Oxford Uehiro Prize in Practical Ethics: Should We Contact Uncontacted Peoples?: A Case for a Samaritan Rescue Principle

This essay was a joint runner up in the Oxford Uehiro Prize in Practical Ethics Graduate Category

Written by University of Oxford student Brian Wong

Uncontacted peoples refer to individuals who live (by choice or by circumstance) without coming into contact with broader, greater civilisation.[1] I make the idealised[2] assumption that our act of contacting seeks to provide these peoples with goods, opportunities, and access to advantages. I outline a tentative argument as to why it can be obligatory to contact[3] uncontacted peoples, basing my claim on a Samaritan modification to Singer’s Rescue Principle[4]. I will then raise several limiting conditions to the above Samaritan considerations. Noting that the key difference between contacting uncontacted peoples and more ‘uncontroversial’ Samaritan acts lies in the heightened epistemic indeterminacy in the former, I conclude by presenting a principle of epistemic prudence that has wider implications for broader instances of paternalistic intervention.[5]

Consider first the relevant facts: in 2013, it was estimated that there were over 100 uncontacted tribes around the world, residing in densely forested areas of South America, Central Africa, and the Indian Ocean. These tribes generally live without access to modern medical or communicative technology, and are more vulnerable to natural threats (e.g. predators, disasters etc.).[6]

I now turn to the case for our contacting these uncontacted peoples. Let’s begin with Singer’s relatively uncontroversial Rescue Principle (RP).

RP: It is morally obligatory to prevent some bad state of affairs S, without any costs of comparable moral significance.

Consider the Samaritan Rescue Principle (SRP): It is morally obligatory to bring about certain goods P* to which individuals have strong claim rights to, without any costs of comparable moral significance.

The SRP supplements the RP; it stipulates that agents have, beyond a tentative duty to prevent bad states of affairs, a duty to bring about specific types of goods (“presumptive goods”) to which their recipients have strong claims. Consider the Student – Sam graduates from primary school and leads a reasonably fulfilling life as a manual worker. However, his career options will be vastly expanded if he is accepted into a high school, which involves little to no costs to society. Sam does not appear to qualify for assistance under the RP, but the SRP would cohere with our intuitions that it seems only justified for Sam to be offered a place at a high school. A right to reasonable life chances and opportunities[7] may be something to which individuals have strong claims, and yet the failure to meet such a right is not bad enough to warrant the activation of the RP.

If one endorses RP, one also has good reasons to endorse SRP.[8] Both principles share similar underlying motivations – it is undesirable that some bad states of affairs occur, but also that certain individuals do not have access to particular goods viewed as significant. Note here that some states of affairs can be undesirable, and thus generate reasons for avoiding them, because they are bad by being comparatively worse than their next most likely counterfactual. Other states of affairs can motivate avoidance of them because they lack the presence of certain presumptive goods, and are thus undesirable in an absolutist, baseline-oriented sense. Whilst the notion of badness is comparative and the claim about presumptive goods seemingly thresholdist, both accounts fundamentally posit that it is morally obligatory to bring about a state of affairs that is more desirable than its counterfactual along some metrics, whether it be the accessibility to presumptive goods or the extent of badness. An obese child or primary-graduating Sam may be in a ‘good state’ because the closest alternative features a far worse-off obese child or Sam. Yet even if this were true, this does not mitigate the absolute badness[9] in their obesity or inability to access high school. If we feel that individuals ought to prevent what is undesirable, and we also recognise a non-comparative form of undesirability, then we ought to endorse SRP.[10]

Given the SRP, I now make a two-step claim that contacting uncontacted peoples brings about the certain goods S* to which individuals have strong claim rights to. The first part outlines what such goods may constitute; the second part addresses why contacting uncontacted peoples is necessary or highly efficient in delivering such goods. Under the first, note that SRP is far more expansive as a principle than RP, because it enables us to focus on the imperative to provide particular goods, which may exist alongside the imperative to prevent bad states of affairs. More specifically, then, what are goods that individuals have strong claims to, but the absence of which does not fall under a bad state of affairs? These may be goods such as a long lifespan, health that is typical of an individual living in the 21^st century, or the ability to fully self-actualise given a sufficient range of options. It may not be bad for a person to live only until 50, or to lead happy but unfulfilled careers, or to live in sub-healthy ways; yet such ‘goods’ are certainly presumptively good for them.

Here it seems almost paternalistic to an unreasonable degree[11] to assume that the particular goods I outlined/discussed above are in fact good for the indigenous groups, without further justification. There are two explanations that ground our intuitions here: firstly, these goods are often instrumentally quasi-necessary for unlocking a wide range of other goods (if one is dead at 50, one cannot access more pleasure or satiate one’s preferences); secondly, these goods – whilst not fundamental to survival – are crucial in the maintenance of individuals’ moral personhood and agency[12]. Thus their absence in some may well not be bad or comparatively worse than the alternative, but can still be undesirable in a way that motivates our provision of them. I want to emphasise here that there certainly is a non-trivial probability that these meta-metrics may well be themselves faultily prescriptive. Thus the explanations offered here are defeasible. However, in the absence of potentially trumping considerations or substantiated objections, I suggest that we should take instrumental quasi-necessity and personhood/agency preservation as reasonable external criteria for determining the (intrinsic) values of particular goods.

Empirically, the provision of healthcare, technology, and options to engage in education can provide uncontacted peoples the above presumptive goods; contact with the outside world can expand their lifespans through medical resources, as well as enhancing their access to knowledge, opportunities that are exclusive to interactions with the outside world (e.g. the opportunity to access the Internet), and broader range of lifestyles and recreations. [13] We thus avoid the controversial judgment that uncontacted peoples live in a bad state of affairs, without neglecting the duty to introduce presumptive goods to their lives through initiating contact.

I now turn to the limiting conditions for the above principle. The first condition that must be met is the 1i) existence of presumptive goods with contact that 1ii) outweigh the presumptive goods without contact. Some aptly note that uncontacted peoples have little to gain from contact due to the a) substantial health risks[14], b) power asymmetries that minimise likelihood of gains and maximise probability of exploitation[15], c) the inability of (forcefully) assimilated uncontacted peoples to appropriate any benefits from the process. Others[16] are wary that the marginal gains are outweighed by the opportunity costs of other presumptive goods uniquely associated with isolation from the outside world, such as d) political sovereignty and self-determination, e) social intactness and cohesion, and f) cultural integrity. Part 1i) to this condition addresses concerns including a)-c), whereas the comparative nature under 1ii) addresses d)-f). This condition precludes contact when the outcomes are apparently worse. Note that the SRP is lexically superseded by the RP – if some act P brings about some presumptive goods but also introduces a (comparatively) bad state of affairs, then P is impermissible and not-P is required.

A powerful objection to my argument operates as follows: contact with uncontacted peoples is often carried out without their consent, in a way that renders the contact with them an infringement to their 2i) autonomy, and, if coupled with a preferentialist understanding of welfare[17], 2ii) welfare – specifically their presumptive goods. 2i) suggests an independent reason as to why contact is impermissible; 2ii) challenges the empirical premise above.

A cheap response is to introduce the limiting condition of explicit or tacit consent. Yet this limiting condition appears to cede too much; considering that most uncontacted tribes have resisted external contact[18], it is unclear if accepting this limiting condition would permit any contact at all. Furthermore, this condition also neglects the possibility that minorities within these uncontacted groups may not consent to being governed by their tribal leaders[19], which generates questions of indeterminacy.

A better response to both 2i) and 2ii) is to adopt the limiting condition of retrospective consent – i.e. contact should not have occurred if it is the case that the contacted individuals do not retrospectively consent to having been contacted.[20] In some cases, the consent withheld or granted after a more complete experience of a choice option is often viewed as more valuable than consent offered before experiencing the choice option – this applies specifically to cases where individuals cannot acquire an accurate understanding of the options they are presented with, without experiencing them from an “internal point of view”[21]. Examples for this may include the enrolment of children into Sunday schools, physical education, or mandatory schooling, even if they do not consent to entering into these in the very first place. As such, the above objections are diffused by the fact that some uncontacted persons, after contact, may express retrospective consent to being contacted.

A core issue to this limiting condition concerns adaptive preferences[22]. Consider Stockholm Syndrome, where hostages develop psychological affection for their captors in captivity, and – through the initially coercive conditions – thoughts that rationalise their initial abuse. Likewise, uncontacted peoples – post-contact – may well develop preferences that rationalise their contact (thus granting retrospective consent); it is unclear if such preferences should be taken seriously.

In response, I recommend a potential move towards idealisation, in reformulating the limiting condition as whether an omniscient advisor, given the preferences (or second-order preferences) held by the uncontacted peoples prior to contact, would advise them to consent to the contact knowing the consequences. The idealised forward-looking consent condition resolves the concern.

So far my argument appears to exist in a vacuum, detached from realist concerns[23]. The following areas of ignorance of policymakers may be particularly troubling: 3i) preferences – we simply do not know if uncontacted peoples would retrospectively consent to our contact; 3ii) the good – we may not know what falls under ‘presumptive goods’, such that even if SRP is true, we may still be wrong about what actually are presumptive goods[24], and 3iii) outcomes – we do not know what will happen if we do contact uncontacted peoples. The Elephant in the Room, then, is that my principle seems to have little to no real-life import. Unlike saving a drowning child from a shallow pond[25], contacting uncontacted peoples has high moral risk deriving from indeterminacy under all three categories (i-iii).

My third limiting condition hence calls for epistemic prudence. Consider here the Rare Disease Patient. A patient suffering from illness Y collapses, and is need of medical attention. A trained doctor walks by, and – from the signs exhibited by the patient – has good reasons to believe that i) their own expertise is reliable (reliability premise) and ii) the patient is, as per their expert judgment, suffering from a Disease X (judgment premise). In applying the treatment for X to the patient, the doctor fails to treat the patient appropriately, and the patient thus dies. We do not fault the doctor here, for they are behaving in an epistemically prudent manner. Rather similarly, even if there exist facts of which agents initiating contact have no knowledge, so long as within the information available to the agents, epistemic prudence is exercised, contact with uncontacted persons can still be obligatory and permissible.[26] This principle has implications for wider discussions about paternalistic interventions – paternalistic agents ought to ensure that the goods they seek to promote are indeed goods, but also that the benefits introduced outweigh the costs.[27]

An objection to my third limiting condition may be the following. Western Colonialism – within the information available to colonial conquistadors, they may have held the active beliefs that they have both the obligation and permission to (forcefully) colonise lands and ‘transform’ the hearts and minds of the colonised such that they would become ‘civilised individuals’[28]. The third limiting condition of epistemic prudence does not appear to rule out instances of colonialism, given that it suggests that the intentions of the colonisers were sufficient in rendering the colonial acts both permissible and even obligatory. This appears to be both an absurd and deeply troublesome implication of the theory at hand.[29]

There are three key responses to the Western Colonialism objection. Both are compatible with (and indeed somewhat follow from) the view that colonialism was and remains morally abhorrent and impermissible. The first response is an empirical contestation – there exists evidence to suggest that with regards to colonialism, both i) the initial contact enacted by colonisers and ii) the subsequent actions undertaken by colonial regimes and administrators were intentionally malicious, grounded upon not concerns for the interests of the colonised, but blatant self-interest orientated towards the colonial powers’ metropoles.[30] Even to the extent there was some intention to improve the colonised regions, the public and political will was predominantly motivated by ulterior motives that exemplified little to no exercising of epistemic prudence with respect to the colonised’s interests and desires: the judgment premise (i.e. the colonised would benefit from the act of colonialism) was not met; even if it was, the reliability premise falls, because the colonisers – given their engagement in practices of epistemic silencing, quieting, and injustices[31] – had no good reasons to believe that their own expertise was reliable[32]. The second response is that past acts of colonialism must be assessed in conjunction with the first and second limiting conditions (the presence of idealised forward-looking consent and overall greater presumptive benefits than harms). Even if colonialism had been overall beneficial as a whole, in terms of introducing more presumptive benefits, it is deeply unthinkable or implausible to argue that the colonised would have offered idealised forward-looking consent in response to the atrocities committed by their colonisers.

The final response notes that at best the above argument makes the case for pro tanto obligations to contact uncontacted peoples on grounds of the SRP. Note here that even under the RP, the RP is constrained and offset by concerns of moral costliness (e.g. if saving the drowning child would cause ten distant children to drown, the pro tanto duty to rescue the drowning child is outweighed by the moral costliness of the ten extra deaths); similarly, the SRP can and should be deemed as offset by more urgent moral concerns – e.g. minimising violence, infringements upon bodily autonomy etc. (which occurred rampantly under colonialism). All three of the above responses suggest that accepting my account does not require us to re-evaluate or re-conceptualise colonialism as permissible and obligatory (especially when all things are considered). Thus the Western Colonialism objection falls.

In conclusion, the SRP suggests that we should contact uncontacted peoples, providing that the contact is i) comparatively better with regards to presumptive goods provision; ii) retrospectively consented to under idealised conditions, and iii) epistemically prudent. The condition of epistemic prudence requires both reliability and judgment aptness. The conclusion of this paper is not only immediately relevant to the ethics of uncontacted peoples, but also to debates about paternalism and state neutrality.[33]

[1] Walker, Robert, and Kim Hill (2015), “Protecting Isolated Tribes”, Science, Vol. 348

[2] Valentini, Laura (2012) “Ideal vs. non-ideal theory: a conceptual map.” Philosophy Compass, 7 (9). pp. 654-664

[3] Let us assume that the contact between the greater civilisation and uncontacted peoples here is largely benign – i.e. we are not discussing extreme instances of deliberate murder or resource extraction.

[4] Singer, Peter (1972), “Famine, Affluence, and Morality”, Philosophy and Public Affairs, Vol. 1, No. 3

[5] Quong, Jonathan (2010), Liberalism without Perfectionism

[6] Holmes, Bob (2013), “How many uncontacted tribes are left in the world?”, New Scientist

[7] Nussbaum, Martha (2000), “The Capabilities Approach”

[8] See Kahane, Guy and Julian Savulescu (2016) “Disability and Mere Difference” for a similar argument in their refutation of the Mere Difference account of disability.

[9] I assume here that personal badness (i.e. ‘badness for’) translates into overall/holistic badness (i.e. ‘badness in general’), in the absence of countervailing reasons that prohibit such translation/conversion.

[10] This claims-centric view also coincidentally addresses a challenge frequently launched towards unrestricted Maximising Consequentialism (that we ought to bring about the best state of affairs we possibly can), i.e. the Asymmetry[10]: it is not wrong to not bring an additional individual into existence, because non-existent persons, given their inability to formulate intelligible “I-desires” or preferences, can develop no strong claims to the goods they could only hold preferences about if they actually existed. A non-existent person has no claim to the right to life (Elliot, 1989).

[11] I thank Prof. Cécile Laborde for raising this objection to me over a casual conversation.

[12] See Rawls, John, Political Liberalism (1993), on how the possession of moral powers is a prerequisite of individuals’ having freedom in a substantively valuable sense. I shall not enter into the quagmire concerning theories of autonomy here.

[13] Walker and Hill (2015)

[14] Kluger, Jeffrey (2015), “Is It Ethical to Leave Uncontacted Tribes Alone?”, Time

[15] Kluger, 2015

[16] Survival International (2015), “Survival International attacks Science editorial as ‘dangerous and misleading’”, Survival International

[17] …and what presumptive goods may entail

[18] …often through exerting violence, see Kluger, 2015

[19] See Will Kymlicka’s Multicultural Citizenship: A Liberal Theory of Minority Rights (2016); I will refrain from entering the complex discussion as to the extent to which an individual can meaningfully withdraw consent from living under a political arrangement (e.g. the tribal leaders of the uncontacted tribes) without any knowledge of alternatives to said political arrangement.

[20] See George Sher’s Beyond Neutrality: Perfectionism and Politics (1997)

[21] Sher, 1997

[22] More commonly known, perhaps, as the ‘false consciousness’ argument channeled by certain Marxists in explaining the intrinsic oppressiveness of capitalist structures in disaggregating class consciousness and suppressing recognition of the problematic nature of capitalism.

[23] As Charles Mills (2005) may argue, I may well be disguising my ideological commitments through the veil of ideal theory.

[24] Cf. a child seeing a can of cyanide and thinking it is cough syrup or good for their health.

[25] Singer, 1972

[26] In practice, the bar to satisfying the reliability premise may well be very high – our imaginations of the good are culturally tainted and socially constructed, and both states and NGOs may have distorted interests that result in epistemic inadequacies. Yet merely because the bar is high does not preclude the existence of the obligation, once the bar is attained.

[27] Word count: excluding footnotes, 2,000 words

[28] See Fanon (1967), Black Skin, White Masks for his analysis of the (psychological and physical) brutality of colonialism, and how it was legitimated through the narrative of inducing civilisation and salvation.

[29] Mills (2005; 1997) notes the prevalence of racist, whitewashed assumptions made in analytical/liberal political philosophy. It is worth noting that one of the ways in which racism is discursively enforced is via the omission and entailment of certain deeply troubling conclusions with respect to racial equality.

[30] See Tharoor (2017), Rodney (2011) for extensive discussions of how the Western colonial powers had systemic incentives in underdeveloping and depriving their colonies in order to enrich their colonialist metropoles.

[31] Fricker (2007)

[32] It is worth clarifying here that ‘good reasons’ here embed a quasi-externalist conception of (justified true) belief here – it does not suffice that the agent involved thinks that they have good reasons; these reasons must be good themselves (see internalism vs. externalism debate in epistemology).

[33] With thanks to Samuel Chan (University of California, Berkeley Campus), Prof. Cécile Laborde (Nuffield College), Ho-Yin Yuen (University of Oxford) for their comments, feedback, and insights (in conversations).