by Amartya Sen
There is a similar distinction within ethical and political evaluation, comparable to the contrast between the illuminating and diverting roles of positionality, respectively. In pursuing theories of relation-based personal responsibilities that demand a special role, for example, of parents in looking after their own children, giving special prominence to the interests of one’s own children may be plausibly seen to be ethically appropriate. To take an asymmetric interest in the lives of one’s own children may not be, in that context, a subjective folly – rather the reflection of an ethical perspective that is objectively sought (which is linked, in this case, with the positional relevance of parenthood).*
In that framework, there would therefore be something of a lacuna in thinking of ethical objectivity only in terms of ‘the view from nowhere’, rather than ‘from a delineated somewhere’. There can be special relevance in positional features that a capacious ethics must adequately acknowledge and appropriately take into account. Indeed, one’s duty to one’s children, to consider the same example again, does not have its source just ‘in the mind’, and real significance may be attached to it in specific approaches to ethics.
When questions of agent-relative – more generally, position-relative
* The relevance of personal connections and relationships is a subject of considerable importance and complexity in the discipline of moral evaluation. Bernard Williams has discussed many of the underlying issues with force and clarity, particularly – but not exclusively – in his critique of utilitarianism; see his ‘A Critique of Utilitarianism’, in J. J. C. Smart and B. Williams, Utilitarianism: For and Against (Cambridge: Cambridge University Press, 1973), and Moral Luck: Philosophical Papers, 1973–1980
(Cambridge: Cambridge University Press, 1981), especially the essay entitled ‘Persons, Character and Morality’.
160
p o s i t i o n , r e l e va n c e a n d i l l u s i o n
– assessments and responsibilities are examined, as they will be in Chapter 10, the illuminating aspects of positional objectivity will be of relevance. And yet in other contexts, giving such a special prominence to the interests of one’s own children must be seen, from the perspective of a non-relational ethics, as a clear mistake. For example, if a public official in pursuit of his civil duties gives greater importance to the interests of his own children, that could be seen as a political or ethical failure, notwithstanding the fact that his children’s interests would be more prominent to him, thanks to their positional closeness.
What may be needed in this exercise is a ‘positionally unbiased’
approach. The demand in this case would be for an adequate recognition of the fact that other children may have similarly large and important interests at stake as one’s own children, and the view from
‘a delineated somewhere’ (linked for example with parental relations) would be, in that context, a mistake.
The search for some kind of position-independent understanding of the world is central to the ethical illumination that may be sought in a non-relational approach. When Mary Wollstonecraft pilloried Edmund Burke for his support of the American Revolution without taking any interest in the status of the slaves, as if the freedom that he supported for white American people need not apply to its black slaves (as was discussed in Chapter 5), Wollstonecraft was arguing for a universalist perspective that would overcome positional prejudice and sectional favouritism. The point there is not positional comprehension, but some kind of a transpositional understanding. Taking a ‘view from nowhere’ would obviously be the appropriate idea in that context.
o b j e c t i v e i l l u s i o n s a n d
p o s i t i o n a l o b j e c t i v i t y
Even when a position-independent view is appropriate for an epistemological, ethical or political assessment, the reality of position dependence of observations may have to be taken into account in explaining the difficulty of achieving a positionally unbiased comprehension. The hold of positional perspectives can have an important 161
t h e i d e a o f j u s t i c e role in making it hard for people to transcend their positionally limited visions. For example, in a society that has a long-established tradition of relegating women to a subordinate position, the cultural norm of focusing on some alleged features of women’s supposed inferiority may be so strong that it may require considerable independence of mind to interpret those features differently. If there are, for instance, very few women scientists in a society that does not encourage women to study science, the observed feature of paucity of successful women scientists may itself serve as a barrier to understanding that women may be really just as good at science, and that even with the same native talents and aptitudes to pursue the subject, women may rarely excel in science precisely because of a lack of opportunity or encouragement to undertake the appropriate education.
The observation that there are few women scientists in a particular society may not be at all mistaken, even when the conclusion that women are no good at science – when drawn from that positional observation – would be entirely erroneous. The need for going beyond the positionality of local observations within societies with entrenched discrimination can be very strong here. Observations from other societies where women have more opportunities could confirm that women have the ability to do just as well as men in the pursuit of science, given the necessary opportunities and facilities. The argument relates here to the case for ‘open impartiality’, invoking such ideas as Adam Smith’s methodological device of the impartial spectator, seeking perspectives from far as well as near.*
When the confines of local beliefs are strong and difficult to overcome, there can be a steadfast refusal to see that a real inequity is involved in the way women are treated in their own society, and many women are themselves led to a belief about women’s alleged intellectual inferiority based on the supposed ‘evidence of the eyes’, drawing on a faulty reading of local observations within a stratified society. In explaining the protest-free tolerance of social asymmetry and discrimination that can be seen in many traditionalist societies, the idea of positional objectivity has something of a scientific contribution to make, in giving us an insight into the genesis of an illegiti-
* See the discussion in Chapter 6, ‘Closed and Open Impartiality’.
162
p o s i t i o n , r e l e va n c e a n d i l l u s i o n mate application of positional comprehension (when the need is for a transpositional understanding).
The important notion of ‘objective illusion’, used in Marxian philosophy, can also be helpfully interpreted in terms of positional objectivity.* An objective illusion, thus interpreted, is a positionally objective belief that is, in fact, mistaken in terms of transpositional scrutiny. The concept of an objective illusion invokes both the idea of positionally objective belief, and the transpositional diagnosis that this belief is, in fact, mistaken. In the example involving the relative sizes of the sun and the moon, the similarity of their appearances (positionally objective as this is from here on earth) can lead – in the absence of other information and ideas and the opportunity for critical scrutiny – to a positional ‘understanding’ of the similarity of their
‘actual sizes’ (for example in terms of the time that would be taken to go around them respectively). The falsity of that belief would, then, be an illustration of an objective illusion.
There is an interesting discussion by G. A. Cohen in his book, Karl Marx’s Theory of History: A Defence, about the idea of objective illusion in Marxian theory:
For Marx the senses mislead us with respect to the constitution of the air and the movements of heavenly bodies. Yet a person who managed through breathing to detect different components in the air would have a nose that did not function as healthy human noses do. And a person who sincerely claimed to perceive a stationary sun and a rotating earth would be suffering from some disorder of vision, or motor control. Perceiving the air as elementary and the sun as in motion are experiences mor
e akin to seeing mirages than to having hallucinations. For if a man does not see a mirage under the appropriate conditions, there is something wrong with his vision. His eyes have failed to register the play of lights in the distance.4
* The concept of objective illusion figures in Marx’s economic writings (not just in the more philosophical ones), including Capital, Volume I, and Theories of Surplus Value. Marx was particularly concerned with showing that the common belief about the fairness of exchange in the labour market is, in fact, illusory, and yet that claim to fairness is ‘objectively’ accepted by people who see how things exchange in terms of equal values at market prices. Even the exploited workers who, in Marxian analysis, are robbed of part of the value of their products, might find it hard to see that there was anything other than exchange of ‘equal values’ in the labour market.
163
t h e i d e a o f j u s t i c e Here the observations, which are taken to be objective, relate to the positional features of breathing the air with a normal nose, seeing the sun with normal eyes, observing the play of light in the distance with a normal vision and so on. And these observed features are indeed positionally objective, though misleading or mistaken in terms of other – contextually more compelling – criteria of truth that can be invoked once we go beyond positional perspectives.
h e a l t h , m o r b i d i t y a n d
p o s i t i o n a l va r i a t i o n s
Marx’s own use of the idea of objective illusion was primarily in the context of class analysis and it led him to his investigation of what he called ‘false consciousness’. A very different type of example concerns the self-perception of morbidity, and this can be particularly important in analysing the health situation in developing economies. For example, among the Indian states, Kerala has by a big margin the longest life expectancy at birth (higher than China and closer to Europe), and professional medical assessment gives much evidence of Kerala’s successful health transition. And yet in the surveys of self-perceived morbidity rates, Kerala also reports by far the highest rates of self-perceived morbidity (both on the average and in terms of age-specific rates). At the other end are states like Bihar and Uttar Pradesh with very low life expectancy and excep-tionally high age-specific mortality rates (and little evidence of any health transition), and yet astonishingly low rates of self-assessed morbidity. If the medical evidence and the testimony of mortality rates are accepted (and there are no particular reasons to rule them out), then the picture of relative morbidity rates as given by self-assessment must be taken to be erroneous – or at least highly problematic.
Nevertheless, it would be odd to dismiss these self-assessed morbidity rates as simply accidental errors, or as results of individual subjectivism. Why is there such a systematic pattern of dissonance between mortality rates and self-perceived morbidity rates? The concept of objective illusion is helpful here. The population of Kerala has 164
p o s i t i o n , r e l e va n c e a n d i l l u s i o n a remarkably higher rate of literacy (including female literacy) than the rest of India, and also has much more extensive public health services. Thus in Kerala there is a much greater awareness of possible illnesses and of the need to seek medical remedies and to undertake preventive measures. The very ideas and actions that help to reduce actual morbidity and mortality in Kerala also have the effect of height-ening the awareness of ailments. At the other end, the populations of Uttar Pradesh or Bihar, with less literacy and education, and severely undersupplied with public health facilities, have less discernment of possible illnesses. This makes the health conditions and life expectancy much worse in these states, but it also makes the awareness of morbidity more limited there than in Kerala.
The illusion of low morbidity in the socially backward states in India does indeed have an objective – a positionally objective – basis, for a population with limited school education and medical experience.* The positional objectivity of these parochially mistaken diagnoses commands attention, and social scientists can hardly dismiss them as simply subjective and capricious. But nor can these self-perceptions be taken to be accurate reflections of health and illness in an appropriate transpositional understanding.
The possibility and frequency of this type of objective illusion have some far-reaching implications for the way comparative medical and health statistics are currently presented by national and international organizations. The comparative data on self-reporting of illness and
* The empirical work on this has been based substantially on Indian data and their interpretation; see the discussion, and the extensive literature cited, in my joint books with Jean Drèze, India: Economic Development and Social Opportunity (Delhi and Oxford: Oxford University Press, 1995), and India: Development and Participation (Delhi and Oxford: Oxford University Press, 2002). However, there is empirical information from elsewhere in the developing world that fits broadly into this reading; see my Development as Freedom (New York: Knopf, and Oxford: Oxford University Press, 1999), Chapter 4. This line of explanation is reinforced by comparisons of self-assessed morbidity rates in the USA with those in India (including Kerala). In disease by disease comparison, it turns out that while Kerala has much higher self-assessed rates for most illnesses than the rest of India, the United States has even higher rates for the same illnesses. On this, see Lincoln Chen and Christopher Murray, ‘Understanding Morbidity Change’, Population and Development Review, 18 (September 1992).
165
t h e i d e a o f j u s t i c e the seeking of medical attention call for critical scrutiny, taking serious note of positional perspectives.*
g e n d e r d i s c r i m i n a t i o n a n d p o s i t i o n a l i l l u s i o n s
Another interesting case relates to the dissonance between the ranking of perceived morbidity and that of observed mortality of men and women. Women have, on the whole, tended to have survival disadvantages compared to men in India (as in many other countries in Asia and North Africa, such as China, Pakistan, Iran or Egypt). The mortality rates have typically been, until very recently, higher for women of all age groups (after a short neo-natal period of some months) up to the ages of thirty-five to forty years, contrary to what one would expect biologically, given the medical evidence for lower age-specific mortality rates for women than for men, when they receive symmetric care.†
Despite the relative disadvantage in mortality rates, the self-perceived morbidity rates of women in India are often no higher –
sometimes much lower – than those of men. This seems to relate to women’s deprivation in education, and also to the social tendency to see gender disparity as a ‘normal’ phenomenon.‡ Happily (I use the expression here in a way that utilitarians might not approve of),
* I shall not pursue further this important practical issue; see, however, my book, Development as Freedom (New York: Knopf, and Oxford: Clarendon Press, 1999), Chapter 4.
† The life expectancy of women in India has recently overtaken that of men, but the ratio of women’s to men’s longevity in India is still considerably below what can be expected under symmetric care. Kerala is an exception in this respect too, with female life expectancy substantially higher than men’s (with a ratio similar to that in Europe and America).
‡ On an earlier occasion, I have discussed the remarkable fact that in a study of post-famine Bengal in 1944, widows had hardly reported any incidence of being in
‘indifferent health’, whereas widowers complained massively about just that (see my Commodities and Capabilities (Amsterdam: North-Holland, 1985), Appendix B). On related issues, see my Resources, Values and Development (Cambridge, MA: Harvard University Press, 1984), and also, jointly with Jocelyn Kynch, ‘Indian Women: Well-being and Survival’, Cambridge Journal of Economics, 7 (1983).
166
p o s i t i o n , r e l e va n c e a n d i l l u s i o n women’s unhappiness about their health has systematically increased across the country, which indicates a declining hold of positionally confined perception of
good and bad health. It is interesting to see that as the subject of women’s deprivation has become politicized (including by women’s organizations), the biases in the perception of the deprivation of women has become less common. A better understanding of the nature of the problem and illusions about women’s health has, in fact, substantially contributed to the reduction (and in many regions of India, elimination) of sex bias in mortality.*
The idea of positional objectivity is particularly crucial in understanding gender inequality in general. The working of families involves some conflict as well as some congruence of interest in the division of benefits and chores (a feature of group relations that can be called
‘cooperative conflict’), but the demands of harmonious family living require that the conflicting aspects be resolved implicitly, rather than through explicit bargaining. Dwelling on such conflicts would generally be seen as abnormal behaviour. As a result, customary patterns of conduct are simply taken as legitimate and even reasonable, and in most parts of the world there is a shared tendency not to notice the systematic deprivation of females vis-à-vis males in one field or another.
p o s i t i o n a l i t y a n d t h e t h e o r y o f
j u s t i c e
This issue is quite important for the formulation of a theory of justice and, more specifically, for exploring a theory that gives a special role to public reasoning in the understanding of the demands of justice.
* The phenomenon of ‘missing women’ in India, China and many other countries in the world, reflecting the number of women who are ‘not there’, compared with what could be expected in the absence of any gender bias, should have gone down sharply as a result of fairly widespread progress in the world in the reduction of sex bias in mortality. Unfortunately, the relatively new phenomenon of gender bias in natality (through sex-selective abortion of female foetuses) has worked in the opposite direction. The changing picture is discussed in two articles of mine: ‘Missing Women’, The British Medical Journal, 304 (March 1992), and ‘Missing Women Revisited’, British Medical Journal, 327 (December 2003).