Ambiguity as a manifestation of complexity

Building on the uncertainty literature, we* conceptualize ambiguity as the type of uncertainty that emerges from complexity.

Knight¹ first introduced the distinction between risk and uncertainty in economics, defining risk as a situation in which the possible outcomes are known and the probabilities associated with each outcome can be calculated. An example of risk is casino betting². Uncertainty is defined as a situation in which the possible outcomes are known but the associated probabilities cannot be calculated. Uncertainty can be associated with hurricanes in the Mexican Gulf, which are a recurrent event (known outcome) whose probability is unknown. Uncertainty can be associated with disruptive events such as tsunamis³, which are known to occur, but cannot be precisely predicted. The novelty introduced by Knight is that uncertainty is conceptualized as a matter of degrees of (lack of) knowledge.

Building on the idea of degrees of uncertainty, Wynne⁴ identifies four levels of uncertainty: (i) risk, in which the odds are known, (ii) uncertainty, in which “we know what we don’t know”, (iii) ignorance, in which “we don’t know what we don’t know”, and (iv) indeterminacy, in which “causal chains or networks are open”. Ignorance may be associated with geopolitical threats derived from “insecure political and unstable economic environments” ⁵ with regard to energy security, and with the emergence of new technologies, such as fracking, which introduce unforeseen possibilities and risks. Indeterminacy is associated with systemic changes, or what some authors call phase changes. “Threats like delivery disruptions or global warming of more than 2ºC can be seen as phase changes, because in addition to having a direct impact on consumers they also change the way in which the system works”¹³ [p.11]. As uncertainty increases, knowledge becomes more and more incomplete. In the context of indeterminacy, uncertainty is irreducible, as systemic changes reduce the possibility of knowing.

Stirling⁷ reorganises Wynne’s levels of uncertainty in a 2×2 matrix, which ranges from known to unknown outcomes and known and unknown probabilities, and introduces ambiguity as an additional type of uncertainty. Once again, risk is defined as known outcomes and known probabilities, uncertainty as known outcomes and unknown probabilities, and ignorance as unknown outcomes and unknown probabilities. According to Stirling, ambiguity is defined as a situation in which probabilities are known but outcomes are unknown, because of divergent and contested perspectives on the justification, severity or wider meanings associated with a perceived threat⁸. The concept of ambiguity thus suggests that uncertainty may be mobilised to pursue different goals, not because of a lack of information, but because a plurality of representations can be accommodated.

The analysis of uncertainty is particularly relevant to understand the interface between science and policy. Science and technology play an important role both in overcoming and identifying knowledge gaps, and in creating new sources of uncertainty⁹. The conceptualisation of uncertainty, and of different degrees of uncertainty has been widely used as a means to support decision-making ^10–13.

Knight’s definition of risk has been criticized in the literature for not taking into account “how people perceive uncertain phenomena and how their interpretations and responses are determined by social, political, economic and cultural contexts, and judgments”² [p. 237]. The concept of risk is thus not fully captured by a technical definition of calculable probabilities and effects. Renn et al.¹⁴ criticize the reduction of risk to calculable probabilities for this concept may lead to the use of “technocratic, decisionistic and economic models of risk assessment and management” [p. 234]. Risk management involves more than a technical characterization of probabilities, outcomes and associated harm. As the case of the mismanaged L’Aquila earthquake has shown, public concerns, the communication of uncertainty and the risk of generating panic in the population are paramount to governance in uncertainty¹⁵.

In the realm of uncertainty, the linear model of science speaking truth to power is questioned, not only because incomplete knowledge is produced in the context of irreducible uncertainty, but also because social and political threats indicate that decisions cannot be reduced to rational, utility-maximising, get-the-facts-then-act models^16,17. With regard to energy security, distinguishing between different levels of uncertainty is critical for decision-making. For example, after the Fukushima accident, experts declared that new designs for nuclear power plants took into account the risk of earthquakes¹⁸. Better designs, however, do not solve the uncertainty linked to the unpredictability of earthquakes and tsunamis. Safer designs refer to advances in the reduction of risk at the level of the power plant. Uncertainty about earthquake forecasting (higher levels of uncertainty) cannot be factored in new nuclear designs. The undistinguished reference to uncertainty with disregard to the level of uncertainty in question thus leads to a false perception of control over uncertainty, and a diffuse understanding of the role, and limits, of scientific knowledge in decision-making.

In the management of uncertainty the question of scales of analysis emerges as a fundamental factor. The existence of multiple scales of analysis invokes complexity. Complexity is defined as the existence of multiple legitimate representations that cannot be reduced to one another ^19–21. The complexity literature often refers to complex systems. A system is defined as complex if the whole is more than the sum of its parts, that is, the representation of the whole presents emergent properties that are not observed at the level of its constituent parts. The two scales of analysis, whole and parts, are not equivalent to one another.

An example of non-equivalent representations in the case of energy is that of electricity accounting. In 2005, electricity accounted for 36% of the consumption of Primary Energy Sources (PES), but only for 18% of the total consumption of Energy Carriers (EC) in Spain ²². The difference is due to the fact that the first measurement refers to the primary energy needed to produce electricity, and the second measurement refers to the end uses of energy carriers, that refer not only to electricity but also to heat and fuels. An analogy can be made with meat production and consumption: one thing is to measure the primary inputs needed to produce meat, in terms of feed for example (primary energy sources in the analogy), and a different matter is to measure meat consumption in the diet (energy carriers). Complexity signals the need for a semantic process to handle the meaning of different formal representations²³. As Shove²⁴ argues, the use of generic measures for energy “reproduces understandings of energy as an all-purpose resource” [p. 3]. The semantic distinction between PES, EC, and end uses, and between electricity, fuels and heat makes it possible to assess energy not as an abstract category, but as a set of practices and material arrangements.

Building on Stirling’s conceptualisation of ambiguity as a type of uncertainty, we* define ambiguity as the uncertainty created by the existence of multiple non-equivalent representations of the same issue. Ambiguity is thus not just a matter of different opinions, as may be the case of contrasting perceptions about the rise in the price of oil (perceived as beneficial by oil producers and problematic by oil importers). In linking ambiguity to complexity, we* argue that this type of uncertainty is caused by the existence of incommensurability in the knowledge base. That is, although a lot of information can be produced about, for instance, energy security, there is no univocal way to combine these representations, as can be seen in the proliferation of energy security indicators^5,25,26 and of academic papers dedicated to the conceptualisation of energy security^6,27.

The role of ambiguity in policy has been widely analysed in the literature^7,9,12,28,29. According to Matland²⁸, ambiguity is necessary to limit conflict (ambiguity of goals), and to define policy when there is uncertainty over the technology needed or over the role that various organisations are to play in the implementation process (ambiguity of means). With reference to the EU, Zahariadis²⁹ explains that ambiguity is an integral part of the policy making process in contexts where there is a plurality of, often contrasting, interests, a multiplicity and high turnover of actors, and highly bureaucratic systems that lead to a fragmentation of the policy process. As a consequence, we* further conceptualise ambiguity as functional uncertainty. Understanding the function of ambiguity in the policy process has important consequences for the interface between science and policy. Whereas numerous studies argue for more holistic and integrated conceptualisations of energy security^3,25,30, Matland warns against the dysfunctional effects of clarity in policy implementation.

 *this text is taken from a paper co-authored by Louisa Jane Di Felice

References

1. Knight, F. H. Risk, uncertainty and profit. (Hart, Schaffner and Marx, 1921).
2. Taleb, N. N. The black swan. (Random House, 2007).
3. Kiriyama, E. & Kajikawa, Y. A multilayered analysis of energy security research and the energy supply process. Appl. Energy 123, 415–423 (2014).
4. Wynne, B. Uncertainty and environmental learning: reconceiving science and policy in the preventive paradigm. Glob. Environ. Chang. June, 111–127 (1992).
5. Zhang, H. Y., Ji, Q. & Fan, Y. An evaluation framework for oil import security based on the supply chain with a case study focused on China. Energy Econ. 38, 87–95 (2013).
6. Winzer, C. Conceptualizing energy security. Energy Policy 46, 36–48 (2012).
7. Stirling, A. Keep it complex. Nature 468, 1029–1031 (1993).
8. Stirling, A. in Negotiating change (eds. Berkhout, F., Leach, M. & Scoones, I.) 33–76 (Edward Elgar, 2003).
9. Funtowicz, S. O. & Ravetz, J. R. Uncertainty and quality in science for policy. Ecological Economics 6, (Kluwer Academic Publishers, 1990).
10. Refsgaard, J. C., van der Sluijs, J. P., Brown, J. & van der Keur, P. A framework for dealing with uncertainty due to model structure error. Adv. Water Resour. 29, 1586–1597 (2006).
11. Saltelli, A., Guimaraes Pereira, A., van der Sluijs, J. P. & Funtowicz, S. O. What do I make of your Latinorum? Sensitivity auditing of mathematical modelling. Int. J. Innov. Policy 9, 213–234 (2013).
12. Smith, A. & Stirling, A. Moving outside or inside? Objectification and reflexivity in the governance of socio-technical systems. J. Environ. Policy Plan. 9, 351–373 (2007).
13. Wardekker, J. A., van der Sluijs, J. P., Janssen, P. H. M., Kloprogge, P. & Petersen, A. C. Uncertainty communication in environmental assessments: views from the Dutch science-policy interface. Environ. Sci. Policy 11, 627–641 (2008).
14. Renn, O., Klinke, A. & Van Asselt, M. Coping with complexity, uncertainty and ambiguity in risk governance: A synthesis. Ambio 40, 231–246 (2011).
15. Benessia, A. & De Marchi, B. When the earth shakes… and science with it. The management and communication of uncertainty in the L’Aquila earthquake. Futures (2017).
16. Funtowicz, S. O. & Strand, R. Models of science and policy. Biosaf. First-Holistic Approaches to Risk Uncertain. Genet. Eng. Genet. Modif. Org. 263–278 (2007).
17. Pielke, R. A. When scientists politicize science: Making sense of controversy over The Skeptical Environmentalist. Environ. Sci. Policy 7, 405–417 (2004).
18. Diaz-Maurin, F. & Kovacic, Z. The unresolved controversy over nuclear power: A new approach from complexity theory. Glob. Environ. Chang. 31, 207–216 (2015).
19. Ahl, V. & Allen, T. F. H. Hierarchy theory: A vision, vocabulary, and epistemology. (Columbia University Press, 1996).
20. Kovacic, Z. Investigating science for governance through the lenses of complexity. Futures (2017). doi:10.1016/j.futures.2017.01.007
21. Zellmer, a. J., Allen, T. F. H. & Kesseboehmer, K. The nature of ecological complexity: A protocol for building the narrative. Ecol. Complex. 3, 171–182 (2006).
22. Giampietro, M. & Sorman, A. H. Are energy statistics useful for making energy scenarios? Energy 37, 5–17 (2012).
23. Allen, T. F. H. et al. Mapping degrees of complexity, complicatedness, and emergent complexity. Ecol. Complex. (2017). doi:10.1016/j.ecocom.2017.05.004
24. Shove, E. What is wrong with energy efficiency? Build. Res. Inf. 0, 1–11 (2017).
25. Sovacool, B. K. & Mukherjee, I. Conceptualizing and measuring energy security: A synthesized approach. Energy 36, 5343–5355 (2011).
26. Jun, E., Kim, W. & Chang, S. H. The analysis of security cost for different energy sources. Appl. Energy 86, 1894–1901 (2009).
27. Jansen, J. C. & Seebregts, A. J. Long-term energy services security: What is it and how can it be measured and valued? Energy Policy 38, 1654–1664 (2010).
28. Matland, R. E. Synthesizing the Implementation Literature: The Ambiguity-Conflict Model of Policy Implementation. J. Public Adm. Res. Theory 5, 145–174 (1995).
29. Zahariadis, N. Ambiguity and choice in European public policy. J. Eur. Public Policy 15, 514–530 (2008).
30. Cherp, A. & Jewell, J. The three perspectives on energy security: Intellectual history, disciplinary roots and the potential for integration. Curr. Opin. Environ. Sustain. 3, 202–212 (2011).