape Architecture at Wageningen University.

References

Abbott

(2005) Understanding and managing the unknown: The nature of uncertainty in planning. Journal of Planning Education and Research 24: 237–251.

Abbott

(2009) Planning for complex metropolitan regions: A better future or a more certain one. Journal of Planning Education and Research 28: 503–517.

Abbott

(2012) Planning as managing uncertainty: Making the 1996 livable region strategic plan for greater Vancouver. Planning Practice & Research 27(5): 571–593.

Albrechts

(2004) Strategic (spatial) planning reexamined. Environment and Planning B: Planning and Design 31: 743–758.

Alexander

(2015) There is no planning – only planning practices: Notes for spatial planning theories. Planning Theory 15: 91–103.

Alfasi

Portugali

(2004) Planning just-in-time versus planning just-in-case. Cities 21(1): 29–39.

Allmendinger

(2009) Planning Theory. London: Palgrave Macmillan Ltd.

Ansaloni

Tedeschi

(2016) Ethics and spatial justice: Unfolding non-linear possibilities for planning action. Planning Theory 15(3): 316–332.

Balducci

Boelens

Hillier

et al . (2011) Introduction – Strategic spatial planning in uncertainty: Theory and exploratory practice. Town Planning Review 82(5): 481–501.

10.

Basta

(2014) Siting risky facilities: Probabilism, determinism and beyond. Planning Theory 13(1): 44–64.

11.

Batty

(2013) The New Science of Cities. Cambridge, MA: The MIT Press.

12.

Bertolini

(2010) Coping with the irreducible uncertainties of planning: An evolutionary approach. In: Hillier

Healey

(eds) The Ashgate Research Companion to Planning Theory. Farnham: Ashgate Publishing, pp. 413–424.

13.

Bhaskar

(1998) The Possibility of Naturalism: A Philosophical Critique of the Contemporary Human Sciences. 3rd edn. London and New York: Routledge.

14.

Brugnach

Dewulf

ARPJ

Pahl-Wostl

et al . (2008) Toward a relational concept of uncertainty: About knowing too little, knowing too differently, and accepting not to know. Ecology and Society. 13(2): 30.

15.

Campbell

(2006) Just planning: The art of situated ethical judgment. Journal of Planning Education and Research 26: 92–106.

16.

Christensen

(1985) Coping with uncertainty in planning. Journal of the American Planning Association 51(1): 63–73.

17.

Connell

(2009) Planning and its orientation to the future. International Planning Studies 14(1): 85–98.

18.

Davoudi

(2015) Planning as a practice of knowing. Planning Theory 14(3): 316–331.

19.

Davy

(1997) Essential Injustice. Heidelberg: Springer.

20.

Davy

(2008) Plan it without a condom! Planning Theory 7(3): 301–317.

21.

De Roo

Silva

(2010) A Planner’s Encounter with Complexity. Farnham: Ashgate Publishing.

22.

DeLanda

(2006) A New Philosophy of Society: Assemblage Theory and Social Complexity. London and New York: Continuum.

23.

Dewulf

ARPJ

Craps

Bouwen

et al . (2005) Integrated management of natural resources: Dealing with ambiguous issues, multiple actors and diverging frames. Water Science and Technology 52(6): 115–124.

24.

Fischer

Ravizza

(1998) Responsibility and Control: A Theory of Moral Responsibility. Cambridge: Cambridge University Press.

25.

Friend

(1993) Planning in the presence of uncertainty: Principles and practice. Journal of Infrastructure Planning and Management 476(21): 1–9.

26.

Friend

Hickling

(2005) Planning under Pressure: The Strategic Choice Approach. Oxford: Elsevier, Butterworth Heinemann.

27.

Friend

Jessop

(1969) Local Government and Strategic Choice: An Operational Research Approach to the Processes of Public Planning. London: Tavistock Publications.

28.

Funtowicz

Ravetz

(1990) Uncertainty and Quality in Science for Policy. Dordrecht: Kluwer Academic Publishers.

29.

Gardiner

(2006) A perfect moral storm: Climate change, intergenerational ethics and the problem of moral corruption. Environmental Values 15: 397–413.

30.

Gardiner

Caney

Jamieson

Shue

(eds.) (2011) Climate Ethics: Essential Readings. Oxford: Oxford University Press.

31.

Graham

Marvin

(2001) Splintering Urbanism: Networked Infrastructures, Technological Mobilities and the Urban Condition. London & New York: Routledge.

32.

Gunder

(2008) Ideologies of certainty in a risky reality: Beyond the hauntology of planning. Planning Theory 7(2): 186–206.

33.

Haasnoot

Kwakkel

Walker

et al . (2013) Dynamic adaptive policy pathways: A method for crafting robust decisions for a deeply uncertain world. Global Environmental Change 23(2): 485–498.

34.

Harman

(2008) DeLanda’s ontology: Assemblage and realism. Continental Philosophical Review 41: 367–383.

35.

Healey

(2007) Urban Complexity and Spatial Strategies: Towards a Relational Planning for our Times. London: Routledge.

36.

Hillier

(2010) Introduction to Part Two. In: Hillier

Healey

(eds) The Ashgate Research Companion to Planning Theory. Farnham: Ashgate Publishing, pp. 235–250.

37.

Hillier

(2013) On Relationality and Uncertainty. disP – The Planning Review 49(3): 32–39.

38.

Hofer

(1996) When to separate uncertainties and when not to separate. Reliability Engineering & System Safety 54: 113–118.

39.

Innes

Booher

(2010) Planning with Complexity: An Introduction to Collaborative Rationality for Public Policy. London & New York: Routledge.

40.

Islam

Susskind

(2013) Water Diplomacy: A Negotiated Approach to Managing Complex Water Networks. New York: RFF Press, Routledge.

41.

Kato

Ahern

(2008) ‘Learning by doing’: Adaptive planning as a strategy to address uncertainty in planning. Journal of Environmental Planning and Management 51(4): 543–559.

42.

Klijn

Knoop

Ligtvoet

et al . (2012) In search of robust flood risk management alternatives for the Netherlands. Natural Hazards and Earth System Sciences 12: 1469–1479.

43.

Knight

(1921) Risk, Uncertainty and Profit. New York: Augustus M. Kelley.

44.

Kwakkel

Walker

Marchau

VAWJ

(2010) Classifying and communicating uncertainties in model-based policy analysis. International Journal of Technology, Policy and Management 10(4): 299–315.

45.

Lau

(2015) Tackling uncertainties in plan implementation: Lessons from a growth area in England. Town Planning Review 86(1): 7–28.

46.

Lempert

Groves

(2010) Identifying and evaluating robust adaptive policy responses to climate change for water management agencies in the American west. Technological Forecasting and Social Change 77: 960–974.

47.

MacIntyre

(1988) Whose Justice? Which Rationality? Notre Dame, IN: University of Notre Dame Press.

48.

Mees

HLP

Driessen

PPJ

Runhaar

HAC

(2012) Exploring the Scope of Public and Private Responsibilities for Climate Adaptation. Journal of Environmental Policy & Planning 14(3): 305–330.

49.

Milly

PCD

Betancourt

Falkenmark

et al . (2008) Stationarity is dead: Whither water management? Science 319: 573–574.

50.

Munthe

(2011) The Price of Precaution and the Ethics of Risk: The International Library of Ethics, Law and Technology, vol. 6. Heidelberg: Springer.

51.

Nalau

Preston

Maloney

(2015) Is adaptation a local responsibility? Environmental Science & Policy 48: 89–98.

52.

O’Riordan

Cameron

(1994) Interpreting the Precautionary Principle. London & New York: Routledge.

53.

Olssen

(1999) Michel Foucault: Materialism and Education. Westport, CT: Bergin & Garvey.

54.

Rauws

(2015) Why planning needs complexity: Towards an adaptive approach for guiding urban and peri-urban transformations, vol. 4. PhD Thesis: Series InPlanning, University of Groningen, Groningen.

55.

Rauws

Cook

Van Dijk

(2014) How to make development plans suitable for volatile contexts. Planning Practice & Research 29(2): 133–151.

56.

Renn

(2008) White paper on risk governance: Toward an integrative framework. In: Renn

Walker

(eds) Global Risk Governance: Concept and Practice Using the IRGC Framework. Dordrecht: Springer, pp. 3–73.

57.

Salet

Bertolini

Giezen

(2013) Complexity and Uncertainty: Problem or Asset in Decision Making of Mega Infrastructure Projects? International Journal of Urban and Regional Research 37(6): 1984–2000.

58.

Savini

Salet

Majoor

(2015) Dilemmas of planning: Intervention, regulation, and investment. Planning Theory 14(3): 296–315.

59.

Schenk

Vogel

RAL

Maas

et al . (2016) Joint fact-finding in practice: Review of a collaborative approach to climate-ready infrastructure in Rotterdam. European Journal of Transport and Infrastructure Research 16(1): 273–293.

60.

Scholtes

De Neufville

(2011) Flexibility in Engineering Design. Cambridge, MA: The MIT Press.

61.

Simon

(1961) Administrative Behavior: A Study of Decision-Making Processes in Administrative Organization. New York: The Macmillan Company.

62.

Skinner

DJC

Rocks

Pollard

SJT

(2013) A review of uncertainty in environmental risk: Characterising potential natures, locations and levels. Journal of Risk Research 17(2): 195–219.

63.

Stacey

(2007) Strategic Management and Organisational Dynamics: The Challenge of Complexity. Upper Saddle River, NJ: Prentice Hall.

64.

Susskind

Field

(1996) Dealing with an Angry Public: The Mutual Gains Approach to Resolving Disputes. New York: The Free Press.

65.

Susskind

McKearnan

Thomas-Larmer

(eds) (1999) The Consensus Building Handbook: A Comprehensive Guide to Reaching Agreement. Thousand Oaks, CA: SAGE.

66.

Taleb

(2007) The Black Swan: The Impact of the Highly Improbable. New York: Random House.

67.

Tasan-Kok

(2008) Changing interpretations of ‘flexibility’ in the planning literature: From opportunism to creativity? International Planning Studies 13(3): 183–195.

68.

Thompson

Bedik-Keymer

(2012) Ethical Adaptation to Climate Change: Human Virtues of the Future. Cambridge, MA: The MIT Press.

69.

Van Asselt

MBA

Rotmans

(1997) Uncertainty in perspective. Global Environmental Change 6(2): 121–157.

70.

Van Asselt

MBA

(2000) Perspectives on Uncertainty and Risk: The PRIMA Approach to Decision-support. New York: Springer.

71.

Van Asselt

MBA

Vos

(2006) The Precautionary Principle and the Uncertainty Paradox. Journal of Risk Research 9(4): 313–336.

72.

Van de Riet

(2003) Policy analysis in multi-actor policy settings – Navigating between negotiated nonsense & superfluous knowledge. PhD Thesis, TU Delft, Delft, The Netherlands.

73.

Van den Hoek

Brugnach

Mulder

JPM

et al . (2014) Analysing the cascades of uncertainty in flood defence projects: How ‘not knowing enough’ is related to ‘knowing differently’. Global Environmental Change 24: 373–388.

74.

Van der Keur

Henriksen

Refsgaard

et al . (2008) Identification of major sources of uncertainty in current IWRM practice: Illustrated for the Rhine Basin. Water Resource Management 22: 1677–1708.

75.

Van der Vlist

(1998) Duurzaamheid als Planningsopgave: Gebiedsgerichte afstemming tussen de ruimtelijke ordening, het milieubeleid en het waterhuishoudkundig beleid voor het landelijk gebied. PhD Thesis, Wageningen University and Research Centre, Wageningen, The Netherlands.

76.

Van der Vlist

Ligthart

Zandvoort

(2015) The replacement of hydraulic structures in light of tipping points. Journal of Water and Climate Change 6(4): 683–694.

77.

Walker

Haasnoot

Kwakkel

(2013) Adapt or perish: A review of planning approaches for adaptation under deep uncertainty. Sustainability 5: 955–979.

78.

Walker

Harremoës

Rotmans

et al . (2003) Defining uncertainty: A conceptual basis for uncertainty management in model-based decision support. Integrated Assessment 4(1): 5–17.

79.

Walsh

Glendinning

Castán-Broto

et al . (2015) Are wildcard events on infrastructure systems opportunities for transformational change? Futures 67: 1–10.

80.

Watson

(2003) Conflicting rationalities: Implications for planning theory and ethics. Planning Theory & Practice 4(4): 395–407.