Scenarios as Narratives of Reactive Sequences: Lessons from Historical Social Science
Scenarios are one of the most widely used methods in futures research, and there is a wide variety of scenario methods, as documented by Bishop et al. (2007), for example. Scenarios are commonly used in practically oriented foresight work as structural descriptions of the future which help to anticipate possible future circumstances. In this context, the focus is on the impacts of trends, variables and macro-level drivers.
In an introduction to a recent special issue of Futures, Mermet, Fuller and van der Helm call for critical discussion of the theoretical bases of futures work, drawing on both work within the futures field and in other academic fields (Mermet et al., 2009). This paper contributes to this goal by discussing how concepts from historical social sciences could enrich the theoretical basis of scenario building.
In this article, scenarios are examined as tools to anticipate possible social change processes from an event-based perspective. The starting point of the paper is the parallel between futures studies and historical social science as fields which study change over time. In particular, three central concepts from historical social science are considered as important for scenario construction: 1) narrative explanation, 2) reactive sequences and 3) causal mechanisms. A tentative a scenario framework is presented to incorporate these concepts into scenario building.
Scenarios are one of the central tools of futures research. According to Bell (1997), 317, scenarios are used by all futurists in some form and scenarios are the end product of all futures research methods. Despite the wide spread of scenario methods, or perhaps because of it, there is no consensus on what scenarios are and how they ought to be created. Several recent articles have divided scenario approaches into scenario types (Bishop et al., 2007; Bradfield et al., 2005; Börjeson et al., 2006; van Notten et al., 2003). These typologies are based on different criteria. For example, Börjeson et al. (2006) focus on purposes and aims, while Bishop et al. (2007) focus on scenario techniques, and van Notten et al. (2003) ambitiously consider project goal, process design and scenario content. Bradfield et al. (2005), in turn, focus on three historical schools of scenario development.
In this paper, I will focus on what van Notten et al. (2003) call scenario content: the characteristics of the actual scenarios rather than the practical aims of the exercise, the process design or the organisational surroundings. More specifically, I am interested in the theoretical and social scientific basis of the scenario content and the claims that are made within scenarios.
Two aspects of scenarios are particularly important:
A basic distinction among scenarios is the distinction between scenarios as snapshots and scenarios as sequences or development paths. Bishop et al. (2007) call the first category end state or 'day in the life' scenarios, and the second category is titled chain scenarios or future histories. It would perhaps be more appropriate to have separate terms for these types, but both types of scenarios are well represented in futures research and both have important roles.
According to Bishop et al. (2007), the Global Business Network method of building scenarios has become dominant and it has somewhat shadowed other scenario techniques. Following this approach, scenarios are conceived as structural descriptions of a future operational environment, focusing on key variables and uncertainties (Schwartz 1996). The aim of scenarios is to impact strategic thinking today and to promote organisational learning (Van der Heijden et al., 2009). Staley (2007), 77–78 makes the case for structural descriptions most explicitly, arguing that a "scenario is meant to be synchronic rather than diachronic", describing "the rules of a game, rather than [...] the specific sequence of play". In other words, a scenario is conceived as a thick structural description of a future end-state.
Of course, the success of the GBN technique is partly due to the fact that it is an excellent, insightful technique, but partly it is a path-dependent and self-reinforcing development as new futurists are taught the GBN method or a variation of it and they may never get acquainted with alternative approaches. In my view, it is a richness rather than a hindrance that there are many different scenario approaches. The evolution of scenario building should not lead to one dominant approach but to a diversity of approaches.
In this paper, scenarios are considered as sequences of events or paths of development rather than end-points. This is because the perspectives of time, sequence and duration are important in understanding social change from a social scientific perspective (Sztompka, 1993). The notion of scenarios as sequences goes back to the roots of scenario thinking in Herman Kahn's writings on scenarios (Kahn et al., 1967). Scenario techniques within this tradition include probability trees, sociovision and divergence mapping (Bishop et al., 2007).
Concerning the second distinction between external/strategic, the arguments in this paper mostly relate to external scenarios, focusing on the developments in the societal surroundings.
Scenario approaches need to be grounded in a solid social scientific basis, and the development of scenario techniques should follow developments in the social sciences. The argument has been made many times that history and futures research are parallel fields with strong links (e.g. Kaivo-oja 2004, Staley 2007). In particular, historical social science provides a fruitful juxtaposition with scenario building. In recent decades, there have been discussions within historical social science which could have much to offer for scenario building. In the following, I will assess the meaning of three central concepts from historical social science: reactive sequences, narrative explanation and causal mechanisms. I am not claiming that these concepts have never been used in futures research, but their implications for scenario thinking have th