Amartya Sen as a social and political theorist – on personhood, democracy, and ‘description as choice’

ABSTRACT

Economist-philosopher Amartya Sen's writings on social and political issues have attracted wide audiences. Section 2 introduces his contributions on: how people reason as agents within society; social determinants of people's (lack of) access to goods and of the effective freedoms and agency they enjoy or lack; and associated advocacy of self-specification of identity and high expectations for ‘voice’ and reasoning democracy. Section 3 considers his relation to social theory, his tools for theorizing action in society, and his limited degree of attention to work by sociologists and to capitalism and power structures. Section 4 characterizes a style marked by conceptual refinement, emphases on complexity and individuality, including personal individuality, and reformist optimism. Section 5 shows the features from Sections 3 and 4 at work in his conception of personhood that advocates freedom to make a reasoned composition of personal identity. Similarly, Section 6 addresses his conception of public reasoning and neglect of the sociology of democracy. It contrasts the ideal of a reasoning polity with features in many countries. Sen's programmes for critical autonomy in personhood and for reasoned politics play, nevertheless, a normic role, while his analytical formats help investigation of obstacles to more widespread agency, voice, and democratic participation.

KEYWORDS:

• Democracy
• personhood
• identity
• deliberation
• capability approach

1. Introduction

Amartya Sen works as an inter-disciplinary philosopher-economist who has reconnected economics and ethics. He is also an acute social commentator, outside the disciplinary streams of sociology, anthropology, or political science. His work can be understood as participating in projects of normatively oriented social economics, moral philosophy and political theory that build from bases older than Western ‘social theory’ (Callinicos 2007; Giddens 1971), including from Adam Smith (1723–1790), Condorcet (1743–1794) and John Stuart Mill (1806–1873) (see, e.g., Duncan 1973; Gordon 1991). Besides then contributing towards a humanized economics, he has pursued a vision of a reasoning polity inspired by the Millian notion of democracy as ‘government by discussion’.

Central in Sen's work has been the ‘social choice’ perspective established by economist Kenneth Arrow (1921–2017), extending Condorcet. It focuses on ‘methods of marshalling information, particularly those relating to the people involved, to arrive at correct social judgments or acceptable group decisions’ (Sen 1986, 1073). This normative ‘social choice theory’ does not contain much ‘social theory’. Sen's magnum opus in social choice theory (2017a) has no reference to sociologists. Its theorizing, centred like Adam Smith's on a model of reflective individuals, needs deepened pictures of the individual and of culture, to better understand politics, including crowds, populism, and really-existing democracy. His links to sociology, anthropology, psychology and other social theory have been relatively thin (see, e.g., Gasper 1997, 2000, 2002; Gasper and van Staveren 2003; Holmwood 2013; Sen 1990). Nevertheless, Sen has moved, far beyond a simple conception of persons as traditionally held in economics; and, further, has in recent years published books of social analysis dealing with issues of identity and democracy (Sen 2005a, 2006, 2015; see also 2021). We will consider their treatment of persons in society.

The paper analyses Sen's emergence and profile as a social and political thinker who discusses personhood as multi-dimensional, champions the freedom of persons to make a reasoned composition of their own identity, and advocates ‘government by discussion’. Section 2 sketches the basis for this in his main contributions in social and political analysis. Section 3 then looks at his relation to existing social theory, including connections, contributions and blind spots. Section 4 characterizes his distinctive style, marked by emphases on multi-dimensionality, heterogeneity, and individuality. The style seems to bring a feeling of distance from much social theory, perceived as being too collectivist. Arising from this, Section 5 discusses his ambitious notion of personhood. Section 6 then considers his associated focus on an arguably idealized version of public reasoning and his neglect of the sociology of democracy. In this case, Sen's extraordinarily discerning style of conceptualization seems underused; rather than stressing numerous variants of democracy he has articulated a generalized advocacy. However, one might generously view this as an exercise in ‘description as choice’ (Sen 1982), a putting of emphasis on desirable states that are considered possible, even if they are far from widely prevalent yet.

2. Sen's main areas of contribution to social analysis

This section presents four prominent areas of work by Sen that can be seen as contributions to social analysis: regarding the nature of reasoning actors, and the social determinants of people's access to goods; the edifice of human capability theory; and his ambitions for liberal specification of identities and for a deliberative form of democracy. We will see subsequently how the later ones have grown out of the earlier ones.

How people reason about choices, as social actors

Sen became famous from the 1970s, through showing the extreme narrowness of so-called Rational Choice Theory (Sen 1977, 1985). This asserted that only one type of choice (maximizing/optimizing, typically in terms only of self-interest) was rational, as opposed to understanding rationality as choice on the basis of reasoning that one can sustain in the face of critical scrutiny (Sen 2009, 180). Some elements of his enriched conceptualization of reasoning and choice deserve mention here. Sen neither stands the person outside society nor merges the person into society.

First, people's ideas, including their preferences are not formed and fixed outside society. Sen remarked for example how many women imbibe masculinist ideology and are even the proximate drivers for female feticide (e.g., 2005a, 239–240). However, he considers that, in general, people have the potential to achieve not only autonomy of agency – ability to act independently – but critical autonomy, ability to think independently (Sen 2005a, 239). (Sen's own definition of agency as ‘the pursuit of goals and objectives that a person has reason to value and advance’ (2005a, 221) does not exclude critical autonomy but is less clear than these terms from Doyal and Gough 1991.)

Similarly, second, Sen makes much use of the notion of ‘adaptive preference’, which refers to when one's thinking adapts to normalize one's situation; privileged people can become blind to their privileges or consider them normal, and disadvantaged people may sometimes count only their blessings, not their deprivations. Much such adaptation involves internalization of social norms, like a belief that girls require fewer opportunities than boys. But Sen and associates have developed also the counter-concept ‘capability to aspire’ (e.g., Hart 2013), which modulates Appadurai (2004)'s notion of ‘capacity to aspire’, and even see it as a meta-capability that underpins many specific capabilities.

Third, going further, people (usually) do not reason only in terms of personal self-interest; they commit to social ideals, including through personal reflection not only socialization. Adam Smith ([1790] 1976) had distinguished, first, sympathy, where one feels for/with others, e.g. feels better when they prosper; second, generosity, where one sacrifices some of one's well-being for known others; and third, public spirit, where one sacrifices for a wider group, adopting the standpoint of the nation. Sen from the 1970s advanced a concept of ‘commitment’, as supplement to ‘sympathy’; it spans ‘generosity’ and ‘public spirit’. He later recognized Smith's set of concepts as more refined; but noted also the cosmopolitan spirit of human rights, not confined by national boundaries.

Fourth, deriving from Marx's idea of ‘objective illusion’ (Sen 2021, 216), he highlights the cognitive independence but specific location of each observer, even if not the social moulding of cognition. His notion of ‘positional objectivity’ refers to when different observers have the same perception when viewing from the same observational position but have differing perceptions when their observational positions differ. Position-based comprehension often leads to biases, which can sometimes be overcome but not always fully. This has implications for a theory of democracy. Our positional biases become more extreme when not balanced by the broader attention and learning that arise from sympathy, generosity and public spirit.

Much of this enriched conceptualization can be seen as basic social theory that had become ignored in modern economics. Drawing on reflective philosophers and earlier social economists, Sen helped to (re-)introduce such distinctions into formal economics and choice theory. His discussion of reasoning, norms, ‘commitment’ and objectivity continues though in the same philosophical style as those predecessors, without much reference to evidence from modern psychology on thought processes, accompanying emotions, their evolutionary basis, or so on (unlike, e.g., Nussbaum 2001; van Staveren 2001).¹

Entitlements analysis – socializing economics

Entitlements analysis places these reasoning social-economic actors within an institutional setting and examines their forms and degrees of vulnerability. It is ‘socially disaggregated, institutionally aware, analysis of effective command over specific necessities’ (Gasper 1993, 679), including of its various channels and determinants, notably the rules and institutions that control access. It grew out of Sen's work on explanation of famines (1981) and became applied to investigate hunger and poverty more generally (Sen 1984; Drèze and Sen 1989). It then exercised a formative influence in vulnerability studies. Adger (2006) identifies Sen's influence on major bodies of literature on vulnerability; and notes similarly ‘the findings of Janssen et al. (2006) on the importance of Sen (1981) as a seminal reference across many areas of vulnerability research’ (Adger 2006, 270). Characteristic of the approach is to distinguish many social groups, including not only in terms of income or economic class but also by occupation, gender, and more. This attention to social differentiation, beyond what had been normal in economics theory and practice, and to the institutionalized social norms that mould entitlements, gives channels for enriching economics with social analysis.

The capability approach (CA) – humanizing economics and influencing social analysis

Building on entitlements analysis of who gets what, Sen has explored who can live (or die) how, with reference to what are important elements of human lives and what human meanings they carry. His capability concept reflects a concern with reasoned freedom not only actual activities, and the importance of having both choices and effective power to attain reasoned desired outcomes. He has sought to recapture the potent ‘freedom’ concept, not abandon it to his right-libertarian contemporaries Milton Friedman and Robert Nozick.

People remain unempowered as a result of a variety of complex processes. … Quiet acceptance – by the victims and by others – of the inability of a great many people to achieve minimally effective capabilities and to have basic substantive freedoms … We have to see how the actions and inactions of a great many persons together lead to this social evil, … [and how] the remedy too can come from the co-operative efforts of people at large. … ‘active citizenship’ can be a very effective way of seeking and securing solutions to these pervasive problems of powerlessness and unfreedom. (Sen 2012, ix–x)²

Capability and entitlement analyses together provide both ‘a critical [evaluative] yardstick against which to assess the social world, its structuring principles and individual effects[, and] a series of items suited for empirical inquiry – opportunities, resources, entitlements, conversion factors, achievements’, that structure a research programme ‘for an understanding of the mechanism[s] underpinning [or inhibiting] individual agency’ (Zimmermann 2018, 942; my additions). These features lead Gangas (2014, 2020) to argue that, for example, the concept of capability deprivation provides a more refined, flexible, empirical research programme for understanding the multi-dimensional idea of ‘alienation’ than do the now overextended and insufficiently specified usages of that term in contemporary sociological theories.

Gangas argues similarly in relation to ‘agency’: ‘the normative components of Sen's notion of capabilities are not necessarily locked into an individualist approach to agency, typical of economic thought, but, rather, contain a social core that has been prefigured primarily by Parsons and to a lesser extent by Giddens’ (Gangas 2020, 9). Anthony Giddens (1984) defined agency as the ‘capability to act’. Talcott Parsons stressed ‘capacity’ as central in his theory of action in society and as a condition for achieving real citizenship. His ‘vision of capacity, no matter how abstractly formulated, can be viably likened to the normative program of capabilities, aiming at a social self, free of major deprivations and sufferings’ (Gangas 2016, 34). Sen's description of agency goes further, for example through his attention to information availability and thus to how far agents can compare current arrangements with other possibilities. ‘Sen's idea of capability can render concrete the abstractions entailed in [Parsons’ and Giddens’] formulations of agency’ (Gangas 2016, 33), also because Sen focuses on specific issues such as morbidity, mortality, education, voting and employment, and provides more plentiful and more operational concepts for discussing what influences them. The CA project can in return, Gangas suggests, be strengthened by connection to Parsons: to his exploration of the institutional infrastructures needed for personal capacity and his ‘mapping of the normative patterns which amplify [or inhibit] actors’ capabilities to choose the lifestyle(s) they have reason to value’ (Gangas 2016, 35; my addition).³

Democracy and the reasoning citizen – an ideal of politics and public life

Lastly, Sen has been a theorist and champion of democracy in general and more especially a deliberative style of democracy. This includes a stress on ‘voice’, Albert Hirschman’s (1970) more political partner-notion to ‘agency’ (e.g., Sen 2005a). Voice reinforces and relies on agency. The empowered active citizen demands, negotiates and helps to construct effective entitlements.

Connected to Sen's picture of public reasoning is a view of how participants in public arenas should see themselves (Sen 2006): they should recognize their plural affiliations plus develop a reasoned individuality. This reflects a strong liberal insistence on seeing people as ‘persons with the “capacity to act and the freedom to choose”’, including in respect to their self-description, ‘rather than being reduced to simplified categories, such as [only] workers, unemployed, mothers, etc. (e.g. Sen 1999a, 295–296)’ (Kremakova 2013, 398). We examine this in Sections 5 and 6.

Before that, Section 3 considers how Sen’s contributions to social analysis relate to approaches in other social sciences, especially sociology; and Section 4 argues that they embody an intellectual style that emphasizes multi-dimensionality, diversity, and ‘description as choice’.

3. Sen and social theory – criticisms and some responses

Sen’s relationship to sociology and anthropology

‘[T]he capability approach has remained largely unnoticed by sociologists’, warned Kremakova (2013, 394). Several other authors have commented on the lack of reference by sociologists to Sen's work; for example, Venkataraman (2016) specifically for Indian sociology. The same may apply amongst anthropologists, with Arjun Appadurai an exception. Holmwood noted weakness of the reverse connection too: Sen ‘makes very few references to sociological research despite it having clear relevance to his interests’ (2013, 1171), including in his books on sociological topics (2005a, 2006, 2015).⁴

Holmwood goes on: ‘Sen construes sociology as a … supplier of empirical instances to illustrate the argument about capabilities’ (2013, 1176); he does not engage sociological theory. The relative lack of interest in reverse, from sociologists for Sen, might also reflect, Holmwood suggests, their traditional preoccupation with economic class. That led often to relatively low focus (unlike in the social policy field) on the wide variety and diverse incidence of social inequalities that Sen investigates. Hartley Dean, himself professor of social policy at LSE, spoke though for many readers when assessing the capability approach and Sen's work more widely:

[Capability] is essentially a liberal-individualist concept. Despite its attractions … the “capability approach” obscures or neglects three key realities: [1] the constitutive nature of human interdependency; [2] the problematic nature of the public realm; and [3] the exploitative nature of capitalism. … [An] emancipatory politics of needs interpretation … would be better served by a discourse of rights than a discourse of capabilities. (2009, 261)

In contrast to Dean, some sociologists and other social scientists argue that Sen's capability thinking provides helpful tools for explanation as well as for normative purposes, although it indeed ‘must be complemented with theories that enable a better understanding of social structure, human sociality, collective living and the meaning of social action, for example within the philosophical traditions of Hannah Arendt and Paul Ricoeur’ (Kremakova 2013, 413, referring especially to Deneulin, Nebel, and Sagovsky 2006). Several have initiated this work of partnering, drawing for example also from Dewey (e.g., Kramm 2019; Zimmermann 2006, 2018), Parsons (Gangas 2020) and Bourdieu (e.g., Hart 2013; Pham 2019). Gangas agrees on ‘the marked omission of sociology in core principles of CA’ (Gangas 2020, 7) and warns that continuation of its ‘persistent repulsion of sociological theory’ (13) could leave CA in a dead-end. But he believes the limitations are remediable and worth remedying, for CA's conceptual system expresses a concern with purposeful, just, flourishing human lives, that can help to motivate, focus and complement existing sociological theorizing.

Some of Dean's arguments carry weight; some reflect misunderstandings or concern gaps that can be filled. Regarding human interdependency, Dean remarks that ‘the capability approach to equality is framed in terms of freedom, but not solidarity. … the freedom to choose, not the need to belong’ (2009, 267). He suggests that this gives an inadequate understanding of human living. Being, for humans, is constituted by and critically reliant on dependencies, support by others, not only freedoms; humans are vulnerable and needy, not only autonomous agents. Our affiliations to and dependence on others are not merely a route to our own capabilities but ‘are constitutive of our individual identities and the frameworks of meaning by which we value various functionings’ (2009, 268). Recognition of needs to belong is indeed fuller in the work of Nussbaum, who highlights affiliation as central, than in Sen's. His use of freedoms language can become overextended (Gasper and van Staveren 2003); but by freedoms he explicitly means capabilities to achieve (reasoned) values, which will certainly include numerous affiliations. Nor does he have anything against freedoms which arise from dependencies. His theory of identity highlights the multiplicity of our affiliations but argues that these should be open to reasoned review; and is aware of dangers that can arise from the pursuit of needs to belong, including from stereotyping, excluding and victimizing those who are deemed not to belong.

Dean observes further that ‘our ability at any particular moment to function as we choose may necessarily be achieved at the expense of others’ freedom’ (2009, 273). This is true, but the equivalent applies also for fulfilment of the need to belong. Every value may conflict sometimes with other values or with the same value as enjoyed by other people. Sen never proposes that freedoms cannot conflict or are the only type of value. The criticisms by Dean regarding capitalism and the public realm do, however, have more strength, as we see later.

Links, limitations, and opportunities in relation to social theory

We noted that Sen's declared intellectual base is ‘social choice theory’, which despite its name contains little social theory; it is a branch of choice-theory. He has an economist's strong emphases on choice and agency, but at the same time has always been critical of a simplistic picture of persons common in much of economics and politics. So, how far does he consider human sociality, social structure, and socially derived mental programming and cognitive and emotional constraints, and what potentials exist then for cooperation with sociologists?

We should describe a presumption of separation and detachment of persons from each other as ‘ontological individualism’ and reserve the term ‘methodological individualism’ for analysing from the perspective of individuals. Sen adopts the latter, not the former. He is aware of the importance of affiliations and belonging; but does he excessively rely on methodological individualism? Gangas, for example, suggests that Sen's ‘methodological approach over-relies on economics and thus lacks fine-tuning to sociology and in particular to theoretical models that address the problem of market embeddedness in society’ (2020, 8). Holmwood concurs. He suggests that Sen has far more influence in analytical philosophy and political science because these disciplines are, like economics, more methodologically individualist than is sociology. We see many PPE degree programmes, very few PPS; and some ‘welfare economists and social choice theorists have been influential in forming … a kind of social science that was unsympathetic to sociology’ (Holmwood 2013, 1173). Sen shares the ideal of Condorcet, the reasoning society able to change its behaviour, as opposed to what he considers a Malthusian notion of animal-type societies bound by fixed laws of motion (Sen 2009, 112; 2017a, Ch. A6).⁵ Holmwood proposes though that ‘Sen's approach to markets is similar to that of [the economic sociologist] Polanyi and that we might regard the idea of capabilities as giving substance to the latter's idea of “complex freedom”’ (Holmwood 2013, 1183).

As Kremakova notes (2013, 399), Sen is well-aware of various social influences, interactions, negotiations, and constraints. Entitlements analysis and capability theory consider many in detail. Let us unpack ‘social constraints and influences’, to ask which have been covered and which not.

• Enthusiasts for Sen's language of freedom and agency are often weak in their attention to social structures, argues Andrew Sayer amongst others (e.g. Deneulin 2006). He warns that ‘[CA's] radical implications are mostly being missed, largely on account of attempts to use its normative theory without an adequate account of the social structures that enable or limit human capabilities in particular situations’; CA has been typically combined with ‘inadequate theories of society, particularly regarding the external conditions enabling or limiting capabilities’ (Sayer 2012, 580).

• Gangas suggests in similar vein that CA ‘says very little about role placements and role dispositions’ (2016, 114) and about the institutional configurations required to sustain fulfilment of priority capabilities and functionings. Sen (1999a) writes of a series of essential ‘instrumental freedoms’ but without, unlike many sociologists, much depth on the institutions that might sustainably support them (e.g., Parsons and Smelser 1956).⁶

• Many authors (e.g., Deneulin 2006) note that collective actors rather than individuals are essential in attempts to change or counteract social structures, including norms and roles.

• Sayer, Gangas, Deneulin et al. all suggest though, unlike Dean, that the above common gaps or weaknesses in use of CA are remediable and being remedied.

• More attention is given in CA to the social influencing of preferences, including some attention to unconscious internalization of social norms, an easy step to make beyond ‘adaptive preference’. So Gangas sees in Sen's The Idea of Justice (2009) some move towards Durkheim's ‘response to the Hobbesian problem of order. [Recognition of the] force of [collective] affective sentiments … the normative glue of the collective’ (2020, 198).

• A more general possible area of relative current CA weakness concerns insufficient ‘treatment of the social construction of meaning’ and, more broadly, the meaning systems which people absorb and which form them as people (Kremakova 2013, 404; cf. Salais 2009). Gangas suggests that Sen has somewhat resisted such a focus, for he fears oversimplification and stereotyping of ‘cultures’ and emphasizes rather that people can and do make their own constructions of value and identity (see, e.g. Sen 2004). Bonvin and Laruffa (2018) stress, however, Sen's interest in the social emergence of public purposes. His reasons for democracy include not only its inherent worth and instrumental roles (it mobilizes information and provides pressures to respect the interests of each participant), but its promotion (potentially) of social learning about one's fellows and of formation of shared purposes (Sen 1999b). People can rise above their original limited positional objectivity. Salais (2009) adds though that Sen's social choice theory approach to public reasoning has lacked but needs a notion of common good. This could help to extend the notion of ‘commitment’, to clearly include Smith's ‘public spirit’ and to apply in all groups within which one is interconnected.

Overall, more appropriate than claiming that all requirements are already fulfilled by Sen, is the response by authors like Caroline Hart (2013). For her work on aspirations and education Hart combined CA with Bourdieu's concepts of habitus, fields of interaction, and multiple forms of capital (social, cultural, symbolic, as well as economic) that affect the attainment of capabilities and functionings. Sen provided her with tools to consider a wide range of human capacities and outcomes; Bourdieu provided tools to investigate the social contexts and processes involved in accessing and progressing in higher education. Ideas of habitus allowed her to discuss social formation of the person, including formation of subconscious assumptions, in her studies on emergence and application of expectations and aspirations and hence of capabilities. Sen is himself clear that he does not offer a total theory and that his ideas can be combined with and enriched by other perspectives.

On power and capitalism

Sen does sometimes discuss power, but rarely power-systems (Hill 2003); and his remarks often suggest a distance from relevant literatures.⁷ He is shrewdly critical of the quasi-religious vacuity in ideas of ‘dictatorship of the proletariat’ and similar utopian proposals which do not find ‘countervailing powers’ necessary (Sen 2019). But many critics (e.g., Dean 2009; Gasper 2009; Harvey 2014) note his relative lack of attention to capitalist power-relations in particular, in contrast to his regular praise of market relations as promoting both opportunity-freedom and process-freedom (Sen 1993, 1999a). Gangas acknowledges that ‘For all its merits, Sen's system operates within the parameters set by current market mechanisms. … Sen's model abstracts from the systemic structure of global capitalism’ (2020, 92). The impacts of capitalist and market relations on the character of people and politics are treated relatively lightly.⁸ The Idea of Justice does not, for example, discuss the degree of acceptability of the conditions under past and present capitalist systems that have moulded the formation of wants and more. Gangas implies though that we should distinguish between, first, gaps in theorizing that could be filled, like this one; second, possible distortions that can be remedied or counteracted; and third, any biases that are irremediable (see also Gasper 2007, 2009).

Bonvin and Laruffa (2018, 223) agree that Sen’s 1993 ode to ‘Markets and Freedoms’ was overly optimistic about the ability of a democracy to incidentally tidy up the social effects of an unconstrained market; ‘the market’ needs to be constitutionally constrained. Using stronger language, capitalism will otherwise dominate the public realm. In his later work, Sen himself has outlined the folly of a ‘I am against poverty, but I am not bothered by inequality’ stance; for severe inequality brings power differentials that distort most other matters (2005b, xvi). Most countries, and the world as a whole, have sufficient capacity to solve problems of basic health, for example, yet they do not do so (xvii).

More usually though, Sen presents a generalized, optimistic notion of what democracy can achieve: ‘public discussion’, leading to ‘a better understanding of the lives of others’ (2009, 344). Elaborating this, Bonvin and Laruffa (2018, 224) claim that compared to markets, which just take preferences as givens (a comment which ignores the immense efforts by capitalist businesses to mould preferences), democracy has a potential for questioning and reforming preferences, for promoting the ‘capacity to aspire’, and for expression of views and interests not only in proportion to purchasing power. Given Sen's intra-establishment ‘gentle persuader’ style and his relative lack of a theory of established power, one can ask how far do high-tone emphases on freedom, democracy and reasoning reflect a too relaxed ‘top-table’ reformism? Or can they represent instead a long-term radical normative programme? Sections 6 and 7 will return to this.

4. Sen's intellectual style

Sen's argumentative style matches his ideal of government by reasoned deliberation: courteous, patient, systematic, interactive. He politely avoids picking out ‘bad guys’ responsible for social ills. Typically he identifies bad ideas as the ‘bad guys’ and expresses a Millian faith in education and in ‘government by discussion’ – the phrase coined by Walter Bagehot in Victorian Britain to express a theme from John Stuart Mill (Duncan 1973; Gasper 2009). Bad ideas include, notably, oversimple conceptualization and neglect of multidimensionality and heterogeneity. We will see these emphases at work in, not least, Sen's theory of personhood (Section 5).

Dissector of concepts

Sen's work, in field after field, nearly always involves patient conceptual clarification, while also stressing the limits of attempted definitions and the need to recognize irremovable degrees of ambiguity. The Argumentative Indian (2005a) dissects many significant terms, such as ‘modernity’, ‘secularism’, ‘Hinduism’, ‘India’, ‘the world’; and ‘identity’, which is investigated further in Identity and Violence (2006). He shows how the concept ‘secularism’, for example, if it means non-discrimination between religions, has very diverse variants according to the degree of closeness or distance, support or indifference, given equally to religions by a State. Ashish Nandy's anti-secularism arguments struggle under such close examination (Sen 2005a, Ch.14). Against any reductionist picture of ‘the’ Hindu tradition, Sen presents its enormous variations, heterogeneity, and recency as a notion. The concept of ‘India’ receives a similar analysis, given the huge variations within Indian thinking, society and culture. Within India, ‘Hindu’ and ‘Muslim’ cultures are profoundly overlapping and interpenetrated; the ‘two nations’ theory, established by Hindutva leaders from the 1930s, was based on comprehensive ignorance of Indian history (Sen 2021). So he rejects ‘The Smallness Thrust Upon Us’ (Sen 2015, Ch.3), both the small picture of India and the small picture of Hinduism. Instead, India is seen as a confluence of many religions and streams.

Likewise, in conceptualizing the world, Sen suggests the fatuity and danger of Samuel Huntington’s (1996) partition of humanity into eight ‘civilizations’ that are presumed to be (i) primarily defined by religious tradition, (ii) fundamentally different, (iii) mutually antagonistic and (iv) unable to effectively communicate and co-deliberate (Sen 2005a, Chs. 8, 13). Partition in terms of religious tradition downgrades non-religious aspects of identity; we share a world civilization, albeit with numerous branches and blends, not a Huntingtonian jungle of separate worlds (Sen 2021, xiv). The future of global cooperation relies on our not seeing people in simplistic identity-boxes and on instead recognizing our multiple commonalities. Huntington's inadequate approach to societies links, Sen considers, to an inadequate approach to conceptualizing persons. Section 5 will consider his own approach.

Explorer of complexity, heterogeneity, intersectionality – and advocate of liberal hope

Sen's insistence on complex conceptualization leads to stresses on heterogeneity, diversity, intersectionality, and on ‘description as choice’, the choices that we must make in description (Sen 1982, Ch.20). Entitlements analysis and capability analysis look not at ‘the social actor/subject’ or even ‘the worker’/'the peasant’ or so on, but at multi-dimensional real people: gendered, having particular ages and professions and histories and (dis)abilities. Sen insists on using multiple descriptive and explanatory dimensions in analysing a society and rejects preoccupation with say economic class alone (2005a, Ch.10). He carefully assesses correlations and interconnections between dimensions of disadvantage but considers them insufficiently strong and invariable to justify focus on just one dimension. Instead, he proposes and undertakes intersectional study of how diverse factors operate jointly, not in isolation.

He stresses the inevitable choices then involved in making descriptions of complex multi-dimensional realities, including choices in categorization and about what issues to include. For example, he does not present his emphasis on the presence of argumentation and reasoning throughout Indian history as being the only relevant picture; instead, he consciously emphasizes this strand because it has been neglected (in part due to British imperial dominance and its manufacture of reductive stereotypes about India) and because it contributes to a basis for peaceful progress, democracy, and secularism (2005a, Preface). Sometimes though he claims more, for example that ‘the Indian subcontinent has a particularly strong tradition in recognizing and pursuing a dialogic commitment’ (2005a, 75). This seems somewhat strained, as if Ashoka and Akbar were typical and as if their approaches had prevailed in India. It illustrates a further characteristic in his work, which we may call (to use a phrase from his friend and uncle by marriage Albert Hirschman) ‘a bias for hope’ (Hirschman 1971). He chooses to emphasize what he considers desirable potentials.

Not everyone has Sen's appetite for conceptual clarification and epistemic refinement. Hence not only is his type of social analysis at risk of being a minority pursuit, but the types of people that we need to theorize about include many who are different from those he imagines, as we will discuss in Sections 5 and 6. His chosen description seems to reflect a normative vision.

5. A programmatic conception of the person

We saw that Sen deconstructs the economics notion of the rational individual. He adds awareness of plurality within the self, including through the idea of metapreferences: preferences about preferences; for example, a preference to not have one's preference to smoke. The idea involves an awareness that people have some capacity to reflect on, reason about, assess and modify their preferences (Davis 2007). Davis identifies similarly in Sen a conception of people as reasoners about their various actual/potential elements of social affiliation and identity, seen also in Sen's concept of ‘commitment’, ‘an act in which individuals freely self-constrain themselves to others’ (Davis 2007, 329). Sen does not, according to Davis (2003, 66), go further to think about the different mental frames and ‘utility functions’ corresponding to a person's different social roles; but he does clearly reject economics’ frequent presumption that people are purely individuals with independently given preferences and have no ‘sense of identity with anyone other than themselves’ or give little attention to it (Sen 2006, 20).

Besides the common economics picture of unsocialized individuals, Sen rejects also an overemphatically sociological picture of totally socialized individuals who bear a single socially given identity. In Identity and Violence (2006) Sen argues against classification of people primarily, even exclusively, in terms of a single social identity, whether within a country or across countries, say as ‘Hindu’ or ‘Indian’. He calls such a picture of ‘singular affiliation’ (2006, 20) the ‘illusion of a unique and choiceless identity’ (2006, xv). In reality people exist as intersections of many cultural streams.

There is no single proper way of grouping people, since we have multiple relevant characteristics, and deciding which ones are most relevant in a specific context must be reasoned out in that context. Sen learns here from his own complex geographical and intellectual life-trajectory (2021). Reflective reasoned choice of which aspects of one's plural identity to highlight in a given context is a key aspect of a human life and human freedom (2006, xiii, 24, 38). Similarly, a real multiculturalism is more than mere coexistence of several (supposedly) fixed monoculturalisms (2006, 156); it involves people's ability to learn and select from multiple cultures, make reasoned informed choices (pp. 114, 150), and live examined lives (p. 160).

Sen calls thus for people to be able to categorize themselves rather than be pigeonholed from birth (2005a, 55). We are not bacteria or ants, near-identical programmed ‘robots’. He admits that choice of identity is constrained but insists that virtually all people have significant degrees of choice (2005a, 351). What many people and some communitarian theorists consider authenticity Sen identifies as failure. His phrase regarding India, that ‘in our heterogeneity and in our openness lies our pride, not our disgrace’ (Sen 2005a, 138), applies also to within each person. For him, community membership is taken as instrumental, not definitional, to being (Gasper 2002, 452).

Sen presents the choices by individuals in characterizing themselves as an exercise in reflection, in the same fashion as choices by analysts in categorizing persons. Yet if many of the different components of identity come as social memberships, then they come with social expectations and pressures and as outcomes of socialization. His picture of self-characterizing individuals may often be more programmatic than descriptive, an exercise in liberal assertion. For he considers ‘pluralism [as] both an irreducible aspect of reality and, as it turns out, a value’ (Gangas 2020, 22).

Writing before Sen's 2005 and 2006 books appeared, Davis thought: ‘Sen does not actually have a theory of the individual’ (2003, 164; emphasis in the original), neither a theory of the social formation of individuals nor a theory of how individuals can become and be independent choosers despite being to a major extent socially embedded. The anthropologist Mary Douglas claimed earlier that Sen adopted a picture ‘that wants emanate from individuals and that basically individuals are the same the world over’ (Douglas et al. 1998, 228). Maniar asserts even now that the ‘Capabilities Approach assumes that subjectivities are already formed and need to be responded to’ (2019, 9). We saw that Sen has in fact long discussed the impact of socialization but also believes in most people's agency capacities, including potentials for critical autonomy.

Since Davis’s 2003 comment above, Sen continued his elaborations, making explicit what he had in mind.⁹ He has, for example, written on his admiration for Adam Smith's Theory of Moral Sentiments (Sen 2009), showing that he aims for a social not asocial psychology. As Holmwood (2013) remarks, though, there is more to be said in social psychology now than Adam Smith could provide. Many authors suggest extensions. Gangas, as we saw already and will discuss further, considers that a compatible framing for Sen's ideas is Parsons’ theory of ‘the action system and its normative ideal of “institutionalized individualism”’ (2016, 30).¹⁰

Davis's proposed enrichment is

that having a personal identity is having a capability, … being able to not be lost in different social relationships is a matter of having capability to move across them and maintain one's sense of self. What is needed is not only that one remains a distinct individual (my individuation criterion) but remains one across changing circumstances (my reidentification criterion). (personal communication, 2 September 2019)

Indeed: ‘the entire capabilities-as-freedoms framework depends on the one central freedom or capability of being able to sustain a personal identity’ (Davis 2007, 331). More fully:

If individuals … are enmeshed in countless cross-cutting social group relationships, one way we may understand individual freedom is in terms of their being able to move back and forth comfortably across these relationships … [in other words, to not be] fragmented and lost amid these relationships, but … sustain an individual identity across them. Sen's idea [is] that individuals develop their capabilities by exercising a freedom to explore possibilities in life … Agency freedom … can accordingly be thought central to understanding the personal identity of socially embedded individuals. [The ability to investigate, span and balance/reconcile their multiple relationships] would identify them as independent individuals. (Davis 2003, 179; my expansion)

Sen's more recent work expresses a similar view.¹¹ Thus, a few years after arguing that Sen lacked a theory of the individual, Davis could now write: ‘Sen claims that individuals are able to make first-person reflexive representations of themselves to counter others’ third-person representations of them’ (2010, 174). This model of personhood in Sen's 2005 and 2006 books is ambitious. In it, people are reflexive and thoughtful about everything, including about which group affiliations and identities they might accept and combine. He says people can or should be able to say who they are, choosing-cum-composing their own identity out of their multitude of characteristics, affiliations, ideals and loyalties. Their distinctiveness as persons arises precisely out of this reflective mix-and-match-ing. Is this conception of personal identity presented as descriptive as well as normative/programmatic?¹² For many, perhaps most, people in most of history and possibly continuing nowadays, the independent sense of self may not be or have been strong enough for Sen's vision to serve as a general description. Certain given identities are declared as unavoidable and overriding. Intense nationalistic affirmations, for example, are not in decline; and the notion that markets transform or divert these or other sectarian passions into mere interests or options that people calmly reason about and balance may not be confirmed by experience.

Sen persuasively advocates a ‘large India’ conception, that does not neglect Buddha, Ashoka, Kabir, Nanak and Akbar, above a ‘small India’ notion that stresses only Rama (Sen 2005a, 75). He presents the ‘small India’ notion as an inferior chosen description, a neglect of relevant facts. But he does not enter the minds of those who affirm ‘small India’. Similarly, his incisive dissection of conceptual oversimplifications in discussions of secularism is not an existential investigation of the angers, fears, experiences, and social manipulation that may drive Indian anti-secularism. Will self-definition and drawing from multiple cultural strands prosper in many environments? Sen replies that numerous people throughout recorded history, and certainly nowadays, have evidently been busy with their own self-definition, for otherwise there would be no ‘need for the policing of adherence and loyalty’ which is so typical in communitarian activism (2015, 50). And the growth of such self-definition is a central feature, both descriptively and normatively, of ‘development’ (2015, 51).¹³ His slogan ‘development as freedom’ can refer then as much to this theme of self-definition as to questions of economic choice or political democracy.

6. Faith in democracy as ‘government by discussion’, in a world of wealth and emotions

Sen's theorizing on democracy reflects the character of his social theorizing and also its gaps. It lacks ‘in-depth discussion of the political and power dynamics of reasoning processes’ (Deneulin and McGregor 2010, 513). His high optimism on democracy reflects a high optimism regarding reasoning self-determining individuals.¹⁴ The distance though between, on one hand, the eloquent rationalism in The Idea of Justice (2009) and in the enlarged edition of Collective Choice and Social Welfare (2017a) and, on the other hand, the screaming and opportunistic worlds of many countries’ contemporary democratic politics and their mainstream media and echo-chamber ‘social’ media is great. Reading Sen leaves us wondering about how to connect ideals and realities and about what causal features separate reality from the picture of the ideal. We can view his theorizing then more as a normative vision than as a full understanding of contemporary societies. For that understanding too, his ideas nevertheless contain valuable potentials, and newer work is developing these. Providing inspiration for audiences outside the seminar room may, however, require further ingredients (Gasper 2009, 2018).

Ideal theory of democracy

Sen itemizes a series of benefits of democracy, including the airing of information and provision of pressure to respect the interests of each citizen; ‘ … democracy is not only a blessing in itself, but can also be the most important means to pursue public ends’ (Sen 2005a, 194). He often uses this same language for discussing a free press, or ‘political voice’, or electoral democracy, or ‘public reason’/‘public reasoning’ in general. He appears optimistic about the degree of convergence that public reasoning can bring. The historical record makes some of the claims appear exaggerated (Shapiro and Hacker-Cordón 1999), for example concerning how far (actual) democracy counteracts market-generated inequalities or regulates centrally imposed declarations of what is culturally appropriate.

However, Sen refers mostly not to mere ballot-box democracy but to orderly reasoned open dialogue, an ideal of ‘government by discussion’. He proposes that ‘the view that democracy is best seen as “government by discussion” has gained widespread support’ (2009, 324) – meaning that this is his and many others’ recommended form, not that such a description fits most actual practice of self-described democracies. He criticized, for example, the Brexit decision-making procedure, in which a vote by a minority of the total electorate after an unmoderated polemical campaign was given more than consultative status, and effectively replaced calm and careful consideration through the legally authorized representative Parliament after full and moderated discussion (Sen 2017b).

So,

for him, democracy is not a matter of registering individual preferences as they are at a specific point in time, but a mechanism to allow the effective integration of all relevant positional objectivities in collective decision-making processes. In the course of such processes, individual preferences may be transformed and revised. (Bonvin and Laruffa 2018, 225)

Bonvin and Laruffa acknowledge, however, that the prerequisites for this deliberative ideal – equal voice and a shared willingness and ability to listen openly and reflect, etcetera – are nearly always absent. James Fishkin's ‘deliberative polling’ exercises are instructive exceptions (e.g., Fishkin 2009; not discussed in Sen 2017a). Just as Sen has sought to reclaim the ‘freedom’ label, so he wishes to retain the kudos of ‘democracy’ for his rationalist ideal. But his real focus is ‘the role of reasoning in social choice’ (2017a, 453), and actually-existing democracies are often distant from, and even antipathetic to, ‘government by discussion’, given our actual contexts of nation, faction, capitalism and class, perhaps even increasingly distant in the era of social media. Should the same term, ‘democracy’, be used for both his ideal and the actuality?

Sen brings a razor-sharp eye to examining notions like ‘recognition’ (2005a, 35) or ‘modernity’. He might not be equally strict for ‘democracy’. His fondness for the ‘government by discussion’ version seems to conduce to sometimes using the term ‘democracy’ as almost automatically good, instrumentally as well as inherently. His response to disappointment with its fruits is then typically to call for ‘a more vigorous practice of democracy’ (2005a, 36). We should consistently distinguish at least Democracy-A (actually existing), including Democracy-B (ballot-box, in many electoral variants), from Democracy-D (deliberative; ‘Democracy as Public Reason’, the title of Ch.15 in Sen 2009) and a Democracy-I which operationalizes Democracy-D plus other dimensions of a normative ideal. Those dimensions include the requisites for open and respectful discussion, such as fulfilment of basic human rights, as specified in the complex methodology of the International Institute of Democracy and Electoral Assistance (www.idea.int; Beetham 1994).

How actual democracies diverge from ‘democracy as public reason’

First, democratic systems may contain many predominantly partisan citizens. Even if these have equal powers, their free political participation brings biases towards those who are in a position to reap great benefit through a particular scheme, for they have sufficient incentive to mobilize and invest to establish and defend it, even when it brings a much larger aggregate loss but one that is spread thinly over numerous other people (e.g., Sen 2017a, 400). Perhaps especially ‘in a poverty stricken and illiterate democracy, the “intermediate regimes” are more alert and effective … to extract “rent”’ (Ghosh 2006, 313). This long-known danger applies in all political systems, but democracy gives it considerable scope and momentum (as well as potentially offering wider channels for resistance). Democracy theorists call then for highly socially responsible citizens. ‘Commitment’, including public spirit, is not an optional extra but instead a necessary foundation of well-functioning democracies, and indeed of healthy personalities.

[T]he health and stability of a modern democracy depends, not only on the justice of its basic institutions, but also on the qualities and attitudes of its citizens: e.g. their sense of identity and how they view potentially competing forms of national, regional, ethnic, or religious identities; their ability to tolerate and work together with others who are different from themselves; their desire to participate in the political process in order to promote the public good and hold political authorities accountable; their willingness to show self-restraint and exercise personal responsibility in their economic demands, and in personal choices which affect their health and the environment. Without citizens who possess these qualities, democracies become difficult to govern, even unstable. (Kymlicka 2002, 285)

Ghosh's review of The Argumentative Indian concludes, sadly, that Sen provides:

[no] answers to the questions: Why contemporary [Indians] are so deceitful in the land of the Buddha, so immodest in the land of Ashoka, Kabir, Guru Nanak, Sri Chaitanya and Ramakrishna, so unlawful in the land of Chanakya, so greedy in the land of Harsha, so intolerant in the land of Akbar, so ignorant in the land of Tagore and Amartya, so violent in the land of Gandhi, so indisciplined in the land of Vivekananda, Netaji and Patel and so retrogressive in the land of Nehru. (Ghosh 2006, 315)

Conceivably those historic figures emerged precisely in reaction to longstanding prevalence of the problems mentioned. ‘Commitment’, concern for the situation of others, can arise in powerful negative forms like the widespread conscious exclusion of Dalits.

Second, citizens have in reality very far from equal powers, and democratic systems do not exist in isolation. Sen speaks for a public deliberative space equally accessible to all, but this has never existed, not least under capitalism, and the space and access have been reduced further under neoliberalism, argue Dean (2009) and Nancy Fraser (1997).

Third, partly reflecting the sharply unequal contexts around democratic systems,

[any] consensual agreements achieved in the process of public deliberation – whether in the course of participative poverty assessments or through citizens’ juries or focus groups – may elide fundamental conflicts and hidden oppression. They may do nothing more than reflect prevailing hegemonic assumptions. (Dean 2009, 270)

For example, throughout three generations of democratic India the balance of public policy has never been pro-poor. The state of ‘government by discussion’ in contemporary India, the largest democracy, in the USA, the second largest, or in the UK with its ‘Mother of Parliaments’, is disturbing in many respects. Sen finds baffling the widespread preoccupation in the USA with the supernatural, in the world's richest country (2021, 28), but draws no implications for his Millian faith in democracy.

Democracy in India – Sen’s unwilling hero

Sen is keenly aware that Indian democracy (and Indian media) has over 75 years shown relatively little interest in hunger or mass education or mass morbidity (e.g. 2017a, 403 ff.). His ‘no famines in a democracy’ hypothesis does not appear transferable beyond famines. India's extreme inequality, including in education and health, seems to undermine its democratic forms (Gasper 2018). For famines too, the thesis has had to be refined from ‘no famine in a democracy’ to ‘no famine in “a functioning democracy with regular elections, opposition parties, basic freedom of speech and relatively free media”’ (Sen 2009, 342). Even that description of functioning democracy still matches some famine cases.¹⁵ Still, the theme remains important even if the thesis correctly concerns ‘less’ not ‘no’ famine.

The key required mechanism is not mere balloting but democracy's ‘ability to make people take an interest, through public discussion, in each other's predicaments, and to have a better understanding of the lives of others’ (Sen 2009, 344). But as he knows, ‘ability’ means here only a potential. He has explored the paradox of an India with vast food surpluses and yet vast malnutrition, and noted ‘how little public attention it gets, when it gets any at all’ (Sen 2005a, 212–213). Instead, old myths prevail, that since ‘we’ have plenty of food available, the supposedly idle poor must be failing to work. He holds that rich farmers have dominated the democratic polity to obtain high food prices (2005a, 214–215; 2015) and that in response we need more democracy. By that he intends more involvement of well-informed, well-mobilized poorer groups, capable of acting and unafraid to act. But what issues mobilize the poor in reality? His other proposed remedy for failings of democracy is more education, but India's democratic political system has produced too little and too weak education for the poor.

Sen's belief in dialogical reasoned persuasion matches his oft-repeated call for case-by-case ‘critical scrutiny’ not general slogans or even generalized rules. The ideal reasoner proposes ‘let's reason’ for each specific case, more than ‘let's institution-build’ in the form of pre-set fixed guarantees. Yet countering institutionalized injustices requires institutionalized forms for justice. Invariable insistence on situational reasoning can reflect a socio-political naivete. To leave the operationalization of justice to be discussed afresh in each situation, without constitutionally fixed prioritizations, leaves too much power to the powerful, argues Nussbaum (2003; cf. Gasper 2007). While Sen has called Ambedkar, the Dalit father of India's Constitution, the ‘father of my economics’, his writings on justice give far more prominence to Smith and Mill than to Ambedkar.

Ambedkar knew that the Constitution would be hobbled if not accompanied by a battle against caste: disgust-based social hierarchy. Sen illuminates human complexity in many respects but may understate it in some others and thus under-specify human personhood and potentials. He lacks theories of the demos, of the crowd, and of populism (cf. Canetti 1962). His essay on the national trait of preoccupation with ‘first in class’ (‘The Country of First Boys’, in his 2015 book of that name) does not situate it within a perspective on hierarchy and caste. In his and Drèze's conspectus of contemporary India, An Uncertain Glory, ‘the agents are not strongly highlighted. We read a thoughtful diagnosis of mistakes, oversights, blind spots; but less about the agents who commit them, and their passions and perceptions, likes and dislikes’ (Gasper 2018, 284). The book notes too that ‘we know very little about which institutions matter’ for development (Drèze and Sen 2013, 36); and this remark might apply also to Western forms of political democracy. Drèze and Sen acknowledge the insufficiency of India's Democracy-A: the ‘strongly incriminating evidence against taking Indian democracy to be adequately successful in consequential terms’ (Drèze and Sen 2013, 244). Nor has Democracy-D reached far, including during decades of Congress rule; public discussions in India, the book showed, have been completely dominated by affluent groups and their concerns. Indian mass-media provide little coverage of the lives of half the country, while ‘the relatively privileged seem to have created a social [and mental] universe of their own’ (Drèze and Sen 2013, 268).

Drèze and Sen present injustice, corruption and lack of dignity as problems to be calmly analysed and counteracted. Persuasive power in actually existing democracies seems, however, often to come less through ‘critical scrutiny’ and appeals to a broad ideal of freedom than through mobilizing ‘myths’ about an imaginable desirable attainable future, anger-inducing enemies, and inspirational heroes to be loyal to (Gasper 2018, 285).

Mobilizing capability theory to better understand democracy

Bonvin and Laruffa (2018, 227) argue that:

in contrast to the idealistic approach to deliberative democracy, the CA has great potential for developing a more realistic analysis of democratic processes – one that takes into account the teachings of sociological studies (Bonvin, Laruffa, and Rosenstein 2018) … [First] The purpose is not to reach a view from nowhere where citizens are called to put aside their positions and interests, but to let situated views communicate and debate. … .

Second, the CA apparatus – its ‘series of items suited for empirical inquiry – opportunities, resources, entitlements, conversion factors, achievements’ (Zimmermann 2018, 284) – can help in understanding the limits of formal democracy in converting formal rights into equal participation. It ‘allows a sociologically grounded understanding of how individual agency and social structures interact in the formation of preferences’ (Bonvin and Laruffa 2018, 228). The ‘allows’ here reflects the need to build on but extend Sen's thinking, such as the ‘adaptive preference’ idea, to identify determinants of people's aspirations and their capacity to speak.

Bonvin and Laruffa connect the requirements for effective democracy to Appadurai's ‘capacity to aspire’ (rendered as ‘ability to form one's preferences’; Bonvin and Laruffa 2018, 222) and to ‘capability for voice’, the ability to speak and be heard. They focus on incremental improvements from wherever one starts, not with defining an ideal.

… this specific understanding of democracy departs [i.e. diverges] both from market mechanisms and from ideal conceptions of deliberative democracy. … it develops rigorous empirical tools for the sociological investigation of democratic processes and how these are impacted by material, symbolic and deliberative inequalities. In our view, the CA's main contribution [here] is to shed light on … how preferences are formed (in connection with the notion of capacity to aspire) and how they are then called to confront and coordinate themselves when collective decision-making takes place (this calls for examining the differential degree of capability for voice enjoyed by the various stakeholders). As such, the CA opens up a new field of research for a more appropriate sociological understanding of democratic processes. (Bonvin and Laruffa 2018, 230)

Thus, Sen's work could contribute not only as a normative vision but towards rendering such a vision more realistic, practical and influential. Capability approach concepts might help to structure a workable research programme, usefully refined but not too abstruse, hence perhaps able to mobilize interest from contributors across the social sciences. Much must be added, regarding the formation of selves, crowd psychology, and the power systems which can drive identity formation in ways far different from the reflective, sophisticated, open-minded ideal articulated by Sen. ‘ … the reasoned deliberation that Sen is looking for requires much more than reason’ (Giri 2002, 239).

7. Conclusion

Sen became famous outside economics through his book Poverty and Famines, which analysed the neglect by British colonial rulers in 1940s wartime Bengal that led to millions of deaths. The analysis was more than an exposition of the social mechanics of loss of acquisition-power; it was a demonstration of ethically better alternatives. In later work he stressed how, at exactly the same time, powerholders in Britain itself implemented such alternatives in regard to food distribution, and how this contributed to a social revolution of increased democracy, cooperation and public spirit (e.g., 2021, ch.25). Sen's social theorizing can be seen as aiming to illuminate and promote this sort of social and political change.

His work is an outlier in terms of social theory and has limitations, but has major strengths too. His main sources in social theorizing have been from pre-modern and early-modern eras. Davis (2003, 166) argues that the picture of the timeless reasoning individual found in, for example, Descartes and Locke antedates industrialization and the corresponding transformations of society and increased perception of persons and society as transformable in time; and he presents the challenge for the modern social sciences as how to understand the socially embedded individual not Descartes and Locke's disembedded reasoner (Davis 2003, 182–183). So, while Sen proceeds from the intellectual legacies of Smith and Mill (and Dewey and Tagore), and adds various tools, he leaves much scope for combining with the insights of others. He transcends homo economicus and the mutilated loners assumed in ‘the so-called Rational Choice Theory’ (Sen 2007, 342); he invokes human proclivities to sympathy and social commitment, abilities ‘to reason, argue, disagree and concur’ (2009, 415), and desires for freedom, affiliation and identity. Given though his relative neglect of modern psychology, sociology and anthropology, we need to add in many areas.

Gangas’ book (2020) on Sociological Theory and the Capability Approach provides a major example of taking up this challenge. It warns of ‘underestimation of social structures and institutions by CA’, with specific reference to Sen, and the consequent ‘risks of getting trapped into … an overestimation of its normative potential’ (219, 259). The book seeks to connect CA to high social theory, including Weber and especially Durkheim and their synthesizer Parsons, as well as to the philosophies of Kant and Hegel that nourished their thinking and to many recent sociological theorists. Gangas argues that Parsons in particular provides a relevant social theory basis. Parsons reflected on the capacities required for the social actor to participate as a full citizen; his conception of ‘institutionalized individualism’ ‘offers a template for a theory of social institutions and role-complexes, upon which the capable actor's goal orientations are normatively grounded and evaluatively sanctioned’ (Gangas 2020, 124). Parsons’ concept of capacities was not, however, as helpfully concrete and operational as Sen's capability notion. Sen's work offers sociology ‘a resourceful conception of agency aligned to a vision of a good society’ (Gangas 2020, 111). Parsons, Etzioni, Giddens and others were close to such a conception of agency, and Durkheim had made clear the importance of such visions in cohesive societies and also in social science. We study society in large part because we consider human life valuable and improvable, and we must then pay attention to what improvement means. The capability approach – or, more fully specified, a ‘human development approach’ (Nussbaum 2011) – offers, Gangas suggests, one normative framework for a normatively deepened sociology. Overall, using the CA can help one to ask about social structures and how they influence what people can do; and sociological theories can enrich ‘visualizing how the capable self's choices and freedoms are supported [and constrained] by social institutions’ (2020, 10; my addition). Such a manifesto leads beyond Sen's own work and gives an agenda for research oriented to test how far that can be taken.

Sen has not essayed a natural science of society. From his emphases on choice, reasoning, freedom and improvement, we can see his essays in social theory as attempts to encourage better reasoning and better social choices, rather than to describe present-day central tendencies. His highly optimistic pictures of more reflective, more open, more creative human personhood and human polities might be seen as exercises in ‘description as choice’: highlighting possibilities that he considers desirable and real even if they are, undoubtedly, far from presently predominant.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Des Gasper

Des Gasper is professor emeritus of Human Development, Development Ethics and Public Policy at the International Institute of Social Studies (The Hague), Erasmus University Rotterdam, Netherlands. His publications include The Ethics of Development: From Economism to Human Development (Edinburgh University Press 2004; SAGE India 2005) and Development Ethics (International Library of Essays in Public and Professional Ethics, Ashgate 2010 and Routledge 2016, co-edited with Asuncion Lera St. Clair).

Notes

1 Similarly, Sen's remarks on needs theory (e.g. 2013) do not connect to the needs literatures in psychology and social policy.

2 Here and often elsewhere Sen says more about the powerless than about powerholders and power-systems.

3 Zimmermann (2021) argues that pragmatist sociology is an equally important required base.

4 Slightly more frequently mentioned in The Argumentative Indian are a few social scientists whom Sen criticizes, such as Ashish Nandy and Samuel Huntington. His papers on social exclusion (Sen 2000) and culture in development (Sen 2004) are wider-ranging but still oriented more to institutional economics than to ‘social theory’.

5 Ragkousis (2023) questionably contends that Sen's extensive use of formal reasoning is in tension with this freedom- and learning-oriented perspective.

6 The distinguished Indian socio-economist C.T. Kurien contrasted his own careful specification of institutional context, as a precondition for worthwhile economic theorizing, with Sen's more abstracted approach (Kurien 1996). Sen defends

[a] general approach [that] can be used in many different ways, depending on the context and the information that is available. It is this combination of foundational analysis and pragmatic use that gives the capability [and entitlements] approach its extensive reach. (Sen 1999a, 86; my addition)

His case studies of hunger and food policy, for example, contain far more institutional specification than does his broad theorizing.

7 An example: ‘Farmer points to what he calls “structural violence”’, writes Sen (2005b, xiii), using quotation marks for the term coined by Johan Galtung back in the 1960s; and he continues to make this attribution (‘ … Farmer's notion of “structural violence”’, p. xv).

8 In contrast, Charles Lindblom, theorist of The Intelligence of Democracy (1965) in terms very similar to Sen's, moved on to consider also the profoundly undemocratic corporate and plutocratic capture of politics, in works like The Market System (2001).

9 See, e.g., remarks already in his 1988 interview with Richard Swedberg (Sen 1990, 260), on personal identity as a plurality of memberships.

10 Citing Parsons and Smelser (1956, 49).

11 Whereas Davis engages in depth with social theories (notably, about collective intentionality) in order to do so, Sen does not. His work is thus extended by Davis's.

12 Similarly, in ‘The Smallness Thrust Upon Us’ Sen writes that in ‘determining the relative importance of [our] diverse diversities, and … understanding the priorities between them … [t]hese choices cannot be settled – as some communitarians have claimed – as a matter of passive “discovery”’ (2015, 45). Is the ‘cannot’ here descriptive as well as normative?

13 Zimmermann (2021) traces this view back to John Dewey (‘for Dewey: being human means the capability to develop one's individuality’, p. 182), just as she traces Sen's usage of ‘capability’ to Dewey (Zimmermann 2018).

14 ‘ … I do not believe that, in general, dissociation of choice from reasoning is a sweeping characteristic of the world in which we live’ (Sen 2007, 343). This standpoint seems distant from much in modern psychology, neuroscience and behavioural research.

15 See Currie (2000), Banik (2007); also Sen's discussion of the role of ‘the local Bengal government’ (2009, 340) and the elected Bengal Assembly during the 1943 Bengal famine.