Beyond Reason? The limitations of rationality – a tautological and asocial concept

Patrick Brown

School of Social Policy, Sociology and Social Research, Giles Lane, University of Kent, Canterbury, CT2 7NF, UK.

Abstract

Notions of rationality are frequently applied by social-scientists as a basis of describing, understanding and evaluating the actions of individuals and organisations in relation to risk-contexts. This paper will critically examine the utility of this contested concept for risk research. The gradual shift away from the invocation of rationality in a dogmatic sense has culminated more recently in approaches which consider multiple rationalities or understand certain action as functioning between the rational and non-rational. From such a starting point this paper goes further to argue that analyses of organisational, and especially individual, decision-making and behaviour are often hindered by considerations of rationality. It is suggested that the usefulness of rationality in describing and explaining action is limited in a number of ways, not least because of the tautological nature of the concept. Moreover a conclusive evaluation as to whether an action is ‘rational’ must be made ex-post the outcome – thus ignoring the multiplicity of contingencies, potential outcomes and limited knowledge apparent at the time of action. At the centre of the paper is the assertion that no action is irrational, and that ascriptions of such a label are merely indications of a limited grasp of the salient social variables. Three specific risk contexts are used by way of illustration.

Introduction: The problem of rationality as dogma

The concept of risk, and the range of policy and academic responses to it, has shifted many analytical lenses away from social structures and towards the individual. Spurred on by the emancipatory possibilities of progress and reflexivity, scientific and indeed many social-scientific approaches to risk have reinvigorating considerations of appropriate or effective human action in the face of future ‘unknowables’ (Keynes 1921). Whether it is in relation to entrepreneurial and investment decisions (ibid), criminal behaviour (Kemshall 2006), or the pension decisions taken by those reaching retirement (Vickerstaff et al 2008), analyses abound as to the nature and process of correct and less-than-correct – rational and irrational – courses of action. In the midst of an apparent demise of left-right politics (Giddens 1994) and related concerns of rationality at a more meta-system level (eg Habermas 1984), risk has emerged as a dominant political and intellectual ideology (Furedi 2005) which instead focuses attention on the behaviour of specific communities (meso) and moreover individual actors and institutions (micro).

Of course the scrutinising of decision-making and action as rational or otherwise is in many senses timeless. From the concerns of Heraclitus with the lack of reflexivity amongst the ignorant hoi polloi (Sandywell 1996:236), to the new public health of recent NHS policy (Department of Health 2004), societies have been rife with assumptions over the progressive potential for reason to illuminate and rectify the unfounded and/or flawed. The Enlightenment bolstered confidence in the possibilities of reason to progressively elucidate and refine approaches to human problems. Importantly, the vigour by which the Enlightenment thinkers such as Voltaire, Hume and Montesquieu sought to replace a single-dominant metaphysics (Porter 2001:26) seemingly imbued its successor with a similarly dogmatic presumption about the existence of a single reality - around which one undisputable reason could be honed (Gray 2009).

The hegemonic position of scientific reason makes possible the elevation of generalisable knowledge (Habermas 1981; Adorno and Horkheimer 1973) and a corresponding deprivileging and debasement of subjectivity (O’Neill 1995). Such ontological and epistemological gerrymandering (Woolgar and Pawluch 1983) which this dogmatic rationality facilitates is very much the contextual basis of this article. From here the next section will rehearse positions which note a multiplicity of rationalities at work within social systems and lay society (eg Weber 1968; Habermas, 1987; O’Neill, 1995). Yet the difficulty for any one of these forms to offer a satisfactory account of human action amidst uncertainty has led more recent work (Zinn 2008) to suggest decisions take place across a spectrum spanning the rational (calculative), non-rational (belief, hope), and in between these ideal-types (trust, intuition, emotion). A positive critique of this position will begin by problematising the demarcation between rationality and alternatives. The phenomenology of Schutz (1972) will be used to illustrate that calculative and emotive responses are similarly grounded in prior socialised experiences.

It will be contended that the obscuring of ‘the social’ by the application of rationality hinders understanding – blinding analyses to structural factors. These limitations, alongside the tautological nature of the concept, form the backdrop to the central thesis of this paper: that social scientific accounts of action amidst risk contexts would be more complete by moving beyond notions of rationality. The final three sections will explore and expand on this theorisation through brief, specific, policy-focused case-studies: restorative justice, pension planning and financial markets. These examples underline the significance of multiple layers of social experience in determining the manner by which knowledge is both accessed and processed. The decisive influence of the social on how knowledge is communicated and received illuminates the structural factors which bear on actors. Conversely notions of rationality would seem to fetishise the utility of refined or generalisable knowledge, at the expense of comprehending the inherent rationality in all agency and the social structures which bear upon this.

From one rationality to many: the multi-dimensionality of human action

The dogma of rationality which Gray (2009) sees as emblematic of the Enlightenment, epitomised for him within the works of Saint-Simon and Compte, has come under considerable and sustained critique. Kant was a contemporary of the Enlightenment who, in his concerted attempt to free humanity from a dogma of metaphysics (Kant 1929: 30), builds on the work of Leibniz and Hume to distinguish between different types of judgement and their relation to human experience. In differentiating between knowledge forms, and moreover reinforcing the limitations of the knowable, Kant lays the groundwork for a contestation of rationality.

In a not dissimilar fashion to Kantian dualisms between the theoretical and the practical, Weber (1968) distinguishes between purposive/instrumental rationality and other formats relating to norms and values, emotions and tradition/habit. Whilst Weber’s use of the term rationality has been criticised as imprecise and fluid (Swidler 1973), his description of human action as motivated by combinations of differing rationalities nevertheless facilitates a more nuanced understanding of behaviour. One that is more open to acknowledging the groundedness of that which would appear untenable from a purely purposive-rational perspective. This perspective is central to Weber’s analyses of religious affiliations, yet his work on bureaucracy also has much light to shed on the nature of 'rational' decision-making.

For Weber, the form of rationality that develops within bureaucracies is very much driven by their social structures - thus reflecting a more general consideration of rationality as a socio-pathological trait[1]. Yet whilst Weberian accounts of subjective and moral reason tend towards the pessimistic and at times pejorative, two positive critiques of his work (Habermas 1984; Schutz 1972) more readily acknowledge the inherent value and utility of communicative and inter-subjective (as opposed to instrumental-scientific) rationality. Habermas (1987: 204-234) traces a line of thought which illuminates the limitations of Hobbesian (1968) understandings of man as solely driven by purposive-rational tendencies. Building on the empiricism of Locke and Simon and then the extra-contractual emphases of Durkheim and Parsons, a conceptualisation of communicative rationality is developed which understands the socialised bases of action flowing out of socialised norms and values as much as strategically-reasonable interests (see Brown 2008).

Perhaps more importantly for the concerns of this article, Habermas (1971) draws a clear distinction between different types of knowledge as they relate to human experience - as determined by their respective social contexts. Public experiences are those which can be expressed by generalisable validity-claims - therefore accessible to many, and moreover to institutional bureaucracies. Meanwhile knowledge generated from 'private experiences' is barely utterable, with the accessibility of these therefore restricted to a much smaller number of actors more immediately present. Whilst publicly available knowledge, through its generalisability, is capable of having its rationality validified/contested, private or experiential knowledge is often beyond such acknowledgement. Thus critical theory makes clear two key limitations in contesting the rationality of behaviour: a) human action is oriented around multiple-considerations[2] (rationalities) and therefore should not be evaluated on a single, instrumental basis; b) an important part of human experience is beyond generalisability, with this 'invisibility' obstructing empirical examinations concerned with rationality and reason.

Between the rational and the non-rational? Or inherently rational?

Phenomenological insights into human action (Schutz 1972) represent a further challenge to rationality-oriented understandings of human action. The possibility of an actor understanding any notion, or any projected action, is grounded in their direct experiences as it is only via these, the building blocks of cognition, which make possible the inferential constructions required to comprehend the attitudes of others or indeed more abstract notions (ibid). This means of comprehending how actors appropriate knowledge is significant in that direct experiences are typically more concrete for individual actors than more abstract ideas - as the latter require multiple inferential leaps and are accordingly more tentative. So whereas confidence in scientific progress and the Enlightenment of society fetishises the emancipatory possibilities of knowledge and its public communication (as too does Habermas – eg 1984:333) - phenomenology makes evident that human action, and understandings thereof, are decisively rooted in private, socialised, experiential contexts.

O'Neill (1995:179-182), drawing on the work of Schutz (1943), argues against distinctions between what he calls "common-sense" action and a scientific rationality. Given that everyday human behaviour meets the criteria of being 'reasonable' and 'deliberative', and that the ideal-type of rational action is "so idealised that it cannot be found anywhere in the world" (O'Neill 1995: 181; Keynes 1936), it seems that the only aspect which sets lay behaviour apart from its generalisably rational counterpoint is the formers causal complexity. This points towards the tautological nature of the ascriptions of rationality: How do we know actor a is rational actor? Because s/he follows action x given context D. Why does s/he follow path x? Because s/he is a rational actor[3].

A recognition of this system-specific, self-referential and circular conceptualising of rationality in analyses of human decision-making is the first stage in illuminating its problematic nature. If, as implied by O'Neill (1995), lay decisions are set apart by their socially structured complexity - then defining their multi-dimensional manifestations as irrational is merely erecting a schematic obstacle to effective understandings of behaviour; making for a lazy social science. Zinn (2008) posits behaviour as often taking place between the rational and 'non-rational' (via intuition, trust and affect) and in so doing seeks to avoid privileging one over the other. Following Wynne (1996) and Gigerenzer (2007), Zinn acknowledges the mutual effectiveness and necessary complimentarity of the two approaches - even highlighting the superiority of the non-calculative (2008: 444). This recognition of similar utilities and effectual outcomes is doubtless a step forward, yet continued references to two ideal-typical positions within decision-making processes ignore significant commonalities.

All decision-making, whether more calculative, affect-based or 'innate'[4], is profoundly shaped by prior socialised experiences (the biographical). Comprehension of more formal or systemised knowledge systems functions through a number of inferential steps rooted in direct experiences of the individual actor (Schutz 1972). Their continued application is based on previous positive feedback from their use and/or socially constructed confidence in their epistemic authority (eg bio-medicine). The very enactment of the science, and moreover its refinement, would be impossible without common-sense judgements (O'Neill 1995: 160). Thus science, as an imperfect analogy useful in interpreting the natural world (Popper 1959; Archer, Bhaskar et al 1998), is simply a more generalisable configuration of the heuristical judgements of/by which it is composed.

Means of making judgements through less generalisable inferences – for example as intuition, trust, and/or affect (Zinn 2008) - are similarly refined strategies learnt through prior socialised experiences (Schutz 1972). The post-London-bombing experiences of John Tulloch[5] (2008) illustrate how even the most innate reflexes and reactions are learnt/adopted via prior direct encounters. Once again perceptions of the continued appropriateness of intuition, trust or affect are configured around ongoing feedback in social settings and their constructed validation by the lifeworld (eg through interactions with close acquaintances, social networks, socio-cultural norms, habitus and personal reflexivity). The preceding and current paragraphs make evident therefore that all decision-making can be said to be neither perfectly rational nor irrational. Rather all human agency is inherently, and in many senses similarly, rational. Focusing on the nature of this agency in terms of its rationality/irrationality/non-rationality however obscures the decisive social structures which determine the type and extent of knowledge available to the particular actor, and the way such knowledge is accessed/recognised/ processed.

Practical context 1: Restorative justice and offender rehabilitation

Thus far it has been noted that decision-making, both in relation to risk and more generally, is best described as: oriented by values rather than mere instrumentalism (Habermas 1987); driven by far more than purposive-rational efficacy (Weber 1968); and informed by experiential knowledge as derived from private experiences and/or in a more generalisable sense (Schutz 1972; Habermas 1971). The multi-dimensionality and complexity of these processes - not to mention the social contexts they are embedded in - represent a significant challenge to the social scientist and policy-maker; one which should not be explained away by rational, irrational or between-rational labelling.

Whereas the discussion up to this point has been at a relatively abstract level, the arguments developed will hopefully become more clearly formed and nuanced through the practical policy contexts explored from here onward. Three examples are briefly outlined which deal with a variety of 'risk' relationships in terms of how institutions engage with individuals: the institutional management of risky individuals in society (restorative justice); institutions trying to persuade members of the public to change the way they consider risk (extending working-life and pension deferral); individuals working within institutions dealing with risk (investment finance).

The multi-dimensional structuration which bears upon action, as referred to above, is patently evident within the context of restorative justice. This approach – to the criminal behaviour often characterised as a risk to society in many senses – is a more relational, social contextualising approach amongst stakeholders than some alternative approaches to crime (McLaughlin et al. 2003). Yet by its future-orientated, conciliatory approach the criminal act is still posed as deviant (irrational) and correctable through effective communicative action and knowledge exchange (Marshall 2003). The limited effectiveness of restorative justice may be partly explained by its failure to comprehend criminal activity as inherently rational agency, thus blinding itself to the potency of structuration.

Conceptualising agency through the multi-layered socialisation of the actor is central to its effective analysis[6]. The capacity for actors to think about the consequences (social and otherwise) of their actions and to make decisions in a more general sense is vitally dependent on the effective maturation and integrity of their prefrontal cortex (Anderson et al. 1999). Impairment to such neurological functioning inhibits moral judgement and makes ‘anti-social’ and/or violent behaviour more likely (Greene and Haidt 2002). Prima facie this might lead analyses towards decidedly psychological or even phrenological avenues. However the work of Schore (1999; 2001) highlights the extent to which social environments and stimulation during infancy drive the development of this and other parts of the brain’s decision-making apparatus. Hence what might be construed as malevolent or psycho-pathology is in fact socio-neurologically-influenced decision-making.

By recognising the manifold ways in which prior social experiences influence the 'innate', and moreover the ways in which socially-mediated information is processed, the problematic disjuncture between the psychological and the social is bridged. Thus the primary layer of social structuration, which bears on the maturation of the prefrontal cortex, is followed by a complex lattice of further experiential learning and stimulation within an array of social contexts – from the familial to the institutional. These socialisations form the proverbial classrooms in which ideal-typical assumptions are formed as to the dependability (or otherwise) of authority figures/groups and the outcomes (positive or negative) of certain actions towards these. Of course the nature of these very experiences is determined by both the qualities of the persons involved as well as the inclusion/engagement or exclusion of the actor within these learning contexts. Amidst these ongoing experiences a range of attitudinal positions are determined (Schutz 1972) at both cognitive and affective levels.

Such experience-based formulations of societal expectations (rooted in class, gender, ethnicity and age) and the actor’s cognitive and emotional responses to these, in relation to others, are the basic apparatus on which certain modes of decisions/behaviour (labelled by society on occasions as ‘criminal’ or ‘benevolent’ – flawed or appropriate) might be understood. These social factors (alongside other socially influenced mitigators such as substance misuse or mental health problems) thus determine the way information regarding certain environmental contingencies is processed. Meanwhile the contemporaneous socio-economic factors which influence the comparative appraisal of certain avenues of behaviour and understandings around these underline the extent to which ‘the social’ bears on the availability and natureof information. This multi-layered lattice of social experience will thus determine the ex-post attitudinal position of the (criminal) actor.