Public Reasoning without Sociology: Amartya Sen’s Theory of Justice

Conflicting Sensibilities

One of the problems in bringing Sen and sociology into dialogue concerns the relation between economics and sociology. At best, the disciplines are seen as complementary, but concerned with different topics, as, for example, indicated by domains of instrumentally rational action, or of value-rational action, as Weber proposed and many have followed. At worst, the disciplines are viewed as mutually opposed, with, from the perspective of economics, the one awaiting absorption to the other (Hirshleifer, 1985), or, from the perspective of sociology, the other as an ‘ideological’ expression of capitalism, rather than the means to its analysis (Clarke, 1982). The latter position is reconciled to the Weberian ‘resolution’ by suggesting that since capitalism is a domain of expanded instrumental rationality, the modes of analysis and techniques of economics have an expanded and appropriate application to contemporary societies, while sociology has as its domain the emergence of capitalist modernity, its wider institutional embedding and phenomena that are residual, or resistant to, market processes.

The latter position represents, perhaps, a ‘common-sense’ rapprochement that preserves ‘boundaries’ and prevents mutual antagonism. On the sociological side, the rapprochement is not the object of scrutiny or systematic reflection, since its function is that of maintaining a disciplinary modus vivendi, yet this is very close to being a description of the very interrelationships that Sen (2009) is seeking to address. For example, his interest in understanding the ideas of justice that pertain in non-modern settings and from which moderns can learn (for example, with regard to tolerance), and the merits of market rationality as well as its limits, all derive from a strong historical and multi-cultural sensibility that would normally be regarded as the provenance of a sociological interest. How then to explain the neglect of Sen in contemporary sociology? ¹

Sen’s own background in welfare economics and the theory of social choice places him in a branch of economics where the standard assumptions of economics are generally regarded as being most entrenched. The default view of this particular sub-field of economics is that choices that can be resolved by market mechanisms tend to optimality, while other forms of public choice using other mechanisms, such as voting, face various kinds of paradox and ‘impossibility’ (as Arrow, 1951, expressed it). Along with other kinds of claims about the advantages of markets compared with planning, and of the limitations of bureaucratic organisation, etc., the arguments of welfare economists and social choice theorists have been influential in forming not simply a kind of social science that was unsympathetic to sociology, but also a normative framework for public policy that was also unsympathetic to the reform-oriented political sensibilities of most sociologists.

This difference has recently found expression in Burawoy’s (2005) call for ‘public sociology’ in which he argues that economics as a discipline is associated both with the rise of the market economy and with proselytising for market-based solutions. In contrast, and echoing Karl Polanyi’s (1957[1944]) characterisation of the ‘anti-social’ consequences of the self-regulating market, Burawoy argues that sociology has as its aim the defence of civil society against ‘market tyranny’. Again, as in Polanyi’s argument that markets are ‘planned’, Burawoy suggests that ‘market tyrannies’ are frequently state imposed and, thus, involve an alliance between politics and economics. This is expressed in his 11th thesis that:

If the standpoint of economics is the market and its expansion, and the standpoint of political science is the state and the guarantee of political stability, then the standpoint of sociology is civil society and the defense of the social. In times of market tyranny and state despotism, sociology – and in particular its public face – defends the interests of humanity. (2005: 24)

On this powerful recent construction of the sociological imagination, then, the form of public reason expressed by sociology is at odds with that expressed by politics and economics.

Burawoy’s account may be too sharply and polemically put (Holmwood, 2007), but it provides a good illustration of a prevailing view of the relation between the disciplines in the latter half of the 20th century, at least as it is understood by sociologists. Thus, accounts of economics as a discipline (Fourcade, 2009; Yonay, 1998) suggest that a greater heterogeneity of approach evident in the first half of the 20th century gave way in the 1960s to the dominance of a neoclassical orthodoxy around the idea of ‘market efficiency’. However, within economics, as might be expected, there is a view that there is a greater heterogeneity of perspectives than is usually allowed and that orthodoxy is breaking down and not only because of the challenge of the current world financial crisis. Thus, Colander (2000) has argued that the term ‘neo-classical economics’ should be dropped because it no longer describes the assumptions of most contemporary economics. Similarly, Davis (2002) has argued that, while economics did rely on a theory of the individual derived from the classic Enlightenment conception of the self-calculating, atomistic agent, contemporary economics has ‘hollowed out’ even this idea of the individual such that it no longer has a theory of the individual at all. Along with Mirowski (2002), he suggests that economics has become a ‘cyborg science’ concerned only with representing economic processes according to the technically circumscribed constraints of formal modelling.

Notwithstanding these qualifications, Davis does acknowledge that the achievement of a coherent core to the discipline was the consequence of maintaining a stance detached from political engagement in order to secure claims to neutral expertise. However, he also suggests that although economics may lack a ‘monolithic’ ideological presence, it may nonetheless ‘exercise an undetected ideological presence, behind the backs of individuals, as if by an invisible hand’ (2002: 151). Moreover, the ‘hollowing out’ of the economic theory of the individual is continuous with the elaboration of a normative theory of the individual within political philosophy and ethics. In that context, then, it is perhaps not surprising that what once may have been seen as integral to economics as a form of ‘political economy’ (displacing the ‘moral economy’ of pre-capitalist systems; Thompson, 1971) should come to be understood as a matter of the interdisciplinary relations among economics, politics and ethics as formal disciplines with distinct, but commensurable, objects of study.

Indeed, this is evident in the trajectory of Sen’s own thought and of welfare economics within which it has been located. Thus, welfare economics also had its internal critics concerned with the normative assumptions of neoclassical economics and its tendency to elide scientific and normative issues. This had two aspects. The first was to criticise economics for being an axiomatic, rather than an empirical science. The second was to criticise the ethical assumptions implicit within those axioms. For present purposes, what is most interesting about this criticism, which begins just as the neoclassical hegemony within economics is being established in the 1950s, is that it is associated with economists whose primary engagement was with problems of economic development; first, Ian Little (whose Critique of Welfare Economics was published in 1950), and later, I suggest, Amartya Sen himself.

In fact, the debate between ‘axiomatic’ and ‘empirical’ approaches to economic methodology can be traced back to the 1930s and an important exchange between Robbins (1932) and Hutchinson (1960[1938]). ² The former sought to deny intersubjective comparisons of utility for introducing ‘values’ into an otherwise ‘scientific’ (axiomatic) subject, while Hutchinson was concerned to demonstrate the value position that was implicit in the axiomatic approach and that it constituted a form of a priori theorising that was not subject to empirical challenge. The debate is significant because Robbins’ economic methodology was also subject to critique in one of Parsons’ (1934) early essays that was formative in the development of his action frame of reference. The latter, of course, is itself an axiomatic account that sought to integrate sociological and economic elements of action in a single theory of action and, thus, from the perspective of Hutchinson would be subject to a similar criticism as that mounted against Robbins (though, as far as I am aware, neither Parsons nor Hutchinson took notice of each other, with the former no doubt regarding the latter’s argument as an example of the kind of ‘empiricism’ he believed to have been thoroughly dispatched by the arguments presented in The Structure of Social Action, 1937).

As will be clear, I am not promoting a putative sociological argument in Sen in terms of an expanded theory of action of the sort set out by Parsons (and, by the same token, by Habermas, or any other proponent of a priori general theory in which different elements of action are abstracted and formally synthesised). ³ There are a number of problems with this kind of argument that I return to later in the article. However, for present purposes, it is sufficient to note, first, that it relies on a representation of economics as continuing to be grounded in a ‘utilitarian’ theory of action, which, as I have already noted, is now denied by many practitioners, but which remains, nonetheless, a stock argument within sociological critiques of economics. Second, it provides a sociological equivalent of the argument that Sen is concerned to criticise, namely that there is a form of analytical theory of a very high degree of abstraction in which normative issues can be addressed separately from issues of empirical adequacy. In the next section of the article, I consider how Sen seeks to ‘deconstruct’ the more formal approaches to needs and human agency, before going on to a more direct consideration of the sociological significance of these arguments.

Capabilities

The core of Sen’s (1999, 2009) argument to bridge economic analysis and public reasoning is an argument about the nature of the individual as embodying capabilities that may or may not be realised. The standard idea of the rational actor is an asocial individual endowed with the capacity to act in relation to his or her own interests. For Sen, the problem is that the argument becomes tautological, or self-defeating, in the sense that the actor is viewed as endowed with just one capacity (or ‘capability’), which is necessarily realised in any action undertaken (Sen, 1977). At the heart of Sen’s conception of capabilities is the idea that there are a number of universal human needs associated with individual flourishing. ⁴ These needs are not indivisible and they cannot be presented along a single dimension, governed by a single principle, or algorithm (for example, whether of Pareto optimality, or Rawlsian ‘justice as fairness’, or, indeed, a Habermasian criterion of the ‘recognition’ afforded by undistorted communication). Improvements can be made piecemeal and they should not be judged from too lofty a perspective. The tendency with other approaches to welfare or justice is to reify institutions – for example, systems of market exchange – as the embodiment of supposedly key normative principles, when, for Sen, the emphasis should be on the pragmatic consequences of institutions for the realisation of individual capabilities.

Thus, Sen (2009) articulates a paradox within a liberal theory of justice. For most liberal writers, the appropriate question is whether institutions appropriately express theoretically established principles of individualism, while, for Sen, the issue is whether institutions allow the flourishing of individuals and he suggests that these are not the same question. From the latter perspective, a compromise of institutional principles may be necessary for the realisation of capabilities. Thus, Sen is neither for, nor against, markets; neither for, nor against, state direction. Markets have certain things in their favour for encouraging human flourishing (for example, allowing the choice of occupation), but may also have negative consequences (for example, creating insecurity of employment in the pursuit of an occupation) with serious consequences that require political intervention. ⁵ In this way, Sen asserts the universality of (diverse) capabilities while endorsing a necessary relativism of the means through which they might be realised. While ‘employment’ may be associated with freedom, the commodification of labour may also be associated historically with a reduction of capabilities, such that teleological accounts of progress – whether in their Marxist, or economic liberal forms – fail to recognise that what counts for those undergoing such processes is not a putative future freedom for others, but a specific capability deprivation that demands an address in the present in terms of justice.

I shall return to a discussion of how this universalism of capabilities and relativism of their realisation has confounded the (few) sociologists that have engaged with it. However, Sen’s critique of the standard economic axioms on the grounds of the incommensurability of the different capabilities he identifies is open to sociological argument precisely in so far as his argument has empirical content. In effect, Sen construes sociology as a ‘place marker’, the supplier of empirical instances to illustrate the argument about capabilities (for example, concerning issues of gender, or disability, as they bear on access to, and distribution of, ‘commodities’ necessary to realise capabilities). It is important to acknowledge that these examples can be incorrect in their own terms and in terms relevant to his argument about the specific deprivation that he is seeking to identify and promote measures to redress, but need not, by that token, undermine the general argument.

One example will make the point. It is frequently argued (including by Sen) that women are disadvantaged by their obligations for the care of others, for example the care of sick relatives. However, it is not the case that men do not also have such obligations, say towards a partner who is ill. In the UK, for example, what currently appears to be the case is that where an older male partner of working age is ill, the supporting female partner tends to be a full-time carer, but where an older female partner of working age is ill, the male partner tends to be in employment and, thus, there is a demonstrable inequity in caring arrangements (Arber and Ginn, 1991; Arber et al., 2003). However, this situation is open to different interpretations, other than the immediate possibility that men tend not to accept their obligations in the same way as women. Simply, the gendered operation of occupational pensions and the different entitlements of men and women deriving from different work histories are played out such that a compensating income is available in the case of male ill-health and not in the case of female ill-health. Where a female partner is ill, the male partner is disproportionately likely to need to continue in paid work, while, in the case of male ill-health, paid work by the female partner may be less necessary because a compensatory income is available from occupational benefits linked to pension entitlements.

Either explanation would be consistent with Sen’s ‘formal’ argument about capabilities; it matters which is correct from the point of view of developing social policies to enhance capabilities, which, from Sen’s perspective, should be directed at removing inequalities of either of the sorts indicated. The task of public reasoning is both to be clear about the nature of the injustice in terms of capabilities and to be clear about how it comes about so that robust measures can be designed to redress it. More problematic from his perspective, but still amenable to incorporation, would be the issue of whether or not capabilities vary across the life course. Clearly, the capabilities of children are different from those of adults, but to what extent do older citizens have lesser capabilities in the sense that ageing reduces capabilities and thus fewer resources might be required to secure their flourishing?

It should be evident that Sen’s arguments about capabilities have significant consequences for how we think about distribution within a market economy. In the latter, distribution is primarily organised through employment, where the labour contract and the operation of the market stipulate returns to an individual on the basis of his or her labour contribution. ⁶ Yet individuals differ not only in their skills and capacities, but also in their relations with others with claims upon them. For example, nothing in the operation of a labour market guarantees that it will provide incomes sufficient to maintain children. Moreover, where the labour market offers a differentiated range of jobs with different levels of rewards (given that the stakes are high, not least in terms of the consequences for dependants), it follows that there are likely to be exclusionary processes associated with access to them, at the same time as more advantaged jobs provide the means of ‘purchasing’ from the market forms of insurance against risks (thereby, having an impact on ‘political coalitions’ seeking the amelioration of risks through welfare benefits outside the market). In this way, the issues of distribution in a market economy are complex and have a nuanced relation to the idea of capacities outlined by Sen. Of course, there will be issues of class inequalities, gender inequalities and generational inequalities to be addressed and no reason to suppose that these map on to each other in a direct way. It is precisely this that makes sociological argument peculiarly relevant to the practical accomplishment of public policies based on public reasoning and directed at maximising capabilities.

I shall return to Sen’s reliance on ‘sociological’ argument (though I should note that the term ‘sociological’ here stands for empirical arguments supplied, for the most part, by colleagues in development studies and social policy). For the moment, I want to pursue Sen’s normative argument a little further.

Universalism versus Relativism

I have argued that the absence of sociological engagement with Sen is attributable to problems in the relations between the disciplines of sociology and economics over the last few decades. As I have already commented, the development of sociology has, by and large, been associated with the rise of the national welfare state, with economics associated with the rise of the market. While the development of sociology and economics were complementary, if distant, in the post-war period up until the 1960s, partly reflecting a political consensus in western societies on a mixed economy and the role of the welfare state, their respective trajectories seem to have been quite different since then. Heterodox approaches notwithstanding – including attempts to forge a Keynesian-Neo-Ricardian/Neo-Marxist alternative to neoclassical economics – it was the latter that achieved an unprecedented hegemony over economic discourse in the period after the 1960s (Fourcade, 2009). In contrast, although sociology was also divided between heterodox and orthodox approaches, it was the former that won out and forged a radical sensibility for the discipline.

In this context, sociologists were frequently not only ‘anti-capitalist’ but, increasingly, critical of the welfare state precisely in so far is it functioned for capitalism. In contrast, economists tended to emphasise the regulation of capitalism by markets and to represent the operation of markets in their abstract, ideal form as the realisation of norms of efficiency and the disaggregation of power (which was associated with monopolistic appropriation). In public policy, the state was treated as a form of monopoly supplier of services, to be criticised on the grounds of efficiency as well as freedom (Barry, 1990). There was, then, a convergence on the critique of the welfare state, but from alternative normative accounts of autonomy and freedom.

To use the terminology adopted by Sen, the dominant form of ‘public reason’ associated with the discourse of economics was close to the prevailing public policy discourse, while ‘public sociology’ was increasingly displaced (along with sociology’s more orthodox expressions). ⁷ In this context, the sociological argument that had the greatest resonance with the research programme being developed by Sen was that of T.H. Marshall and his discussion of social rights of citizenship. ⁸ Marshall’s (1950) conception of social rights identified the evolution of rights from legal and political rights of citizenship (the core rights of liberal citizenship) to the idea of a social right to a standard of living necessary for full participation as a member of the community or polity. This latter emphasis on ‘partcipation’ also has resonance with more recent arguments within social policy that have sought to extend the idea of capabilities to address processes of exclusion (Lister, 2004), as well as those setting out the idea of voice as a capability (Zimmerman, 2006). ⁹ Marshall also perceived the problem of maintaining this appropriate standard of living in the context of specific vulnerabilities associated with the life cycle and the contingencies of employment. For Marshall, such vulnerabilities could be met by ‘incorporating social rights in the status of citizenship and thus creating a universal right to real income which is not proportionate to the market value of the claimant’ (1950: 47). The idea of a ‘universal right’ is important because of the stigma associated with eligibility tests. ¹⁰

Marshall had also identified legal and political rights of citizenship and social rights as potentially incommensurable and in conflict, speaking of a ‘war between citizenship and social class’, as political rights gave rise to the demand for social rights. In principle, then, Marshall offers an account of capabilities and their institutional expressions that sees development not just in ‘additive’ terms, but potentially also in reconstructive terms. In other words, the role of markets as potentially both the means of the realisation of capabilities and as potentially their negation is something that is a matter of continued negotiation and renegotiation. ¹¹

However, the possibility of a Marshallian sociological ‘bridge’ over to Sen was undermined by other developments within sociology. The renewed sociological interest in Marxism gave rise to a renewed interest in class analysis, sensu stricto. Even if neo-Weberian approaches predominated, they tended to focus on employment relations and to regard poverty as a secondary aspect of the operation of the class system, rather than a topic in its own right. ¹² In this way, the study of poverty was given over to social policy. This is particularly the case once a relative poverty line is determined as a percentage of the average or median household income. Poverty becomes simply a line drawn within a distribution and the main focus of sociological research became social inequality, its structure and reproduction. In contrast, the mutual implication of poverty and social inequality is evident in Marshall’s conception of social rights. He writes that:

… class-abatement is still the aim of social rights, but it has acquired a new meaning. It is no longer merely an attempt to abate the obvious nuisance of destitution in the lowest ranks of society. It has assumed the guise of action modifying the whole pattern of social inequality. (1950: 47)

However, if sociologists now tend to address the structure of inequality, rather than its range and specific vulnerabilities within it, for his part, Sen seems to wish to separate the issue of those vulnerabilities from questions to do with the range of social inequalities and their reproduction, or, at least, this is how he is frequently understood. Thus, Sen is adamant that his concern is with absolute poverty and not with relative poverty. This was the topic of a rather ill-tempered exchange between Townsend (1985) and Sen (1983, 1985), in which the former argued that a sociological approach to poverty was necessarily relative, mobilising a Marshallian argument that the definition depended on ‘standards of the community’, which were necessarily variant. For Townsend, the concern with absolute poverty reflected Sen’s concern with issues of ‘developing societies’, but, in ‘developed’ societies, the minimum standard should be conceived in relative terms. For Sen, for his part, addressing the issue of poverty in relative terms in the way indicated by Townsend would tend to sanction differential standards across national systems and, thus, fail to address issues of global justice (1985, 2009).

To some extent the two spoke past each other and part of the argument reflected the contemporary politics of inequality. Townsend, for example, was concerned that an emphasis on absolute poverty would suggest that poverty was, by and large, overcome by secular increases in living standards in advanced societies and that this was a reflection of liberal ideology that might be concerned about destitution, but was tolerant of inequality as such (a point also made by Marshall). Sen, for his part, suggested that defining poverty in relative terms created an arbitrary definition of a poverty line (for example, a poverty line of household income less than 60 per cent of median income, as in current EU definitions) and that, in terms of ‘capabilities’, there could be individuals above the poverty line, but who were in real poverty because they required a larger income to meet the capabilities definition of flourishing because of some other characteristic. For example, gender relations in the household might mean that household income was distributed differently for men and women, disabled individuals, or individuals with caring responsibilities, might have additional costs that meant that the minimum income needed to meet the same needs (as defined by capabilities) was greater.

On the face of it, then, the difference between Sen and his sociological critics, such as Townsend, would seem to be that the latter emphasised class inequalities, while Sen is concerned with a wider set of inequalities that fall out of the perspective of class analysis. The matter is not helped by the fact that the sociological view has a problem with class inequality as such, whereas Sen does not. In other words, notwithstanding the performance of different countries in terms of the reduction of poverty when defined in relative terms, or in the range of inequalities (normally expressed in terms of the gini-coefficient), the sociological argument tends to be that the different societies all share a common character as class societies. According to the standard sociological account, the building can be understood as being equally capitalist regardless of whether it is a ‘sky-scraper’ or a ‘bungalow’ (to use Marshall’s resonant phrase).

The sociological argument seems to be that the problem of Sen’s argument is that it doesn’t matter to him whether the building is a sky-scraper or a bungalow, and that his approach smuggles in implicit liberal assumptions that means that he is, at best, indifferent between them, but willing to countenance that the sky-scraper may allow the realisation of some capabilities and not be inconsistent with others so long as they are attended to by specific measures. It is not helped by the fact that neoliberal policies over the last few decades have been designed precisely to increase the number of floors in the building and, therefore, that Sen’s resources for combating neoliberal policies (whether domestically, or internationally through agencies such as the World Bank) are weak.

However, I suggest that this standard sociological line against Sen is itself weak once we accept that any straightforwardly anti-capitalist argument lacks pragmatic cogency. The sociological critique reflects a political sensibility favouring lesser over greater inequality of outcomes, but does not deliver on it in terms of conceptual clarity, precisely because it is concerned with the capitalist structuring of inequality, rather than in differences among capitalist countries in terms of its range and consequences. So the question must arise, is Sen’s argument for a capabilities approach as blinkered on this question, as is implied by Townsend? It is well to recall that Sen’s argument for an absolute definition of capabilities goes along with recognition of the relativity of the means to their realisation, where, ‘absolute deprivation in terms of a person’s capabilities relates to relative deprivation in terms of commodities, incomes and resources’ (1983: 153). In other words, if the distribution of commodities or the political measures to address specific needs can be shown to impinge on capabilities, then there would be a capability argument to be made. For example, if we consider that the matter at hand is justifying a system of universal welfare benefits against a means-tested system then that could be done by stating that focusing scarce resources on those most in need would have the advantage of efficiency. However, if that is done by creating a stigmatising dualism between those identified by their reliance on benefits and those who secure their welfare by private, market-based means, then we immediately have recourse to a capability argument. The right not to be shamed is absolute. Equally, if it can be demonstrated that a wider range of inequalities has consequences for health (and other outcomes), as is suggested by Wilkinson and Pickett (2009), for example, then we have a capability argument for lesser, rather than greater inequality. ¹³ Nor need the capability argument be restricted to issues only of inequalities in the distribution of commodities, but also in the access to public goods. For example, while a capitalist economy may provide some with the means to afford second homes in the countryside, an important dimension for the exercise of capabilities might be that of open access to the countryside for those who cannot own a part of it.

In other words, variations within arrangements broadly represented by sociologists as capitalist may be as significant for their impact on capabilities as those that are attributed to capitalism. At the same time, some aspects of capitalist organisation – choice of occupation, self-determination – may make positive contributions to the realisation of capabilities. Walby suggests that the problem with the capabilities approach is that it emphasises ‘choice’ and is ‘equivocal’ about inequality, and so it has the consequence that, ‘because the door is deliberately opened to choice, it is thereby open to the possibility that people may choose inequality through their choice of a form of difference that is linked to inequality’ (2009: 10). In contrast, I have suggested that attention to outcomes is central to the capabilities approach and that what is missing in the sociological approach is precisely the urgency of a normative language to address inequalities in a context where some inequality is widely regarded as positive in the context of capabilities, because it derives from their exercise. The issue of choice versus outcomes may be vexed. But it needs to be addressed systematically and the problem with the sociological priority on outcomes is that it operates with an assumption that less inequality is better than more inequality, without explicitly advocating complete equality, at the same time as accepting that capitalism necessarily produces inequality.

Perhaps the most powerful argument in support of Sen’s ‘absolute’ approach to poverty is precisely that the argument for a ‘relative’ definition is reliant on the ‘standards of the community’. In the context in which it was articulated by Marshall, it may seem that the ‘standards of the community’ favoured greater equality and this was part of his teleological understanding of the shift from civil to political and social rights of citizenship. However, in the aftermath of several generations of neoliberal public policy, the ‘standards of the community’ have been consistent with elected governments that have favoured welfare cutbacks over welfare spending and labour market deregulation over regulation. In this context, the ‘relative’ definition has become an arbitrary line used by policy analysts and advocates. Given that current circumstances also involve public distrust of experts, the crucial task of public reasoning would seem to be to provide substantive arguments for why poverty matters to all of us as an issue of justice. ¹⁴

Conclusion

It is true that Sen’s account of the incommensurability of capabilities can be interpreted through a Weberian conception of different types of action that are mutually incommensurable; for example, instrumentally rational action, and value-rational action, with the latter also including the possibility of incommensurable values motivating the behaviours of different actors (something that led Robbins to deny the intersubjective comparison of utilities). This kind of construction is a staple of the sociological critique of economics, which is then regarded as over-emphasising one kind of action (instrumental) to the neglect of other types (see, for example, Gould, 1989). ¹⁵ However, it is a construction that also encourages the maintenance of distance between the disciplines and separates sociology from productive engagements with parallel developments within economics over the nature of fundamental categories of action and the individual.

There might be an objection that Sen’s account of the human individual is problematic precisely in so far as it is insufficiently ‘social’. Here we encounter a problem at the boundary among disciplines, not just economics and sociology, but also psychology and sociology. However, it is also indicative of a ‘retreat’ within sociology itself from its organisation in terms of the ‘social self’ evident both in the French tradition of Durkheimian sociology, and in the American pragmatist tradition of Dewey and Mead.

It is clear from Sen’s frequent invocation of Adam Smith and his argument that the latter’s Wealth of Nations needs to be understood through the Theory of Moral Sentiments that Sen is arguing not only for a richer psychological understanding of the individual than is normally found within the abstracted understandings of standard economic science, but also that the psychology he is advocating is a social psychology. No doubt, from a sociological perspective, Adam Smith hardly represents the last word on social psychology and there are sophisticated developments of similar arguments about the individual within the sociological tradition (for example, in Mead’s idea of the ‘generalised other’ as an alternative to Smith’s ‘impartial spectator’; see also, Zimmerman, 2006) such that the motivation toward justice can be grounded in a psychology that is neither ‘subjectivist’ (as in economics) nor dependent on a separation of reason and affect (as in much philosophy). If Davis is right that the economic concept of the individual has been ‘hollowed out’ from the Lockean construct of classical economics, then the time is ripe for sociologists to return to the idea of the individual within sociological accounts, with a view to address conceptions of the individual across the boundary currently separating sociology from economics. This would involve returning to the idea of a social self that overcomes the dualism of instrumental and value-rational action and analytical conceptions of action and agency that otherwise dominate in contemporary economics and sociology.

We might aspire for sociology to be more than a ‘placemarker’ for the empirical issues associated with the ‘relativity of commodities and resources’ and to contribute to the idea of public reason. However, this could not be done without recognising that Sen’s argument has force where much of what the sociological alternative currently provides is a sensibility – critical or analytical – without practical engagement with current public debates over inequality and its injustices.