未来学素养

武夷山

 

科学技术已经渗透到生活的每一个角落，要适应现代社会，就得具备基本的科学技术素养。同理，人类未来发展前景的不确定性越来越大，我们完全不知道子孙后代会不会过得比我们还糟糕，因此，要想做一个负责任的公民，要想面对扑溯迷离的未来也能坦然应对，就需要具备基本的未来学素养。

Riel Miller在Futures（未来学）杂志2007年第4期提出了未来学基本素养的框架，如下表所示：

未来学素养	任务或使命	方法或技术
第一层：意识	树立时间意识以及将内在的价值观和期望值外显化	向大家征集其对未来的期望的各种方法
第二层：发现	严格想象：要敢于想象，才能突破框子；要严格，才不会漶漫无所归。	社会科学是必不可少的要素
第三层：选择	构建战略场景。战略场景旨在质疑当下决策所依据的假定，不要把战略场景作为规划目标，而要将其作为能对当前世界的潜力带来新启示的东西。	构建战略场景的方法

 

公众的基本语文素养，是强制性义务教育出生的成果，而公众基本语文素养提高后所开拓的种种可能性，事先谁也想象不到。我们有必要问问自己：在21世纪，未来学素养能产生语文素养曾经产生过的那种巨大作用吗？

 

参考文献：Riel Miller，Futures literacy: A hybrid strategic scenario method, Futures, 2007, 39 (4): 341-362

原文如下（与最终发表版也许有差异）：

Futures Literacy: A hybrid strategic scenario method

V 1.3

(May 2006)

Riel Miller

Senior Visiting Fellow, Danish Technological Institute,

Tecknologiparken Kongsvang Alle 29,

KD-8000 Arhus C, Denmark

Tel. 336 3239 1360

rielm@yahoo.com

 

 

Abstract

This article develops a definition of “futures literacy” and outlines a Hybrid Strategic Scenario method for acquiring the capacities of futures literacy.

Keywords

Possibility Space, Learning Intensive Society (LIS), Spontaneous Society, Hybrid Strategic Scenario (HSS) method, Future Studies, Unique Creation, banal creativity, scenarios, knowledge society, forecasting, foresight, 21st century, knowledge intensity, 21st century Transitions, transition scale change, foresight scenarios, strategic scenarios, strategic analysis, rigorous imagining, isovalue scenarios, isoprobable scenarios.

 

         “The human condition can almost be summed up in the observation that, whereas all experiences are of the past, all decisions are about the future. The image of the future, therefore, is the key to all choice-oriented behavior. The character and quality of the images of the future which prevail in a society are therefore the most important clue to its overall dynamics.”

Kenneth Boulding, Foreword in The Image of the Future, Fred Polak, 1973

Introduction

People think about the future all the time. In the morning when they wake-up and start planning the day ahead. At the dinner table when they discuss how to pay the bills next month or what will happen in the Middle East. Most of these reflections are about the short-term, a few days, weeks or months. Such conversations naturally mix together people’s hopes and fears with a wide range of probable to improbable expectations. Professional forecasters trying to predict tomorrow’s weather or next year’s economic growth handle degrees of probability more carefully. Professionals tend to focus on getting to the highest probability prediction that available models and data can provide. They generally steer away from considering the broader, less predictive question of what might be possible as well as the more normative question of what is desirable. This choice, while understandable in light of our deep hunger to know the future, has important limitations.

Two drawbacks of the search for greater predictive accuracy are worth highlighting. First, there is the familiar risk of adopting forecasting methods and models that depend too heavily on what happened in the past. Yesterday’s parameters may do a good job at tracking past events but experience shows that this approach consistently misses major inflection points related to long-run transformative changes [1, 2, 3, 4, 5]. A second, less remarked danger is that a preoccupation with what is likely to happen tends to obscure outcomes that may be unlikely but still possible and potentially more desirable. At best, seeking the reassurance of greater predictive certainty tends to overlook what is viewed as less likely. At worst it lulls us into a false sense of having exhausted the available options, thereby narrowing the set of choices considered actionable. This, in turn, can impair strategic decision making[1] because it leads to an underinvestment in imagining less predictable outcomes and the avenues for reaching them.

Futurists, in their efforts to satisfy the yearning for “scientific”[2] consideration of the future, have been aware of these problems for a long time [40]. Currently, most futurists formally recognise the impossibility of assigning meaningful probabilities to the way a society or organisation might function in ten or twenty years. Notwithstanding this formal nod to the inadequacy and failures of the predictive quest little progress has been made in developing practical alternatives. Three of the many reasons for this inconsistency between theory and practice were/are: a fear of the future that drives a deep desire to know (divine) what will happen (clients want predictions); recent (post WWII) relative systemic stability and the related success of planning in this context[3]; and lack of experience with, and hence under-development of, the conceptual tools and behavioural conventions that make it practical to embrace non-predictive approaches to decision making with success. Old paradigms do not cede easily and the attachment to predictive approaches rooted in trend analysis, forecasting models, multi-factor calculations, etc. are tightly integrated with the way risk is managed and decisions taken in industrial society.

The premise of this paper, without claiming any degree of probability, is that the world around us today, in its evolving conceptual and practical attributes, is creating a context that, on the one hand, is dispensing with industrial era modes of perpetuating systemic stability, risk management, decision making, etc. and, on the other hand, is embracing complexity, heterogeneity and spontaneity as opposed to simplification, homogeneity and planning.[4] As a result, it is hypothesised that modal or non-predictive narratives (scenarios) [41, 53] could become more important and more practical, for two reasons.

One reason is that if the conditions for change in systemic sustainability (change in the conditions of change [43]) are emerging then the role of “what-if” imagining becomes central to making the choices (individual and collective) that, in turn, are pre-requisites for actively pursuing such “transition scale” change [16]. In other words explicit choices, in both the individual (behaviour, aspirations, etc.) and collective (institutions, rules, etc.) spheres, are fundamental to operationalising societal transformation and hence so are the methods for imagining the choices that are outside the old order.[5]

Second, more specific to this particular historical period, is the growing importance of non-predictive imagining due to a shift towards greater spontaneity in decision making. Learning-by-doing, experimentation’s successes and failures, is generating the concepts and techniques for embracing the richer, more accurate information available to just-in-time decision making.[6] Practice is evolving the networking, diversification, openness, trust, and fluidity that makes it feasible to be spontaneous and rely on complexity rather than fight it.[7] All of which calls for an enhanced capacity to imagine the possibilities of the moment without succumbing to the temptation to plan the future based on probabilistic calculations about the unknowable [53].[8] Instead, as discussed in more detail at the end of this article, the capacity for more imaginative storytelling becomes the way to use the potential of the present more effectively and, assuming that people’s choices are consistent with their values, evolving in ways that realise our aspirations.

Clearly the hypothesis that non-predictive scenarios are gaining in importance is itself rooted in a scenario of how post-industrial society might function [10, 13, 16, 17]. A reminder that there is no escaping the interpretive lenses that always filter, reflexively, our perceptions and analyses. From this self-aware starting point the rest of this article offers a partial overview of one way of making sense of how stories about the future may be developed and used in our emergent, perhaps eventually post-industrial societies.

1.           The Challenge of Choosing a Story to Tell

Stories may be what make life intelligible. Certainly humans invent and tell many kinds of stories for many different purposes. Foresight Scenarios (FS) are just one specialised type of story, most often invoked in the context of complex and ambiguous situations [6, 7, 8, 49]. Like with any story, one of the primary challenges is how to imagine and then select a few distinctive and pertinent narratives from the vastness of the imaginable [9]. There are two commonly used methods for solving the problem of how to choose specific stories about the future.

The first, springing directly from the predictive tradition, takes an initial starting point, for instance population or economic output, and then develops scenarios on the basis of a range of growth rates – low, medium and high. This method can be called the baby-bear, momma-bear and papa-bear approach (Bear for short).[9] The second technique focuses more on preferences and expectations in order to sketch scenarios that capture the stories of the futures that people consider to be: the most desirable, the least desirable and the muddling through version that mixes a bit of good and a bit of bad (which also usually happens to be the one people consider most likely or realistic). This method can be dubbed “the good, the bad and the ugly” approach (GBU for short).

Both of these methods generate FS with the virtue that the stories are usually quite familiar. For instance it is easy to recognise Bear scenarios regarding the future of, for instance, universities that are distinguished by differences in enrolment growth rates – low (baby-bear), medium (momma-bear) and high (papa-bear). Or GBU scenarios that are distinguished by the preferences of people whose values, for instance, lead them to consider the “good” scenario to be one where universities are exclusively citadels of a pure search for knowledge, the “bad” scenario to be one where universities are exclusively driven by the commercial imperatives of funders from the private sector, and a muddling through or “ugly” scenario, usually seen as the most likely, that combines both pure and commercial options.

Many foresight exercises generate a composite scenario matrix by mixing and matching a variety of trends (Bear) and preferences (GBU). For example, Table 1 combines three sets of scenarios (1 GBU  and 2 Bear) for the future of tertiary education. Purely for the sake of illustration, the GBU scenarios, based on knowledge-driven versus commercially-driven models of tertiary education are crossed with two categories of Bear scenarios based on 1) technological change and 2) enrolment growth rates. Building up this kind of matrix can rapidly generate numerous stories, 18 in this case, for the future of tertiary education.

Relying on Bear/GBU methods to produce FS has its strengths and weaknesses. On the positive side, Bear scenarios can lean heavily on readily identified trends and the findings of the “predictive social sciences” like economics and demography. While GBU scenarios, generally developed by teasing out the expectations and values of specific constituencies like business executives, groups of experts, citizens in a local community, etc., usually have the virtue of generating stories that are, at least initially, easy to comprehend.

	The Good, The Bad and The Ugly
Bear Scenarios	Knowledge Driven	Commercially Driven	Mixed Model
Low rate of tech change
Low enrolment growth	Scenario 1	Scenario 2	Scenario 3
Medium enrolment growth	Scenario 4	Scenario 5	Scenario 6
High enrolment growth	Scenario 7	Scenario 8	Scenario 9
High rate of tech change
Low enrolment growth	Scenario 10	Scenario 11	Scenario 12
Medium enrolment growth	Scenario 13	Scenario 14	Scenario 15
High enrolment growth	Scenario 16	Scenario 17	Scenario 18

Table 1: Mixing Bear and GBU Scenarios

On the negative side, there is an obvious problem when the scenario matrix gets out of hand as more variables, rates of change and preferences are added. The explosion of permutations and combinations reflects the richness of people’s perceptions but makes it difficult to distinguish the different stories in meaningful ways.[10] Even more damaging, as anyone who has experienced these processes knows, since the variables are often highly incommensurate and lack explicit and/or tested hypotheses regarding basic definitions, causal inter-relationships, or weighting in terms of the model’s parameters, it becomes difficult to use these scenarios for deeper analysis and decision making [26].

How serious are the limitations to FS arising from a reliance on Bear/GBU methods as a way to select and specify stories about the future? The answer to this question depends almost entirely on the purpose or task to which the scenarios are being put. There are clearly certain tasks, like building temporal awareness and revealing shared or divergent assumptions and goals, where using GBU/Bear methods is highly effective, precisely because it surfaces values and expectations. However, crucially, when it comes to policy analysis and strategic decision making the flaws in these two approaches to building scenarios are serious enough to compromise effectiveness.

Three problems with Bear/GBU methods for generating analytical and strategic FS come to the forefront.

First is the limited realm of the possible when, as in most cases of applied scenario development, the GBU/Bear narratives originate either in people’s current (usually unexamined) expectations and preferences or in the framework of predictive modelling familiar in the social sciences. As a result non-conventional and transformative possibilities are under-represented.[11] Second, as already noted, it is more difficult to push the frontiers of the imaginable in actionable (policy choice) ways when the methods used for specifying the scenario variables, usually a structured conversation amongst pertinent constituencies, are rarely rooted in explicit, rigorous[12] analytical-systemic frameworks.

Third the virtue of distinguishing scenarios using either values or expectations turns into a vice when the objective is to imagine a set of equally preferable (isovalue) or equally probable (isoprobable) outcomes or paths for achieving the same outcome. GBU and Bear scenarios, within respective spheres, are inherently different, not equal, when it comes to values and expectations. That is what distinguishes one scenario from another. But, as a result, it is more difficult to generate equally desirable scenarios (isovalue) that can be used to analyse different ways of getting to similar goals or equally likely scenarios (isoprobable) that can facilitate choosing from amongst outcomes that reflect different values. This drawback of FS generated using GBU and Bear methods is a significant handicap for decision makers who are often called upon to make choices from amongst equally desirable or equally (un)likely futures.

To be clear, the weaknesses identified here are not due to the nature of these scenarios as evocative stories about the future nor the utility of using expectations and/or preferences to develop such narratives. The problem is that an exclusive reliance on predictive and/or value based FS results in an incomplete, and hence inadequate interrogation of the potential of the present. Evidence of this poverty of invention, narrow repetitiveness of findings and modest utility in terms of an impact on strategic decision making can be found, for instance, in much of the scenario work done in the education sector, be it on the future of universities [34] or of schooling [30, 33]. The scenarios that predominate are generally the product of single iteration[13] GBU and Bear exercises, sometimes mixed together to produce composite scenario matrices [32].  Similar drawbacks have been noted with respect to many technology foresight initiatives [35]. Without calling into question the positive outcomes of these efforts for teamwork and leadership, this type of exercise has generally failed to produce analytically persuasive or actionable conclusions for policy, particularly policies intended to take into account transformative societal change [45]. 

There is another, less remarked danger that arises from an uncritical reliance on GBU/Bear methods for selecting stories about the future, it is the difficulty of determining which methods are most appropriate for particular tasks. Existing typologies of scenario activities are largely inductive, arriving at the categories by reviewing existing practices differentiated on the basis of the scenario process or goal [9]. A more deductive approach, moving from general assumptions like the modal attributes of a futures studies methodology I have called “history of the future” [18], can generate a capacity related typology that treats story telling about the future as a capability or capacity that can be more or less sophisticated. Taking the capacity point-of-view for constructing a typology gives rise to what I call a “Futures literacy” (FL)[14] framework for categorising thinking about the future.

2.           A Hybrid Strategic Scenario Method for Developing Futures Literacy (FL)

Futures Literacy (FL) is the capacity to explore the potential of the present to give rise to the future. Like language literacy, FL is a variable or cumulative capacity that can be used for many purposes, “good” and “bad”. One level is not “better” than another but Level 1 is a precondition for Level 2 and Level 2 is a precondition for Level 3. This is no different than reading. Learning the alphabet is the first step in acquiring the capacity to decipher text and then being able to read is crucial for beginning to decipher the messages contained in a text. Again, like reading, FL is not the same as the text that is being read nor producing a text to be read. Certainly the capacity to read a text, in its many senses, can be improved through practice, including writing text. But FL is not “the future”, it is the capacity to think about the potential of the present to give rise to the future by developing and interpreting stories about possible, probable and desirable futures.

One way of learning and practicing FL is to use a hybrid and sequential method. Hybrid because the only way to ensure the congruency between scenario techniques and tasks is to use a range of different approaches. Sequential because getting to Level 3 FL requires the capacities and knowledge acquired at Levels 1 and 2. The Hybrid Strategic Scenario (HSS) method developed here, in keeping with the cumulative nature of FL, involves working through and acquiring the skills associated with all three levels of FL as summarised in Table 2.

Futures Literacy:	Task	Technique(s)
Level 1 Awareness	Temporal awareness, shifting both values and expectations from tacit to explicit – all of which builds the capacity of people, teams and leaders to respond and innovate	A wide range of catalysts and processes generate the discussions and sharing of stories that elicit people’s views on what they want and expect in the future
Level 2 Discovery	Rigorous Imagining (RI) involves two distinct challenges – imagination and rigour, the former in order to push the boundaries and the latter so that what is imagined is “scientific”[15] and intelligible	Escaping from the probable and preferable to imagine the possible demands systematic creativity and creating systematically, non-discursive reflection and social science are essential ingredients
Level 3 Choice	Strategic Scenarios are aimed at questioning the assumptions used to make decisions in the present, not as targets to plan-by but to provide new insights into the potential of the current world as a way to embrace complexity, heterogeneity and the pertinence of spontaneous actions that put values into practice	Strategic scenarios are constructed using the capacities and stories acquired in developing Levels 1 and 2 FL, by combining values, expectations and possibilities into scenarios that follow the narrative rules (see Level 1 FL below) and the methods of “history of the future” [18]

Table 2: Levels of Futures Literacy – Tasks and Techniques

2.1 Level 1 Futures Literacy: Awareness

Level 1 FL is largely about developing temporal and situational awareness – meaning a greater appreciation that change happens over time and that particular constituencies, products or organisations can be situated in time according to their values and expectations [6, 7, 8]. By revealing common goals and shared assumptions, the typically discursive, group processes used to develop Level 1 FL help forge stronger teams and build the kind of confidence needed to make decisions about the future. Much applied foresight is practiced at this level and has an admirably successful track record at improving team and leadership capacities [26, 30]. Bear/GBU scenarios, and particularly the learning processes used to generate them, are highly effective at developing awareness that change happens over time, that people do harbour expectations and values, and that choices might matter.

Techniques for building Level 1 FL are legion and, in most cases, do not require a very elaborate catalyst to get the process going.[16] People are primed to express their views about the kind of future they prefer and expect. However, the ease with which dinner table or workshop conversations carry through to the construction of more elaborate scenarios can be deceptive. It is one thing to discuss expectations and preferences, it is another to weave them into a convincing story.  Equally difficult, as teachers know, is to engage in a learning process that develops and embeds new capacities. Meeting both of these challenges calls for careful elaboration of the scenario process, following the rules of good pedagogy and storytelling. Leaving aside the issue, dealt with extensively in many different fields including future studies [47]  of how to design an effective learning process, the issue of narrative structure or the rules of storytelling when it comes to FS merits a brief discussion. At a minimum because it is an aspect of scenario work that is often overlooked.

The narrative of a foresight scenario, like familiar “artistic fiction”, is more intelligible and, hopefully, meaningful if it follows certain rules. In particular it is important to be clear about the following five attributes of the story:

i)         What is the type or purpose of the story?  Not tragedy or comedy, thriller or romance; but is it for contingency planning, simulation training, team building, optimisation testing, or strategic imagining?

ii)       What is the temporal or chronological frame?  Not beginning, middle and end; but comparative static (two or more cross-sections), dynamic trajectory (time-series) or back-casting (reverse engineered)? And, in practice, what is the time-span under consideration (years, decades, generations)?[17]

iii)     What is the (analytical) point-of-view?  Not first or third person, stream-of-consciousness or dialogue; but is the story told in terms of the choices people make in their everyday lives (micro), does it describe aggregate outcomes (macro) or is it about both (linking micro and macro)?

iv)     What are the main protagonists (ultimate decision makers)[18]?  Not hero and villain; but is the agent using the scenario to assist with making choices at a specific institution (firm, school, hospital, etc.) or about a social/economic system (nation, sector, etc.)? And what are the relationships between the agents as decision makers? Not parent/child or detective/criminal; but is the story of a specific agent like a school or a financial market stand-alone, told in terms of its internal functioning only, or is it embedded and interacting with a story of surrounding societal changes in production, consumption, uses of technology, social status, codes of conduct, etc.?

v)       What  rules apply to the action?  Not is time travel allowed or Matrix (the film) like suspension of the rules of this Universe’s physics; but what are the assumptions and hypotheses that underpin the analytical definitions and causal relationships that make for robust social science?

 Explicit answers to these five questions go a long way towards ensuring that scenarios tell an internally consistent story. Helping both the teller and listener to grasp how time, values, and expectations shape our understanding of the potential of the present. A brief “counter-factual” example, that will also be used to illustrate Levels 2 and 3 FL, may help clarify what Level 1 FL is about in the “Hybrid Strategic Scenario” approach. Imagine that it is 18x0, electricity is in its infancy and a group of futures oriented thinkers from the private and public sectors get together to assess the prospects for this new technology. As they discuss its potential they realise that the stories they are telling are very open ended, not only because it is difficult to envisage how the technology will be used but also because industrialisation and globalisation are in full swing. As they refine the stories about electricity’s future they share specific values about access to technology, the role of technology in society and the political principles they believe should underpin decision making – who and how to decide.

In the end they produce a set of GBU scenarios. In the “good” scenario electricity is used to build much bigger factories than was possible with steam or water power and to illuminate the bigger cities that accompany bigger factories. In the “bad” scenario electricity turns out to be too expensive and dangerous to really play a major role. While in the muddling through “ugly” scenario a bit of both the good and the bad lead to an reduced but still important role of electricity in the “factory era”. Of course it would have been very difficult to anticipate that electric motors would become so small and mobile or that power generation could be so effectively centralised and distributed. The predictive capacity of these fictitious futurists is, as expected from a modal perspective, very low. However, in terms of Level 1 FL goals the process has achieved its key targets. The group has rendered explicit a number of values and expectations. They have discovered elements of a “common language” for talking about change and they have probably made considerable progress in defining the subject – the potential role of electricity in society – more clearly. However even the best Level 1 FL stories suffer from the limitations discussed previously. Level 2 FL or “rigorous imagining” is one approach to overcoming the pitfalls of Level 1 FL.

2.2 Level 2 Futures Literacy: Discovery

Level 2 FL futures literacy is the capacity to overcome the limitations imposed by values and expectations on thinking about the future. It is a technique for conducting the potentially paradoxical task of “rigorous imagining”. This is a crucial and challenging step in opening up new insights into the nature and determinants of today’s potential.  Rigorous imagining depends on carefully and consistently distinguishing possible, probable and preferable. Such distinctions are necessary for rigorous imagining because the task of imagining possible futures is logically and practically prior to the assessment of probabilities and preferences.  Prior from a logical perspective because, as illustrated in Figure 1, preferable and probable futures are subsets of the possible.  Prior from a practical perspective since as already pointed out consideration of preferences and probabilities constrain the imagining of possibilities.[19]

Conceivable Futures (Zone 0)

GBU & Bear Scenarios (Zone 4)

Possibility space stories (Zone 5)

Strategic

Strategic

Strategic

Probable Futures (Zone 2)

Possible Futures (Zone 1)

Desirable Futures (Zone 3)

Figure 1: Strategic Scenarios and Possibility Space Scenarios

Starting with the logical, a useful first step in expanding the range and analytical content of possible futures is to clarify, at a conceptual level, what distinguishes conceivable, possible, probable and desirable scenarios. Figure 1 is an illustration of one way of defining these categories and the logical inter-relations. The largest set consists of conceivable scenarios (Zone 0). A sub-set of conceivable futures are those that are possible (Zone 1). Nested within the set of possibilities are probable futures (Zone 2) and some of the desirable ones (Zone 3).[20] There are, of course, desirable futures that are neither probable nor possible, even if conceivable. Hence Zone 3 spills over from Zone 2 into Zones 1 and 0.[21]

Describing, as in Figure 1, the infinite number of stories that make up the whole scenario space, does not in-and-of-itself solve the problem of how to chose specific stories from within that space. As already discussed, GBU/Bear methods offer one way of identifying the variables and selecting scenarios (Zone 4), but in ways that are limited by the pursuit of scenarios based on values and/or probabilities. The challenge then is to use non-predictive, value neutral methods to identify both the variables and the scenarios that are possible (Zone 5).

Rigorous imagining can be conducted in a multitude of different ways. However, it is one of the distinguishing features of Level 2 FL that many of its story selection methods are incompatible with those of Level 1 FL. The definitional differences are obvious since discovering what is possible (minimally probable) is not the same as determining what is likely (more than minimally probable) or desirable. The practical differences arise from the nature of the competences entailed by different levels of FL. Level 1 FL largely involves shifting knowledge from its tacit to explicit form, i.e. what people already know about time, preferences and expectations. The pedagogy for achieving Level 1 FL calls for structured learning processes that help reveal to people their existing assumptions.

By way of contrast, discovery of the unknown (knowledge that is neither tacit nor explicit but must be discovered) and the application of the methods that combine creativity and social science to the task of discovery usually do not involve the same methods nor competences as Level 1 FL. Certainly, both Level 1 and 2 FL entail learning processes and, as such, share the principles of good pedagogy. However in practice the discursive exposure of underlying expectations and values that works so well for developing Level 1 FL is in marked contrast to the much more reflective (often non-discursive) processes of analytical refinement involved in Level 2 FL.

Without making any implausible claims to value and expectation neutrality, since our basic interpretive structures are built upon values and expectations, the methods used by Level 2 FL aim to provide the most open set of parameters or permutations for telling stories about the future (i.e. increasing the coverage of Zone 5). The challenge is to develop a “space” for imagining possible futures. Like with any canvas or map there are rules and conventions, rooted in specific “ways” of seeing the world, that circumscribe what is painted or projected into the frame. It is important to make these explicit, as per point five (v) of the narrative rules outlined above. However it is equally important to justify, using theories and hypotheses, the selection of the variables that describe the “possibility space.” Thinking of geography it is obvious how even if the territory on a map is unexplored it can still be located, the coordinates of the unknown, using latitude and longitude. In a similar fashion a possibility space delimits, by setting out a coordinate zone, a range of possible “locations” for stories of the future.

Here is a practical example of a four step method for constructing a “possibility space” frame and then selecting specific points (scenarios) within that frame.[22] This example is meant to illustrate the concept without in any way exhausting the many techniques for defining a possibility space. For the sake of brevity there is no detailed discussion of the premises and theory underlying the specific variables.

Step 1. First select a subject,[23] in this case for the purposes of illustration the pervasiveness of a specific technology like electricity (variable A) – with pervasiveness defined in terms of how widely a technology is diffused (extent and diversity of use). When a technology is invented and then commercialised its degree of eventual diffusion remains an open question. On the one hand it is possible that it will not be picked up at all. Alternatively it might become widely diffused, entering all aspects of life – from the workplace to the home.

Ease

of use

Simple

Difficult

Limited &

homogeneous

Unlimited &

heterogeneous

Range

of uses

Electricity

Figure 2: Possibility Space Illustration – Pervasiveness of Electricity

Step 2. The second step is to develop a model, following the general precepts of social science, that defines the attributes (variables) of the selected subject, in this case the diffusion of electricity. Here, by way of example, the hypothesis is that the possibility of diffusion (note this is not probability of diffusion) increases in line with two of the key attributes of a technology: (a) how easy it is to use and (b) the number of different uses to which it can be put [27]. Figure 2 describes the two variable space and the arrow illustrates one hypothesised relationship (path) for a technology like electricity. As it becomes easier to use and is applied to more different uses, the possibility of pervasiveness increases from the lower left corner of the possibility space to the upper right.

Crucially the probability of arriving at any particular point in this space remains undetermined.  No probabilities are attached to any particular point. As a result possibility spaces help to open up the field of points (outcome scenarios) and trajectories for getting to the outcomes (path scenarios). It is one way of being systematic and explicit about the hypothetical “what if”. Forecasters also take this type of approach, only their efforts, including the theories and variables selected, focus on prediction and usually work within the constraints of what is practically quantifiable.[24] The aim of a possibility space analysis is to get beyond the predictive imperative by applying many of the same modelling techniques to the challenge of expanding the set of possibilities for both building and mapping scenarios.

This approach helps to overcome two drawbacks that often render scenarios less useful for strategic purposes. First this technique is careful to focus on Zone 1 (possibility) prior to working on Zones 2 (probability) and 3 (desirability). This is critically important from a strategic point of view because it provides a broader canvas (Zone 5) for imagining both the ends (goals for the future – Zone 3) and means (how to get there – Zone 2). This facilitates efforts to articulate the emergent nature of the future. Second, a key tool for pushing the imagination is the use of the analytical apparatus and findings of social science. Not only do these hypotheses and models often (although not always) make it easier push the envelope of what is imaginable, even to the point that some possibility space scenarios range outside Zone 1, but in addition, because the definitions and frameworks abide by basic rules of social science, it is easier to connect an analysis of the conditions for the realisation of particular scenarios (analysis of the probability that certain choices will be made) to actionable and effective decisions (policies).

Step 3. Having enlarged the set of possible futures for consideration when developing scenarios, the third step is to select specific scenarios from what is still a vast space of possibilities. The question is how?  The criteria for selecting the scenarios will depend, in large part, on the type of story (as per the narrative rules above) and the specific subject. Of course there are the Bear and GBU approaches that could be applied immediately to the broader set of possibilities. These extrapolation and/or preference-based perspectives can be used to make a selection from within the larger possibility space, either by taking the starting point and rates of change as givens or by imposing a specific set of values for differentiating end-points. However, once again, the use of values and predictive parameters runs the risk of pre-empting imaginative and emergent options.

A more neutral and open ended way of selecting scenarios from within the “possibility space” frame is to focus on the functions and/or organisational attributes of the scenarios’ subject setting aside, for the moment and once again, the questions of probability and desirability. This approach to selecting scenarios, which is just one amongst many,[25] has the particular advantage of translating fairly directly into the functionalist/organisational requirements of much decision-making, i.e. decisions that aim to achieve a certain functional outcome (a specific product or end result such as an “educated” population) using a particular form or organisation (like a factory or a school).

Continuing with the illustration, imagine that electricity is a technology that has not yet traced its path across time (in formal terms a “counter-factual” scenario). From this pre-electrical vantage point imagine three hypothetical functions and two basic organisational patterns that can be used to develop the counter-factual scenarios for electricity: weapon (tool of war); local replacement for steam and water power in factories; and autonomous source of power for all kinds of consumer products (most of which have not been invented yet). Out of many possible ways of organising electrical power generation the two most contrasted options are: centralised and decentralised. These three functions and two organisational options generate the six scenarios in Table 3.

 Table 3: Form and Function Scenarios of Electricity Use Scenarios

Step 4. The next step is to take the six scenarios in Table 3 and project them back on to the possibility space as per the mapping in Figure 3. Obviously the location of a particular scenario within the possibility space is determined by how that scenario relates to variables (a) and (b). Again for the sake of illustration and without the in-depth analysis needed for defining metrics and justifying the situating of the variables, scenarios S2, S4 and S6 are mapped higher on the grounds that decentralised generation implies that technical barriers to use have been reduced – hence these scenarios score higher on ease-of-use. While scenarios S4, S5 and S6 are deemed to exhibit a wider range of uses because as a decentralised tool for industry (S4) and a general tool for consumers (S5, S6), electricity is assumed to be used in many different ways. In S1, where electricity is held exclusively by the military as a specialised weapon dependent on the centralised generation of power there is little need to develop ease-of-use, while the range of uses is very narrow. Hence S1 is in the lower left of the possibility space.  

Ease

of use

Simple

Difficult

Limited &

homogeneous

Unlimited &

heterogeneous

Range

of uses

Electricity

S2

S3

S6

S1

S4

S5

Figure 3: Example of a Function/Form Scenario for Electricity

None of these scenarios can be considered strategic (informative for making choices) since they only map function/form scenarios onto a technology pervasiveness possibility space. Of course we know that electricity ended up covering all three functions and, despite the recent appearance of somewhat simpler and more efficient techniques for decentralised power generation, the ease-of-use (and cost) problem was largely solved through centralised provision of electric current. Getting to strategic scenarios, as illustrated in Figure 1, involves reconnecting with probabilities and preferences or Level 3 FL.

2.3  Level 3 Futures Literacy: Choice

Level 3 FL, building on the rigorously imagined scenarios of Level 2 FL, uses values and expectations to assess today’s choices. Level 3 FL integrates the insights of the two previous levels. From Level 1 it takes an awareness of values and expectations. From Level 2 FL it takes the discoveries of “rigorous imagining”. The Level 1 values are crucial for selecting from amongst different scenarios in order to designate one (or an isovalue set) as the strategic goal(s). Expectations are also a key ingredient for Level 3 FL because the likelihood of making the decisions that matter for the strategic goal will depend, in large part, on the expectational landscape. Level 2 imagination is crucial because it takes the assessment of the present beyond the constraints of what is considered “realistic” (for instance the “I don’t dare hope for” or “it can’t be measured” that so often truncates Level 1 thinking). Rigor is also critical because the analytically sound attributes of Level 2 scenarios means that the conditions for change are presented in decision relevant terms. Level 3 FL builds on all of this capacity to think about the potential of the present and makes the link to action.

Once again an illustration may help, continuing with the counter factual scenarios for electricity, assume that the desired goal (the preferred scenario), as articulated for instance by elected politicians, is to make electricity as pervasive as possible (a bit like information technology today). This means that the policy challenge calls for an analysis of what it would take to get to the upper-right of the possibility space (Figure 2). Decision makers would seek answers to a series of discipline spanning questions like:  What are the best technical estimates that this is plausible?  How practical are decentralised forms of electrical generation like, at that time, mini-water wheels, batteries, wind, solar? What are the obstacles, in terms of people’s know-how, risk factors, cost, etc.? Can ease-of-use be expected to improve dramatically without generating much higher costs (training, new user-interfaces, etc.) How serious are the entrenched interests blocking certain decisions and facilitating others?

Looking back we now know from the history of electricity that some of the decisions that locked in the form and function (the history) of electricity’s development were explicit, but that many were not [27]. At a certain point the path dependency, or the cost of changing course (technically, financially, and from the point of view of overcoming political obstacles) became very high [36]. Most of the decisive moves that led to the lock-in of the centralised model were taken without any real awareness of the (what-if) roads not taken. Historians now debate whether at the time there was either the will or the way to explore the technological, economic and social possibilities – scenarios of plausible alternative configurations [37]. Indeed were the terms (language), workplace tools and methods, and consumer goods and living patterns of ubiquitous electricity even imaginable? Strikingly, much the same question can be posed today regarding the future of energy or information technology or of the tertiary education sector.

To go into a little more practical detail, Level 3 FL involves four basic (and familiar) steps:

i) Strategic Goal. First the possibility space scenarios that are part of Level 2 FL need to be looked at in normative terms. What is good or bad about these rigorously imagined scenarios. Which one is preferred (or which ones when the task is to select a set of isovalue scenarios). Here it is very useful to be able to reintroduce the values made explicit during the process of developing the Level 1 scenarios (note it is the specification of values that is useful not the Level 1 scenarios which have, by this stage, been discarded).

ii) Strategic Choices. The second step analyses the similarities and differences between the selected story of the future and the story we tell about the present. The aim is to identify the choices that are both “structurally” and “dynamically” critical (i.e. without this particular change or continuity the structure and/or the dynamics do not hold together (work, function)).[26] How consistent or inconsistent are the choices that underlie the strategic scenario with current decisions?

iii) Strategic Probability. The third step, in light of the second, is to consider the probability that the requisite decisions will be made. Given current expectations (political positions, social conventions, attitudes, etc.) is it likely that the enabling changes will occur? What do decision making models, historical analyses and political science tell about the play of power and conflict, inertia and volition when it comes to the strategic choices?

iv) Finally, the fourth step is to act on the basis of the strategic insights acquired in the previous steps. Such an expectation may be too rationalistic and in any case will depend greatly on the context. Still, the fundamental point of FL and of the effort required to identify scenarios that are at once possible, probable and desirable, is to influence the choices people make. This last step is not at all self-evident. For futurists, but also those who yearn for prediction, it is worth remembering the recent and tragic example of predictions not heeded: the inundation of the city of New Orleans. Predictive or non-predictive, a scenario is only a decision making tool. Tools are not always used, nor are they always used for the ends the designers intended.

The role of FL and the HSS method for acquiring and practicing FL is an open one. By way of a brief conclusion I want to explore two opposing uses: planning versus spontaneity.

3. Conclusion “The conception of education as a social process and function has no definite meaning until we define the kind of society we have in mind.”

              John Dewey, Democracy and Education, 1915 [51] 

Level 3 FL makes it easier to put current choices into strategic context in two ways. First, more conventionally, it helps decision makers to question current goals in explicit, actionable terms by drawing clear contrasts between the assumptions and content of present policy and the outcomes and preconditions of the strategic scenarios. Are the goals the same?  Are the functional relationships that make the scenarios plausible the same?  What has changed to turn the potential into an operational, more practically realisable vision?  How dependent (contingent) are the changes depicted in the strategic scenarios on changes in current policy and/or on new policy and/or on changes in underlying values, individual and collective capacities, institutions, laws, cultural norms, etc.? 

Second, less contiguous with past practices, Level 3 FL may be one of the necessary (although not sufficient) conditions for the emergence of a post-industrial society. In such a society, if FL becomes quasi-universal like linguistic literacy today, then that society may be capable of continuously exploring both the micro (individual) and macro (collective) requirements for realising the potential of the present. This is more fundamental than it might appear because building up this capacity may address one of the primary unresolved problems with post-hierarchical governance (i.e. fuller democracy): how to reconcile bottom-up local desires and power with top-down preferences and power? Or, framed in terms of complex adaptive systems theory, if FL is an attribute of post-industrial society then it may help to improve the chances that the path taken by “blind” evolutionary processes is both consistent with people’s aspirations and makes the most of the potential for innovation [50]. 

In this way the strategic scenarios that are central to Level 3 FL offer a solution to two critical “flaws” of non-hierarchical, complex self-organising systems: the non-teleological nature of such evolution and the non-systematic discovery of potential for innovation or alternative solutions. Level 3 FL responds to these two deficiencies of evolutionary systems by providing a society-wide (ambient) capacity, in the context of the assumed post-industrial society, to detect and engage the potential of the present in four critical areas. First is the dynamics (birth, death, entry, exit) and complexity (multi-level, multi-purpose) of networking. Equipping people and communities (interest, practice) to ascertain, initiate and sustain the requisite common standards for network functioning. Second is the explicitness of values in practice that can turn spontaneous choices into the realisation of our aspirations (without actually being able to know, project or want to impose such aspirations on future generations). Third is the contingency of outcomes (probability) on choices at the level of individual and collective behaviour, rules, etc. And fourth through a systematic examination of the opportunities and need for innovation – where breakthroughs are waiting to happen and where breakthroughs need to happen.

Level 3 FL shares many attributes of “strategic planning” without, however, becoming a quest for predictive fulfilment. This may seem a small nuance and, if the opening assumption regarding transition scale change turns out to be false, a largely irrelevant one. But if FL continues to be seen in the context of a post-industrial “spontaneous society” then Level 3 FL must remain resolutely non-predictive as well as strictly choice contingent.[27] The non-predictive requirement stems from four basic attributes of a spontaneous learning intensive society: i) heterogeneity, ii) complexity, ii) fluidity, and iv) spontaneity [17,18]. Under these conditions prediction is a violation of principle. There is no way nor is there any need to know what will happen (what people will want, do, have or think in the future).[28] What does matter is a recognition of how choice contingent the future is and hence why decisions made now need to grasp, to the fullest, the possibilities of the present (i.e. be futures literate).

Strictly speaking there is no reason to violate the principle of “non-prediction” when making choices meant to change what happens in the future. Everyone accepts the self-evident proposition that the future is the outcome of a series of decisions that were always made in the present, the here and now. The past-time of human society stretches out behind us in a long string of decisions (made or not made) and the future-time in a string of decisions to be made. Following the logic of this “modal” assumption there is no necessary reason to impose predictive expectations on the choices made today since such choices may or may not be followed by consistent or inconsistent choices tomorrow. Furthermore if a choice is made tomorrow that is inconsistent or in outright opposition to the one made today, that is (according to one set of values) the prerogative of those making the choices in the future.

A slight pause on this point is useful since it runs contrary to one of today’s “common sense” principles enunciated by many and perhaps most familiar from Gro Harlem Brundtland’s 1987 Report: Our Common Future, that defined sustainable development “as development that meets the needs of the present without compromising the ability of future generations to meet their own needs.” Inherent in this principle is the belief that today’s generations are in a position to determine what will or will not compromise the ability of future generations to meet their own needs. In one interpretation, a common one in the literature, this rests on a very strong predictive assumption regarding not only the capacities but the values of future generations. An assumption that is usually left implicit.

Thinking about the relationship of the values past to values present, there are few that would say that my choices today should be based on the assumptions made by my grandparents about a women’s role in society, about the filial duties of the son, about the way to ensure a livelihood, about the importance of the nation, about authority or democracy. Yet my grandparents may have made choices, not only to ensure their survival, but perhaps in their minds to perpetuate their way of seeing the world and living in it. As I sit here writing I cannot begrudge them either their conservatism (preservationism) nor their choices about how to survive in a fairly hostile context (early to mid 20th century). What I can do is seek a way to continue their commitment to freedom and responsibility by improving both my capacity and that of the society upon which I depend to put my values into practice, into day to day life, into the string of choices that turn the present into the future. It is also one of the profound criticisms of today’s world, one of the value based reasons for striving to change it, that it is very difficult to reconcile many of our aspirations with both the choices we face and the terrible reality of our times. Advocating change is a moral imperative.

Which is precisely why FL as a capacity may be useful. FL embraces the profound unknowability and contingency of the future by developing the capacity to imagine and analyse the potential of the present in ways that are more consistent with our present values. FL is not about predicting what will happen nor planning the path from A to B but about improving the capacity to imagine and assess the potential of decisions made now to create a future by putting values into practice. By combining an openness to the potential of the future with a greater ability to invent stories that make sense of the present, FL produces strategic insight without prejudicing the autonomy of people in the future to see different options and hold different values. Still FL is only a tool. And like any tool it can be used for “good” or for “bad”. Nor is the tool an end in itself.

Why then does FL matter? There is no way to know. What is clear, looking to the past, is that the implications of generalised language literacy ushered in by universal compulsory schooling, a hugely transformative initiative by any standard, made possible things that no one could have imagined. Champions of schooling argued that formal education was the only way to transmit and reproduce the workings of a complex society [51].   Is futures literacy likely to be as critical a tool for the 21st century as language literacy was for the 19th and 20th centuries? Could futures literacy be a key enabler of a much more spontaneous, networked and learning intensive society? Is this a way to reconcile the power of complex adaptive systems with the aspirations of freedom and responsibility? The only way to find out is to do it.

Acknowledgements

As with all such works, this paper has benefited greatly from the insights provided by others. While exonerating them from any errors, I owe deep thanks to Madeleine Akrich, Dan Atkins, Tom Bentley, Peter Bishop, Kristine Bruland, Hugues de Jouvenel, Laurent Dominati, William Dutton, Pankaj Ghemawat, Michel Godet, Fabienne Goux-Baudiment, David Hargreaves, David Hopkins, Esko Kilpi, Pierre Levy, Michael Mandel, Tony Mackay, Nicola Meek, Wolfgang Michalski, Alain Michel, Isabelle Miller, Morris Miller, Geoff Mulgan, Philip Van Notten, Jay Ogilvy, Erik Overland, Gilles Paquet, Helene von Reibnitz, Jean-Claude Ruano-Borbolan, Ziauddin Sardar, Wendy Schultz, Hanne Shapiro, Richard Slaughter, Mihai Spiriosu, Barrie Stevens, Michael Storper, Edna Yahil as well as two anonymous referees. Lastly, I want to thank my students at the Masters in Public Affairs at Sciences Po for their perception and perseverance.

References

[1]   P.D. Aligica, Prediction, explanation and the epistemology of future studies, Futures 36:10 (2003) 1027-1040

[2]   R.A. Slaughter, Where now for futures studies?, Futures 34:3-4 (2002) 229-233.

[3]   J. Dator, Advancing Futures: Future Studies in Higher Education, Praeger, Westport, CT, 2002.

[4]   W. Bell, Foundations of Futures Studies: Human Science for a New Era, Vol. 1: History, Purposes, and Knowledge, Transaction Publishers, Piscataway, NJ, 1997.

[5]   M. Godet, Creating Futures: Scenario Planning as a Strategic Management Tool, Economica, PLACE, 2001. And M. Godet, Manuel de prospective strategique, Tome 1 and 2, Dunod, 2001, 2004

[6]   K. van der Heijden et. al., The Sixth Sense: Accelerating Organisational Learning with Scenarios, John Wiley & Sons, 2002

[7]   G. Ringland, Scenarios in Public Policy, John Wiley & Sons, 2002.

[8]   J.A. Ogilvy, Creating Better Futures: Scenario Planning as a Tool for a Better Tomorrow, Oxford University Press, Oxford, 2002.

[9]   P.W.F van Notten, J. Rotmans, M.B.A van Asselt,  and Dale S. Rothman, An updated scenario typology, Futures 35:5 (2003) 423-444.

[10]               R. Miller, T. Bentley, Unique Creation, National College for School Leadership, UK, 2003.

[11]               D.C. North, “Dealing with the non-ergodic world: Institutional economics, property rights and the global environment,”  Duke Environmental Law and Policy Forum. X: 1 (Fall 1999) 1-12. (Available at: http://www.law.duke.edu/shell/cite.pl?10+Duke+Envtl.+L.+&+Pol'y+F.+1)

[12]               Presidency Conclusions, Lisbon European Council, 23-24 March, 2000. (Available at: http://ue.eu.int/Newsroom/LoadDoc.asp?BID=76&DID=60917&from=&LANG=1)

[13]               OECD International Futures Programme’s 21st Century Transition series. (Available at:  http://www.oecd.org/dataoecd/12/42/1903212.pdf)

[14]               B. Lundvall, B. Johnson, The Learning Economy, Journal of Industrial Studies, 1:2 (1994) 23-42.

[15]               B. Lundvall, The Social Dimension of the Learning Economy, DRUID Working, Paper 96-1, University of Aalborg  (1996). Also see Appendix 1 for more detail about these variables.

[16]               R. Miller, The Future of Public Sector Reform: Towards a Learning Society? in: Adaptive Public Services: The Next Phase of Transformation?,  DEMOS & Hewlett Packard, London, December 12, 2003.

[17]               R. Miller, Getting the Questions Right: Challenges for 21st Century Policy Makers, Optimum Online, The Journal of Public Sector Management 33:3 (Fall 2003). (Available at: http://www.optimumonline.ca/article.phtml?id=183)

[18]               R. Miller, Where Schools Might Fit in a Future Learning Society, IARTV, Victoria, Australia, November 2003.

[19]               E.A. Jackson,  “A First Look at the Second Metamorphosis of Science”, Santa Fe Institute Working Paper, 95-01-001, 1995. (Available at: http://www.santafe.edu/sfi/publications/Working-Papers/95-01-001.pdf)  

[20]               M. Buchanan, “Complexity Tools in Evaluation of Policy Options for a Networked Knowledge Society”, DG Information Society, European Commission, June 2003.

[21]               J.S. Brown, P. Duguid, Universities in the Digital Age, in: B. L. Hawkins, P. Batin (Eds.), The Mirage of Continuity: Reconfiguring Academic Information Resources in the 21st Century, Council on Library Resources, Washington D.C., USA, 1998.

[22]               R. Miller, Measuring What People Know: Human Capital Accounting for the Knowledge Economy, OECD, 1996. (Available at: http://www1.oecd.org/publications/e-book/9196031E.PDF)

[23]               D.E. Atkins, Revolutionizing Science and Engineering through Cyberinfrastructure: Report of the National Science Foundation Advisory Panel on Cyberinfrastructure, National Science Foundation, February 3, 2003. (Available at: http://www.communitytechnology.org/nsf_ci_report/)

[24]               K. van der Heijden, Scenarios: The Art of the Strategic Conversation, Wiley, 2005

[25]                The Millennium Project has produced a scenario bibliography with over 450 scenario sets available at http://www.acunu.org/millennium/information.html which are organized in the “Futures Matrix” in the following domains: Demographics and Human Resources; Environmental Change and Biodiversity; Technological Capacity; Governance and Conflict; International Economics and Wealth; Integration or Whole Futures.

[26]               R. Slaughter, Knowledge Base of Futures Studies, 2005, in particular Volume 2, including W. Schultz, “Defining Futures Fluency”

[27]               K. Bruland, “Technological Revolutions, innovation systems and convergence from a historical perspective”, University of Oslo, Convergence Project, February, 2001; C. Perez, Technological Revolutions and Financial Capital, Edward Elgar, 2002

[28]               M. Spiriosu, Global Intelligence and Human Development: Towards an Ecology of Global Learning, The MIT Press, 2004

[29]               J-C Ruano-Borbalan, “Les metamorphoses du pouvoir", Le Pouvoir, Editions Sciences Humaines, Auxerre, 2002

[30]               National College for School Leadership, FutureSight, NCSL, 2004

[31]               See Innovation UK http://www.innovation.gov.uk/

[32]               S. Vincent-Lancrin, University Scenarios, OECD, 2003

[33]               D. Istance, Schooling Scenarios, OECD, 2000

[34]               National Education Association, The Future of Higher Education, http://www2.nea.org/he/future/index.html

[35]               A. Wynberg, Wider and Deeper: A Review and Critique of Science and Technology Foresight Exercises in the 1990s, AFI Monograph Series, 2003 http://www.swin.edu.au/afi/publications/monographs/index.htm

[36]               C. Perez, Technological Revolutions and Financial Capital : The Dynamics of Bubbles and Golden Ages, Edward Elgar, 2003

[37]               D. Gugerli, “The Sociocultural Aspects of Technological Change: The Rise of the Swiss Electricity Supply Economy, Science in Context, 8, 3, 1995

[38]               Schooling for Tomorrow proceedings of the 2nd OECD Forum, Toronto, OECD, 2004 http://www.oecd.org/document/60/0,2340,en_2649_34859095_31532732_1_1_1_1,00.html

[39]               R. Slaughter, Futures Beyond Dystopia: Creating Social Foresight, Falmer Press, 2004.

[40]               J. C. Glenn and the Futures International Group, Scenarios, in Futures Research Methodology, AC/UNU Millennium Project, Version 2.0, http://www.acunu.org/millennium/FRM-v2.html

[41]               C. Booth, C., Rowlinson, M., Clark, P., Delahaye, A. & Procter, S. (2005) “Scenarios and Counterfactuals as Modal Narratives”. Paper to the 3rd Conference on Advancing Foresight Methodologies, Roskilde, Denmark, 12th-14th May

[42]               D. Lewis, Counterfactuals, Blackwell, 1973

[43]               K. Popper, The Poverty of Historicism, Routledge, 2002.

[44]               N. Rescher, Imgining Irreality: A Study of Unreal Possibilities, Carus Publishing, 2003.

[45]               Cloudy Crystal Balls: An assessment of recent European and global scenario studies and models, European Environment Agency, November 2000, http://www.eea.eu.int.  P. Van Notten, Writing on the Wall: Scenario Development in Times of Discontinuity, dissertation.com, 2005

[46]               Kennedy, A Guide to Econometrics, Blackwell Publishing, Fifth Edition, 2003

[47]               Slaughter, Richard, The Knowledge Base of Future Studies, Professional Edition 2005, http://www.foresightinternational.com.au/books/kbfsproedn.php

[48]               U. von Reibnitz, Scenario Techniques, McGraw Hill, 1988

[49]               H. de Jouvenel, An Invitation to Foresight, Futuribles, 2004

[50]               E. Kilpi, Complexity and Emergence, personal communication of the author; N. Johnson, The Science of Diversity in a Fast Changing World, http://www.lanl.gov/orgs/dvo/sci_diversity/overview.shtml 

[51]               J. Dewey, Democracy and Education, Macmillan Company, 1916

[52]               K. Boulding, "Basic Evolutionary Model," Evolutionary Economics, Sage, 1981, p. 44.

[53]               I. Prigogine, The End of Certainty: Time, Chaos and the New Laws of Nature, Free Press, 1997