On Relational Injustice: Could Colonialism Have Been Wrong Even if it Had Introduced More Benefits than Harms?
University of Oxford
A certain objection to the view that colonialism is and was morally problematic is that it has introduced more benefits than harms to the populations that have undergone it. This article sets aside the empirical question – that is, of interrogating whether colonialism did bring more benefits than harms; instead, it argues that historical instances of colonialism were wrong even if they had in fact brought net-positive aggregate consequences to the colonised populations. In arguing this, I develop and substantiate a new concept of relational injustice in describing the unique nature of inegalitarian, subjugative relationship defining the interaction between perpetrators and victims in colonialism. Given that moral relations cannot be reduced into the welfare of their respective individual agents, it is hence the case that incidental, unintended gains in individual welfare neither adequately compensate for nor at all rectify the initial relational injustice. There are three objections that are discussed and rejected, such as: i) the purported irrationality in individuals regretting events that left them better-off on aggregate, ii) individuals can opt to waive being in just and equal relations with others in exchange for individual gains, and iii) the advanced account is self-defeating, because it nullifies the possibility for adequate compensation.
Recent debates over the legacy of colonialism—such as that of the British Empire—have often been centered around whether members of colonies have, on balance, benefited from being subject to colonial rule. Such debates are not only epistemically challenging, since they require speculation about how things would have turned out in the absence of colonialism; they also neglect the possibility that colonial projects could have been wrong independent of the harms they bring.
My thesis is that the relational injustice perpetuated under colonialism enacted unoffset wrongs, such that colonialism was wrong even in cases where it introduced counterfactual-comparative benefits. I will first discuss my concept of relational injustice, prior to establishing the empirical premise and explaining why such wrongs are unoffset by consequentialist gains.
Harm is often employed in a counterfactual-comparative manner: A’s treatment of B harms B if and only if B has lower welfare in the world in which that treatment occurs than in the closest possible world without that treatment. I propose, however, that B can be wronged by A’s treatment even when no such counterfactual-comparative harm occurs. On the understanding of wronging that I adopt, A’s treatment of B wrongs B if and only if B is justified in holding resentful reactive attitudes towards A in relation to that treatment (Strawson, 1974). I say that there is unoffset wrongness in A’s treatment of B when it is reasonable for B to hold negative reactive attitudes towards A in relation to A’s treatment of B even though this treatment did not harm B in the counterfactual-comparative sense; B is not worse off—and may in fact be better off—with the treatment than in the closest possible world without it.
I hereby propose a new concept of ‘relational injustice’, which refers to the specific injustice perpetuated when an individual is placed in an unjustifiably lower status in relation to another within a relationship. Anderson (1999) discusses the concept of relational inequality, which involves status disparities that prevent individuals from relating to each other as equals within communities. Relational injustice is a particular form of inequality, which involves two additional features: i) a group characteristic-based form of prejudice1 towards particular demographic groups that is ii) institutionalised through formal structures, such as government or the civil service. At its very core, relational injustice measures the quality of relationships between individuals. Relational injustice differs from distributive injustice, in that its manifestations—biases, psychological exclusion, and imposed deprecation of status—cannot be rectified even in a society where individuals have equal levels of welfare, or access to welfare.
Consider now the specific empirical premise, that colonialism did involve a violation of relational equality, and thus constitutes empirically an event of relational injustice. The significance of this section is as follows: if it can be established that relational injustice has been committed under colonialism, then it follows that colonial projects can be wrongful even if not harmful. Whilst African slavery had existed prior to European colonisation, the process of Western colonisation was embedded within large-scale Transatlantic Slavery that led to the non-consensual and dehumanising transfer of 11 million Africans to the Caribbean and Americas (King, 2010, pg.24). Colonies were governed by structures beholden (by definition) to either their original sovereign state (e.g. the UK, Spain, or France), or a newly emerged substitute (e.g. the US, Rhodesia); the colonised public in these areas were systematically ascribed a lower status than a select group of elites (Belich, 2009, p. 573). Moreover, indigenous populations faced psychological exploitation (from being forced to internalise racialised and deeply bigoted tropes about themselves, to navigating coerced family breakups and emotional torment and humiliation at hands of violent invaders) and social exclusion from elites who entrenched foreign interests and designated them the effective ‘Other’ in governance (Fanon, 1961). The apparent caveat is that not all colonial projects exemplified these phenomena equally, if at all; yet to the extent that they did, they were relationally unjust.
Relational Injustice and Wrongness
Why is relational injustice typically wrongful? My view is that there is something intrinsically valuable in equal relations between persons within a particular society. Suppose we remove all specifications of characteristics and features about individuals within a hypothetical society, and are asked to choose between a world where all relations are deeply egalitarian and grounded upon mutual respect and compassion, and a world where all relations are inegalitarian, with a clearly arranged status order and hierarchy. Now suppose that both societies have achieved the same, optimal state from the point of view of distributive justice. We would intuitively find the former more appealing. I suggest that this is because relational equality is good in itself.
A critic would rightly observe that our intuitions here are underdetermining—it is not clear if we find the unequal world less appealing because of the impersonal value we find in relational equality, or because of the wrongfulness in unequal relations. I suggest that we need not pick either of them—we could find the unequal world less appealing due to both impersonal value-centric and wrongfulness-centric considerations.
So why does such injustice wrong a particular individual, independent of specific impacts on their individual welfare? Could an individual be wronged in such a way that cannot be reduced to a loss of welfare? A potential justification is to view relational equality as the component of a universal claim-right held by all people—all individuals are entitled to being treated equally in their relations, independent of the outcomes associated with such treatment. The placement of an individual within a network of relations that treats them as if they were unequal violates their fundamental claim to being treated as a moral and social equal, given their possession of the prerequisite conditions that render them morally respectable agents: the ability to reason and (self)-consciousness. For more on this argument, see Jonathan Wolff (1998).
This proposition appears to also ground our common intuitions concerning why discrimination along arbitrary lines—even if it does not harm the individual—is intrinsically wrongful.
The potential challenge to my claim flagged above is that we find discrimination to be impersonally bad, as opposed to intrinsically wrongful for any particular individuals—i.e. it is a bad state of affairs irrespective of (and indeed, independent of) whether any individual’s welfare is set back by discrimination. We may find this notion attractive, in that it retains both the intuitive observation that the individuals were indeed better off under discrimination, whilst also allowing us to condemn such discrimination.
Yet I find this rejoinder puzzling. We rarely appeal to the notion of ‘impersonal bad’ when explaining why the existence of discrimination is undesirable; instead, we intuit that there is something person-specific about the wrongness at stake—it is a particular individual who is being discriminated against, and thus our intuitive uneasiness stems from the individual’s experience and treatment, as opposed to a mysterious impersonal property. Whilst discrimination does not necessarily constitute an instance of relational injustice, both discrimination and relational injustice share the same ‘wrong-making’ feature—that is, the failure to treat individuals as moral or social equals, in the absence of normatively valid explanations.
The onus lies in establishing why the wrongfulness involved in relational injustice cannot be offset by aggregate welfare gains accrued to individuals. Consider the following example, which is similar to the one discussed by Woodward (1986): the Discriminated Homeless Person. A homeless individual is denied entry into a homeless shelter from the freezing weather outside, by an explicitly racist manager of the shelter; the shelter later collapses, killing everyone inside. From a counterfactual-comparative point of view, it appears that the homeless individual, whilst suffering from hypothermia, was indeed made better off by the initial denial of entry (which prevented their death).2 However, there remains an intuition that the homeless person has suffered relational injustice at the hands of a formal structure (the shelter) on the basis of some group characteristic (their ethnicity). In other words, they are wronged without being net-harmed on balance.
There are two primary strands of arguments in favour of the view that the wrongs (in both cases of colonialism and the Homeless), are not cancellable by offsetting welfare gains. Firstly, there is the argument from non-fungibility: every relation maps onto a correlation between two individuals—e.g. the coloniser and the colonised, the shelter manager and the homeless. Whilst redistribution of money, goods, or opportunities might have compensated the colonised’s individual welfare, it did not rectify the imbalances that had previously persisted (and continued to persist) in spite of the economic advances and technological innovation introduced under colonial rule3. We intrinsically feel that there are certain items that money cannot buy (See Sandel, 2012)—for instance, relational attributes such as genuine love, compassion, or dignity and respect by other individuals. These are dimensions that exist independently of material benefits or individual welfare, in that they necessarily involve interactions between two or more individuals (e.g. it is unintelligible to discuss acquiring love without another person to love you). Therefore, assuming that such improvements to welfare did not (as per empirics) improve the relational parity between the colonised and colonisers, mere material improvements do not suffice in compensating for the previous wrongs.
Secondly, there is the argument from intentions. Many of colonial states’ greatest advances were unintended side benefits of projects primarily installed to generate revenue for the purpose of the colonisers. As such, material gains to the colonised were often the result of fortune and unintentionally favourable policies; even if this were not the case—as per certain colonies that acquired special economic status, such as Hong Kong and Singapore—the primary end objective of beneficiary economic policies remained ultimately the entrenchment of the colonisers’ interests. Consider the Saviour Burglar: a burglar breaks into a house with a malfunctioning microwave, and wakes up the sleeping houseowner in the process; the burglar steals $1,000 from the owner, but effectively saves the resident as they would have perished had the microwave later exploded. Now consider the Selfish Coach: the coach of a prodigy athlete views their success as the only means to accumulate substantial wealth and fame. As such, they sustain a relationally unequal relationship that nonetheless succeeds in training the athlete into becoming highly successful. The athlete would not have been as famous or well off had it not been for the unequal relations between them and their coach. In the case of the Selfish Coach, the benefits to the victim were accrued as a side benefit; in the case of the Saviour Burglar, the benefits were accidental and unintended. This is important, for whilst these benefits accrue to the victims in both cases above, in neither of the cases are they relevant to the particular relation between the individuals—the coach and the student, the burglar and the homeowner: the student’s success derives from their interaction with and recognition by external sporting organisations and other competitors; the counterfactual harm that the burglar ‘helps’ the homeowner avoid is originally caused by the homeowner’s microwave malfunctioning, as opposed to the burglar. The absence of active intentions to compensate renders the comparative benefits accidental, as opposed to being morally relevant and legitimate as a form of compensation. In both cases, we feel that the coach and the burglar have wronged the athlete and homeowner—in spite of the net benefits their actions brought.
Note here that I am not making the strong claim that relational injustice can never be commensurable with any forms of compensation (although this strong claim does sit well with some of our intuitions concerning the irreparability or non-compensatability of certain relation-specific goods—e.g. a particular romantic relationship, or a unique friendship between two friends); instead, I am merely making the claim that in the context of colonialism, the wrongs of colonialism have not been offset by the accidental and non-commensurate benefits that colonialism has brought.
Wertheimer (2008) discusses a distinction between harmful exploitation and mutually advantageous exploitation—the former denotes instances where exploitation is clearly harmful for the exploitee; whereas with the latter, he refers to instances where exploitation is uniquely Pareto Superior, that it allegedly leaves all parties—including both exploitees and exploiters—better off. Imagine a modified version of the Homeless case, where the manager survives the crash and receives a one-off payment that rewards him for his discriminatory behaviour, whereas the discriminated homeless man is (evidently) better off, having survived the crash by not being in the shelter.
My account offers critics of mutually advantageous exploitation with the necessary explanatory currency to explain why the homeless person is still wronged—the inherently unacceptable nature of how the homeless person is treated relationally is non-fungible with incidental material gains, and, perhaps more pertinently in this case, there exists no intention on behalf of the manager to rescue the homeless person, which suggests that it would be unreasonably generous to the manager to credit to him the incidental benefits the homeless person receives.
Objection I: Exceptional Conditions
A primary objection to the above is the view that it is reasonable and rational for individuals to want to opt to waive relational equality in exchange for greater material benefits in certain cases. For instance, I may consent to selling myself into being a servant or junior assistant to someone (lower status within a relationship) in exchange for large volumes of money; alternatively, I may accept being sexualised under oppressive patriarchal norms in exchange for financial security. These scenarios present apparent cases where, if there exists sufficient payoff, relational injustice is arguably not wrongful.
More specifically, this objection could take two forms: i) exception by consent—where, if I consent voluntarily and autonomously to accepting relational injustice under some circumstances C1, such injustice no longer wrongs me; ii) exception by rationality—where, if it is rational for me to accept relational injustice under some circumstances C2, such injustice no longer wrongs me. Applying i) to the question at hand, it could be reasoned that colonised individuals might not have been wronged if they had consented voluntarily and autonomously to their relational injustice; ii) would imply that if it had been rational for the colonised to accept their injustice in exchange for the greater material benefits, they would not have been wronged by the injustice.
Neither version of the objection is fully successful. Notwithstanding such, whilst i) is empirically erroneous, ii), it must be conceded, is partially valid, but with some notable caveats.
In response to i), note that valid consent from the colonised—even if in exchange for material benefits—was largely lacking across most, if not all, colonial projects. Movements ranging from Gandhi’s non-violence to Ho Chi Minh’s pro-independence struggle; or even large-scale protests that pre-dated independence in Ghana, Rhodesia (now Zimbabwe), and the Caribbean, were indicative of substantial popular discontent in spite of the arguably greater (and unique) socioeconomic benefits the colonial regimes had brought the colonies (Belich, 2009). Moreover, individuals often were not given the opportunity to choose whether they accepted the related benefits—there were minimal options that would permit easy international migration, let along inter-regional movements for individuals to ‘opt out’ of the provided benefits.4 Additionally, the colonised often lacked economic, cultural, and political capital to influence politics and determine the arrangements and shape of the ‘beneficial cooperative schemes’ into which they were entered against their will: merely because they ostensibly accepted and ‘enjoyed’ the fruits of colonialism did not imply that they authentically consented to them. Whilst the reasons supplied here may neither be universally nor necessarily true, they are sufficient in illustrating that the consent-based objection raised above is at best severely limited in its explanatory power.
A more promising line may be ii), i.e. that it would have been rational for the colonised to accept the relational injustice brought about by colonialism, in exchange for the allegedly substantial material benefits. This version of the objection appears attractive, for it appears tentatively counterintuitive to posit that one could be wronged by an act that one would be maximally rational (most reasonable) in accepting and not rejecting.
Yet this objection is vulnerable to the fact that it fails to block the particular intuitions we feel in response to the Selfish Coach, the Saviour Burglar, and the Discriminated Homeless Person. For instance, it appears to be (most) rational for the discriminated homeless person to accept the racist treatment they experience, in exchange for not dying (and potentially incurring substantially greater welfare losses on the whole); yet this mere fact does not diminish the powerful intuition that there is something wrong about the relation of discrimination to which the homeless person is exposed. Merely because an outcome is most rationally desirable does not imply that it does not involve a wrongdoing.5
The underlying explanation for the particularity of the above intuitions consists of the locality of relational injustice (cf. the above discussion concerning its non-fungibility)—even if it is holistically rational for an agent to accept the occurrence of the wrongdoing with their own welfare in mind, the rationality does not mitigate or resolve the particular relation between the wrongdoer and the wronged, which stands independently of the wronged’s individual welfare. In other words, the wrongness of relational injustice is local to the relation, as opposed to being a global property relevant to the individual’s welfare.
Objection II: Possibility of Compensation?
The second objection to my central thesis is as follows: as it stands, it appears that the relational injustice account precludes the possibility of any future compensation. If we are to accept that relational injustices are spatiotemporally sensitive and ‘non-fungible’ with a wide range of material benefits, it appears that such relational injustices may never be offsettable by realistic practical options that wrongdoers may pursue in compensation. Should this be the case, this notion of wrongdoing appears to inherently render any form of targeted compensation futile. Whilst this by no means undermines its internal support and consistency, an account of wrongdoing that does not allow for compensation seems to have limited practical value.
The first counter is meta-theoretical—merely because a theory does not serve to account for a particular type of action that may allegedly follow from it (e.g. compensation) does not undermine the explanatory, interpretive, and analytical virtue of such a theory.
The second counter is a concession—it is indeed possible for an individual to ‘accept’ material benefits in offsetting past wrongdoings involving relational injustice, but this concession at most grants that the wrongdoing can be fully mitigated through subsequent redress, but not offset automatically by the generation of any material benefits. In order for the compensation to not be vacuous, some initial, unoffset wrongs must have been committed—it is not the onus of this theory to defend the claim that it is never possible to compensate for past relational injustices, but merely that it requires a substantially high threshold in order to do so. Whilst it is beyond the scope to discuss the following in detail, I would suggest that an adequate theory pertaining to acceptable compensation for relational injustice must involve: i) relational equalisation—the formation of a relationship that constitutes parity and no power asymmetries; ii) acceptance—both parties must view the compensation (in whatever form it may take) as a genuine and adequate compensation pertaining to the relational injustice, and iii) appropriate intentions—the wrongdoer must possess appropriate intentions that reflect genuine desire to compensate: for example, the racist shelter manager cannot pay the homeless with the primary (ulterior) motive of preventing an ignominious expose: the compensatory move must be accompanied by genuinely apologetic intentions.6
Objection III: The Quasi-Parfitian Regrets Test
A further objection is quasi-Parfitian7: it reasons that it would be erroneous to hold that a person is wronged by an injustice if the person would not have been better off in comparison, even had the injustice not occurred. More specifically, assuming that when we posit that something is wrong we are also making the claim that something is regrettable, to regret something that is comparatively better for one appears to be unintelligible. Note here that the metaphysical assumption is that in the absence of relational inequality afflicting them, colonised individuals would not have acquired the material gains and welfare improvements, either because they would not have existed in the first place, or because the outcomes very well could have been worse.8 As such, it would be unreasonable for the colonised to regret colonialism, without also regretting the comparative benefits they derived from the process.
Note the distinction between harm (a comparative concept) and wrong (a non-comparative concept). This objection assumes that the relevant metric is reasonable regrettability, but neglects whether regrettability maps onto the concept of harm, or wrong. I suggest that it maps onto only the former—for we may find something wrong without wishing that it had not happened. The homeowner and athlete may find the burglar and their coach having wronged them, whilst simultaneously being content with the burglary and training they received. Furthermore, it is also deeply unclear as to why individuals cannot regret particular events, even if they do introduce (incidental) net benefits to them—for instance, a person who wins a lottery after undergoing an agonising surgery may have derived more benefit on the whole, but may still (retrospectively) wish that the surgery had not occurred, because the pain accrued from the surgery is non-commensurate with the 10,000GBP they subsequently win.
Most fundamentally, Parfit’s regrets test could be turned upon the argument—perhaps it is the (non-necessary) entailment between the wrong-making feature (e.g. the break-in, the harsh training, and the relational injustice under colonialism) and the welfare improvements that is most reasonable to regret: after all, it is not inconceivable for there to be a possible world where individuals’ welfare is benefited without the preceding acts of relational injustice. A possible world where the entailment between the two does not obtain is one that is not only clearly conceivable, but also potentially relatively proximate to this world. The upshot is that the colonised can regret the relational inequality they experience without regretting the benefits they accrue ‘as a result’ of the relational inequality.9
In conclusion, the concept of relational injustice offers a useful avenue to accounting for the wrongness of empires independent of counterfactual-based disputes over colonialism’s impact on individual welfare. Even if colonies had become more prosperous under colonisation, this offers no recuse for the errors of empires past.
Acknowledgments: With thanks to Doug McConnell, Fergus Peace, and Eleanor Shearer for their various comments and thoughts. With further thanks to Tom Douglas and Jonathan Pugh for their edits.
Anderson, Elizabeth (1999). ‘What is the point of equality?’, Ethics, Vol. 109 No. 2, pp. 287-337.
Belich, James (2009). Replenishing the Earth: the Settler Revolution and the Rise of the Anglo-World, 1783-1939. Oxford: Oxford University Press. p. 573.
Fanon, Frantz (1961). The Wretched of the Earth. Grove Press.
Fricker, Miranda (2007). Epistemic Injustice. Oxford University Press.
Harman, Elizabeth (2004). ‘Can We Harm and Benefit in Creating?’, Philosophical Perspectives, Vol. 18 No. 1, pp. 89-113.
King, Russell (2010). People on the Move: An Atlas of Migration. Berkeley, Los Angeles: University of California Press. p. 24.
Parfit, Derek (1984). Reasons and Persons. Oxford University Press.
Sandel, Michael (2012). What Money Can’t Buy: The Moral Limits of Markets. Farrar, Straus and Giroux.
Strawson, Peter (1974). ‘Freedom and Resentment’, in P. F. Strawson (ed.), Freedom and Resentment and Other Essays. Routledge.
Woodward, James (1986). ‘The Non-Identity Problem’, Ethics, 96 No. 4 (Jul), pp. 804-831.
Woolf, Jonathan (1998). ‘Fairness, Respect, and the Egalitarian Ethos’, Philosophy & Public Affairs, Vol. 27 No. 2, pp. 97-122.
1. Taken here to be denoting a negative or discriminatory judgment with harmful effects; for more on this, see Fricker (2007).
2. Assuming, of course, that their existence is utility-positive.
3. Even for ‘successful’ colonies such as Hong Kong and Singapore, the public continued to reside under governance systems that were led and controlled predominantly by British civil servants; in British colonies in West and South Africa, many economic and social privileges were restricted solely to white residents.
4. For more arguments that are similar to this, cf. common objections to tactic consent arguments in political theory —e.g. Simmons.
5. Consider, for instance, ‘tragic dilemmas’ in virtue ethics; whilst virtue ethics concerns the character, relational injustice concerns the relation.
6. These conditions appear to be rather demanding; one may correctly ponder if they must also obtain as general conditions for compensation for past wrongs across all cases. One potential reason why they apply particularly strongly to the context of relational injustice is the unique role played by authentic intentions and relational parity within the discursive sphere of relational equality.
7. cf. Parfit’s response to rights-based solutions to the Non-Identity Problem in Parfit, 1984
8. cf. the ‘fragility’ assumption underpinning most versions of the Non-Identity Problem —i.e. a slight change in events could result in substantial changes to future events, which produce individuals with completely distinct de re identities.
9. Let us not diverge into the interesting but ultimately orthogonal discussion of how causation ought to be interpreted.