Chaos and complexity

Some years ago, towards the end of the Balkan conflict, I was stationed in Sarajevo with the NATO peace implementation force. As a result of the hostilities, much of the city’s infrastructure supporting its public services was totally destroyed and daily life had either ground to a halt or become utterly chaotic. In particular, driving along the city’s main highway and neighbouring streets was like manoeuvring around a giant-sized fairground dodgems track. With no electricity supply, there were no traffic lights: a large number of vehicles travelled at dangerously high speeds, their drivers negotiating not only junctions but also pot-holes, other vehicles, pedestrians, and many other obstacles. Nonetheless, seldom was there ever a serious accident or collision. Left to itself, the traffic had become a self-regulating system. So much so that when the traffic lights were eventually made to work the drivers had become so used to this self-regulating system that they appeared to have forgotten what to do. Sometimes, nearly all of the drivers ignored the lights completely: as they changed from red to green and back again they made little impact on the continuous flow of traffic. At other times, the flow of traffic simply petered to a halt, everyone unsure whether or not anyone else was observing the changes. In fact, it seemed that when everything was chaotic the traffic flowed well, but when the lights operated everything became dis-organised: its settled state was a form of organised chaos. Perhaps we might more correctly describe it as undesirably organised. The

problem was not that one system represented order and the other disorder, but that one kind of order appeared undesirable but worked and the other kind of order although desirable clearly did not. Left to their own devices, patterns emerged among the drivers that satisfied everyone’s criteria, and the resulting (dis)order appeared fair and efficient. It may not have equated to what we commonly expect in terms of a properly ordered traffic flow but it was ordered, nonetheless. The point is that our natural instinct in social life is always towards establishing some sort of pattern, or order. MacCormick’s (2007) discussion of the social practice of ‘queuing’ and its

subsequent institutionalisation deals with precisely this issue. However, the main point that I want to highlight here is not that this happens but how and

why it happens, and for what purposes. MacCormick illustrates well our common impulse to recognise, impose and institutionalise patterns but results in a way of thinking that regards order and disorder as opposites. The upshot of this is that what becomes institutionalised is what appears to be classifiable and generalisable according to institutional categories, expected and predictable by a controlling agent. Behaviour at variance with this becomes thought of as incomprehensible, unpredictable or chaotic. However, on the view being argued for here, institutionalisation and surprise are not polar opposites. Because something could not have been predicted does not necessarily imply a lack of order any more than its predictability would imply order. Which is simply another way of affirming that pattern does not exclude novelty. Indeed, far from being polar opposites, order and disorder, like universals and particulars, appear to implicate each other. To many legal theorists, law appears to fall naturally on one side of this

pairing. But if the argument being presented here holds then the dualism is a false one and law might well be less deterministic than it appears to be. In this sense, far from reinforcing the Newtonian, mechanistic world-view underlying modernist legal theory, this argument would suggest that a revision of that understanding is urgently required. The suggestion being presented here is not that our idea of reality as a unity needs to be abandoned, but that the version of that idea of unity as derived from and expressed in a predominantly Newtonian mechanistic vocabulary needs to be discarded in law as it appears to have begun to be in some other disciplines. As Tsoukas observes, diversity, change and adaptability, rather than hierarchy, rigidity and standardisation are coming much more to the fore within contemporary scholarship (Tsoukas 1998a: 293). Gradually, a new language with a new attitude towards and an appreciation of ideas such as non-linearity, disorder and noise, fragmentation, unpredictability and marginalisation is emerging. We find a change in attitude that appears much more receptive to a sense of the chaotic and an awareness of dynamic process, an outlook more in sympathy with notions of the unpredictable and the novel and much less ready to impose a division of order and disorder. But with this new outlook we need to radically rethink our ideas concerning the use of law as a tool for intervening in the world. Much has been said here about the Newtonian approach, but this now

needs to be augmented with a fuller description of what that approach entails. In the first place, it is characterised by the search for the universal, general, timeless ‘decontextualised ideal’ (Toulmin 1990: 30-36). The ontological description is that of discrete, objective units linked through normlike associations discoverable through abstract conceptual representation that can aid predictability and help to minimise elements of surprise. In this way, the subject under consideration becomes controllable. As Tsoukas argues, such a view ‘assumes an objectivist ontology, works with a mechanistic epistemology, and enacts an instrumental praxeology’ (Tsoukas 1998a: 294;

Tsoukas and Cummings 1997: 656). Such a view makes use of idealised models, created through abstraction, to estimate the complex behaviour of real entities. This assumes both that the behaviour of real entities will permit such an assessment of their various contingent factors and that by abstracting from the time-dependent historical pathways of their causal relations fairly accurate prediction is nonetheless possible. And all of this rests on that rather sweeping generalisation that we considered at length earlier; namely, that within institutionalised forms every activity of a certain type can be treated in the same way and that it is legitimate to do this. To take an example, consider Mrs Donoghue in Donoghue v. Stevenson

([1932] SC (HL) 31). In MacCormick’s account of the reasoning in Donoghue we find an argument being constructed in reverse to account for the judges’ reasoning. We are directed to understand how it is Mrs Donoghue as ultimate consumer, not Mrs Donoghue as a vulnerable old Scottish lady or any other combination of actual real-time background qualities and descriptions that is significant. But Mrs Donoghue, the ultimate consumer, does not exist in a social vacuum in the way that this abstract conceptual reconstruction of her would seem to suggest. She, Mrs Donoghue, is not this a-contextual and a-historical representation that is given of her; indeed, one is almost tempted to interject: ‘Will the real Mrs Donoghue please stand up!’ In a similar way, we might argue that in respect of the conceptual reconstruction of the manufacturer of the bottle of ginger beer containing the offending snail. What is clear from MacCormick’s account of the reasoning in that case, is

that it is the purpose for which the institution of law is intended that determines the ways in which the various purposes of the characters involved are related; that is, their relative positions within the legal institutional structure. But what this implies is that the legal institutional answer does not arise as a solution to the social problem from which it derives; rather, the institution reconstructs the problem according to its own aims and purposes and defines and modifies the limits of its relations, thus making it more malleable. However, none of this can be discovered from a simple viewing of these objects in their institutional incarnations. Such an analysis would reveal only the ‘fact’ that the system or institutional answer was created by and given in response to the environmental conditions; it would not reveal the underlying process by which one is modified or adapted by the other. So we need to ask what demands the institutional perspective is making of

the real life scenario. Why does it abstract from concrete reality sometimes this way, sometimes that, and what are the implications of this? On the basis of the analysis engaged in earlier in terms of the importance of practices, what other factors, both within and without the story, influence the decision? How is this particular set of real-life concrete relations related to the broader issues of legal structures and doctrines and social cohesiveness in which this particular scenario is set? If we choose to focus solely on the

decontextualised abstract model we will not even begin to find a way to address any of these questions. Clearly, some degree of generalisation is unavoidable. Here we notice

again the movement from what we might describe as simple data (the multitude of descriptions and items of information that could be given in respect of a particular situation) to legally relevant facts (facts that can register in the legal decision-making context) to legally significant facts (those facts that are important to the actual decision or its future authority and use; that is, as part of the ratio). In MacCormick’s view such generalisation is not only inevitable, it is essential and the very idea of legal institutions presupposes this. But the problem is that in order to see with any degree of clarity how and by what means that real-life concrete situation that is before the court can be represented in these a-temporal non-specific terms much of what makes that real-life event exactly what it is, its uniqueness, has to be dropped from view. The open-ended life narrative that gives rise to a particular episode has to be transformed into a scenario that is presentable before the court. In this process, the episode loses some of its particular features and characteristics and gains others, at least in the sense that the narrative structure imposed upon it by its institutional re-presentation embellishes it with a beginning, middle and an end. Within this structure, individual facts are relevant and important in relation to and in terms of the aims and purposes for which the court is constituted. Thus, it is not difficult to see how, in this context, generalisation and abstraction, this consequent reduction to role and rule, might be seen as an obstacle to a fuller understanding of the complexities of real-life situations, rather than an aid. Thinking about a situation in a legal context provides the thinker with a way of thinking that structures how that situation is thought about. The narrative structure imposed by law corresponds with the a-contextual, a-historical mode presupposed by law but not with the time-bound, context-specific situation experienced by those involved. In this way, the mode of thinking allows the thinker to construct certain expectations but it limits them to a certain type: those that can be expressed as universalistic expectations. Such a privileging of the a-temporal, a-historical, and generalisable comes at a price; namely, the loss from view of the temporal, the contextual and the historical. Accordingly, theorising about law that finds its roots in a post-Enlightenmentinspired mechanistic model of the universe generally assumes that the actual situations which are represented by legal institutions as institutional phenomena operate in a social vacuum. The narrative structure that they are assumed to have is in fact a narrative structure imposed by law, and this is the only narrative structure they are permitted to have; without this, as we have seen in relation to Re A, they cannot be heard in law. Legal decision-making illustrates very clearly how this rational translation

of multi-faceted, open-ended, real-life phenomena into the data appropriate to legal speech and then into legal decisions and thereafter into legal

justifications that validate, corroborate and legitimate the institutional mechanism and structure, takes place. But we are only beginning to understand how such decision-making is decidedly self-referential and why this is a problem. Judges must justify or ground their decisions in law. But, as we have seen, at some point the infinite regression brought about by thinking of law as a system of known rules that can be straightforwardly applied to factual circumstances, persons or events must be called to a halt; which is partly why we are then able to see how judges’ decisions must be rooted in their forms of life, within a historically developed body of collective knowledge that cannot be fully represented. It cannot be otherwise, for, as Derrida, observes, ultimately ‘incalculable justice requires us to calculate’ (Derrida 1992: 28).