![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
Abstract
The modern world is characterized by problems that involve systems with social and physical subsystems. They are entangled systems of system of systems with multilevel dynamics. There is no methodology able to combine the partial micro-, meso- and macrotheories that focus on subsystems into a coherent representation of the dynamics of the whole. Policy requires prediction, but the traditional definitions of prediction are not appropriate for multilevel socio-complex systems. Heterogeneous multilevel systems have subsystems that may behave with great regularity over long periods of time, and then suddenly change their behavior due to weak coupling with other subsystems. Thus systems that are usually highly predictable may be subject to rare but extreme events, and this is highly relevant to policy-makers. New ways of thinking are needed that transcend the confines of the traditional humanities, social and physical sciences. Of necessity, this science will be embedded in the design, implementation and management of systems, and therefore the new science will be entwined with policy. Much policy is interventionist experiment. By themselves scientists cannot conduct experiments on socio-complex systems because they have neither the mandate nor the money to design and instrument experiments on the large scale. Policy-makers – elected politicians and their officers – design the future, making it as they believe it ought to be. New kinds of scientific predictions can inform policy but can only be instrumented and tested if there is goodwill between policy-makers and scientists, where scientists are junior partners. Scientists offer policy-makers theories and predictions of social systems based on logical-deductive methods. Policy is generally made on the basis of rhetoric, with the best possible arguments being deployed to support favored conclusions. To convince policy-makers that a particular scientific theory should be used, scientists move from the logical-deductive to the rhetorical. Thus the full theory of a science of complex systems has to provide a logical-deductive metatheory of the rhetorical and logical-deductive systems that make decisions and implement them. Traditional natural and physical science has avoided rhetoric, which is much better understood in the humanities and social sciences. Thus it is concluded that the science of complex systems must embrace the humanities and social sciences not just because their domains of study are relevant but also because their methods are necessary to understand how science and policy work together in complex social systems.
Acknowledgement
I am very grateful to my colleague Matthew Cook for reading and patiently commenting on earlier versions of this article, enabling me to correct some obvious conceptual errors.Footnote17
Notes
1. It could be argued that Gidden's theory of structuration is a counter-example to this proposition (Giddens Citation1986). Certainly this theory is a good point of departure for a new formalism, but the use of vernacular language to express the theory is not considered here an adequate formalism. In particular, from the perspective of complexity science there is the question of how such a theory could be made operational to give “predictions” in particular cases based on data. The challenge is for complex systems science to provide a more general formalism that social scientists would agree has lost nothing of the original.
2. In his paper “From complexity to perplexity”, Horgan (Citation1995) quotes 31 definitions of complexity given by Seth CitationLloyd in 2001. Horgan illustrates the diversity by a selection, including entropy (disorder), information (surprise), fractal dimension, effective complexity (regularity vs. randomness), hierarchical complexity, grammatical complexity, thermodynamic depth, time computational complexity, spatial complexity and mutual information (between parts). In his 1999 PhD thesis, CitationEdmonds gives over 40 definitions of measures of complexity.
3. It could be said that the term is contested and therefore a social construct, fitted within an interpretive epistemology that provides one of many possible ways to understand the world with all possessing some degree of truth. However there is much more agreement on the meaning of the bulleted list of more detailed characteristics of complex systems. Furthermore, the existence of many interpretations suggests the possibility of knowledge and science at a metalevel above the particular interpretations, as discussed later.
4. In his book The sciences of the artificial Herbert Simon (1965) makes a distinction between artificial and natural systems. Artificial systems are man-made and include everything that is designed and intentionally created by human beings. Here the term “natural systems” is intended to mean physical systems that are not designed and man-made, such as the weather. There are of course gray areas such climate change, which many believe is artificial but created by accident rather than design.
5. For examples of computer simulations being used to solve practical problems see the article “Practical business applications of complex science” in the ASSYSTComplexity Newsletter, Number 2, September 15 2009. Available from: http://www.rzevski.net/09%209%20Assyst_Newsletter_final_04_JHJ.pdf
6. The use of the terms “micro”, “meso” and “macro” here follow the usual ill-defined usage. In an obvious sense individual people could be said to exist at the microlevel, and nations could be said to exist at the macrolevel, with anything in between said to exist at the mesolevel(s). The terms will be made more precise below.
7. By definition, the interventions of policy deliberately attempt to change what exists and create artificial systems.
8. It will be argued that the process of formulating and executing policy is exactly analogous to design, with the possible exception of learning from failure.
9. For the purpose of exposition the roles of policy-makers and scientists are separated. Policy-makers – elected politicians supported by their apolitical officers – are mandated by the electorate and have the moral authority to change society. Scientists do not. This is a clear distinction. Of course the reality is that many officers and scientists are political (possibly abusing their favored positions) and the nitty-gritty of policy is much more messy than suggested here.
10. See Russell (Citation1952).
11. Clair O'Farrell writes: “A scientific practice, in Foucault's account, is a particular set of codified relations between a precisely constructed knower and a precisely constructed object, with strict rules which govern the formation of concepts. Foucault was interested in science for a number of reasons. One of these was that “science” had set itself up as the ultimate form of rational thought. With the Enlightenment, scientific reason became the privileged way of accessing truth. According to this view, for knowledge to acquire value as “truth”, it had to constantly strive to become “scientific”, to construct and organize concepts according to certain rigorous criteria of scientificity. Foucault argues that scientific knowledge is not inherently “superior” or more “true” than other forms of knowledge” (http://www.michel-foucault.com/concepts/index.html 2007). The argument given here does not imply that science is the only form of knowledge. Even within science as defined here there are alternative descriptions of systems. However it will be later suggested that all descriptions and forms of knowledge may be unified at an appropriate metalevel.
12. http://plato.stanford.edu/entries/aristotle-rhetoric/ [Accessed 19 March 2010].
13. “Evolution, cities and planning”, plenary talk by Michael Batty, European Conference on Complex Systems, ECCS'09, University of Warwick, September 2009. Available from: http://www.assystcomplexity.eu/video.jsp?video=55.
14. It could be argued that policy failures are inevitable because we may never understand the future due to the nature of prediction as performative or constituitive. This may be so, but need not be conceded until the idea of metelevel resolution as proposed in this article has been reufuted.
16. Mark Lythgo, The new two cultures, BBC Radio 4, 25 April 2007. Available from: http://www.bbc.co.uk/radio4/science/thenewtwocultures.shtml.
17. See also Cook (Citation2010).