Moral Compulsion To Torture A Child

The Pitfalls of Deontological Absolutism: Why We Are Morally Compelled to Torture to Death One Innocent Child to Create a Peaceful Fabric of Human Destiny
{School of Social and Political Science, University of Edinburgh}

When faced with such a dilemma the majority of persons would likely deny their willingness to undertake such an action, and would do so not by virtue of systematic reasoning but out of an intuitive moral repulsion at the idea. The basis of that repulsion is three-fold: First, there is a deontological, or absolutist, moral objection to using persons as means to an end; Second, that because our torturing this child would be a direct and intentional action, it is less morally acceptable than our failure to prevent harms occurring to others by not acting; and Third, that the use of torture against an innocent victim is a priori too great a cost for any intended good. What this work seeks to demonstrate is that given the parameters of the question posed to us, these arguments not only falter when followed through, but in fact turn inwards upon themselves to morally demand our making of this sacrifice rather than to forbid it. Indeed, much of the weight that causes the collapse of the absolutist argument in our case comes from questionable assumptions about moral calculations, such as that action is more condemnable than inaction or that what we face is a choice between protecting negative rights and asserting positive rights. These and similar assumptions shall be considered in turn, and in doings so it shall be shown that torturing the hypothetical innocent being to death is in fact the morally correct course of action given what is at stake.

Because the cost of granting humanity ‘peace and rest at last’ is so great, namely the torturing to death of a sentient being (let us say it is a human child), the benefits and conditions of doing so must be stated in no uncertain terms. Let us turn to what these are, and why they are essential if we are to respond ‘yes’.

We may borrow from Just War theory a basic framework for moving ahead. We begin with the concept of ‘proportionality’. One interpretation of the ‘peace’ and ‘happiness’ at stake is somewhat Benthamite, wherein individuals are provided with the means to fulfil desires, live purposefully, and enjoy pleasure in the absence of pain (Glover 1977, 63). For the cost of a life, however, this is inadequate. A smile is not worth a life, no matter how broad. The act we propose is an extreme manifestation of those things which it must, if we are in fact to undertake it, prevent, namely harm through violence. If we are to commit an act of extreme violence, it must be at minimum for the purpose of preventing such acts from occurring in the future.

Following from this is the criteria of ‘just cause’. Recall that this is not your standard hypothetical moral dilemma, where we are asked if we would kill one person to save five. This is about inflicting harm and death upon one in order to bestow “preventative benefits” (Hanser 1999, 279-280) to all humanity. In our case, ‘proportionality’ is the level of harm inflicted upon the tortured child as compared with the cumulative suffering of all other humans on our planet. Our ‘just cause’ is to prevent this suffering from occurring. The ‘chance of success’ criteria is fulfilled through the phrasing of the question – it is certain that, if we accept the condition of torturing someone to death, a new world shall come to pass.

As for the criteria of ‘last resort’, there is little to be said. It would take a special kind of utopian optimism to imagine that there are any diplomatic or cooperative means of abolishing in totality violence between both political entities and individuals, or to provide the basis for universal human ‘happiness’, however defined. Having established these criteria, we may now consider each of the three proposed objections to committing this action in turn.

The Deontological Objection

According to Kantian moral philosophy, people must be seen as ends themselves and never as means (Cummiskey 1990, 586). Essentially, the argument against the utilitarian causing of harm mounted here rests on two foundations: first, the sanctity of rational life and the Kantian duty to allow it to thrive; and second, the intrinsic wrongness of certain actions – namely violence – as means to ends regardless of circumstances or consequences. In this section, the questionable moral position of absolutism in the context of our dilemma shall be assessed.

Let us first address the principle of the sanctity of life and how it is challenged by utilitarian consequentialism. This principle holds that taking[1] life is “intrinsically wrong” because there is something special, even sacred, about conscious, rational life that warrants preservation (Glover 1977, 57). This should so far be uncontroversial, but where the deontological perspective comes in is in the form of the “no trade-off view”, which holds that individual life warrants preservation above all else, regardless of circumstances or consequences (1977, 57). Immediately, then, we see a basic conflict arise. If, as John Stuart Mill argues, all rational persons have an equal right to happiness and well-being (1972, 42), what happens when the rights of many can not be guaranteed unless the rights of one are taken away? According to Kant’s normative theory, we are foremost obligated to “promote the conditions necessary for the existence of rational beings” and, conditionally, to “promote the ends or happiness of rational beings” (Cummiskey 1990, 597). Given what is at stake in Dostoevsky’s question – the provision of a state of peace, thus the condition for the flourishing of rational beings – are we not morally complicit in the denial of these conditions if we refuse to create them, when we have the ability to do so? True, we are committing an act in the present in order to prevent harms from occurring in the future, and thus are subject to the controversy of preventive action; but given that these harms are certain, wide-spread, and egregious in the form of the cumulative suffering of humanity, it seems that choosing not to make our sacrifice is not only anti-consequentialist, but anti-Kantian as well. If it is our duty to advance the conditions necessary for the flourishing of rational beings, it makes little sense to not do so because of the agent-centred constraints of absolutism (Cummiskey 1990, 615). But let us not dismiss this argument entirely, and proceed to look at why such a moral onus is placed on the wrongness of enacting violence as a means to an end.

Is violence, even when done to prevent greater violence from occurring, an “absolute evil” (Narveson 1980)? Where an action results in the infliction of harm in order to prevent harm done to others, absolutists point to the significance of what Robert Nozick calls “side constraints”, being constraints on how goals may be achieved focussed around the duty to not use people as means (qtd. in Cummiskey 1990, 591). The emphasis in this case, argues T.M. Scanlon, is agent-centred and concerned with what one does as opposed to what happens (qtd. in Scheffler 1994, 110). An absolutist would therefore argue that it is not morally correct to consider consequences over actions in themselves; but suppose, as Jan Narveson does, that we are faced with a situation in which one of two actions can be chosen, both of which would lead to violence, yet one would cause more violence than the other (Narveson 1980, 119). Would it not be in the interest of the Kantian absolutist to least violate the right of persons to autonomy and freedom from harm? If, as Narveson notes, violence is absolutely evil, then “the reasonable thing to do would seem to be to minimize it – to do what produces the least of this irredeemable evil” (1980, 119). It is not an option to, as the absolutist would prefer, walk away and claim that since you made no decision, and therefore did nothing, you are morally free from fault regarding the consequence, because this idea rests on a false distinction between the morality of action and inaction (explored in the next section). In situations such as these, which embody the kind faced in the question we are attempting to respond to, “the presumption against killing may be correspondingly weaker, or more readily overridden by countervailing circumstances” (McMahan 2003, 189). There can be, as Cummiskey observes, “prima facie constraints on the acceptable means to a goal” which are nevertheless non-absolute when facing a moral dilemma (1990, 591).

The absolutist, at this point, is liable to argue that there is a moral hierarchy which places a greater burden on action than inaction, and which also asserts the preservation of negative rights over the provision of positive rights, thereby forbidding our torturing to death of this child regardless of whether or not it will prevent the suffering of others. What is really important to absolutists, it seems, is that “I am never responsible for other people’s violence” (Narveson 1980, 120). It is to this supposed moral distinction between action and inaction, as well as between negative and positive rights, which we now turn.

Intentionality and the Rights-Based Objection

A consistent factor behind the absolutist position is the issue of agency, which boils down to what Alison McIntyre calls the “moral distinction between causing and allowing” (2001, 229). This distinction arises because “what absolutism forbids is doing certain things to people, rather than bringing about certain results” (Nagel 1972, 130). We can look, for example, to the classic scenario posed by defenders of the Doctrine of Double Effect, wherein a bomber pilot engaged in war may morally drop a bomb which will kill civilians because his intent is fixed on the objective of ending the war, whereas a terrorist bomber is essentially immoral because his intent is the deaths of persons used as a means to an end. For absolutists, the pursuit of a good is less morally acceptable when the harm caused to achieve it is intended rather than merely foreseen (Quinn 1989, 334-335), despite the fact that the person killed without intention is just as dead as the one killed explicitly.

Let us now apply this reasoning to our dilemma of killing one to prevent harm and violent death for many, or not killing one and allowing those other harms to take place. Joachim Asscher argues that “the moral difference between killing and letting die is that killing involves taking responsibility wherein letting die does not”; the former is “terrible”, while the latter is merely “negligent” (2007, 279). However, when you merely ‘let someone die’ as opposed to killing them outright, as we would those multitudes whom we could save through our sacrifice, do you not in fact intend for them to die? As Warren S. Quinn observes, the consequences of not making a choice are just as much a demonstration of positive agency as are the consequences of making a choice, since we are choosing the ramifications of not preventing something over the ramifications of preventing it (1989, 301-302). Take, for instance, the much-heralded trolley case. Standing at the junction where we might pull a switch to put the trolley on a track with one person on it rather than five, the decision not to pull the switch is an act of agency by which you intend that the trolley will continue on its path to save the one, but by which you also intend that it will kill the five (1989, 304-305). The capacity to act was yours, yet you chose not to. You are therefore responsible for saving one life, and for taking five. This raises an important point: Hanser writes that when an agent “prevents someone from having his life saved”, as the absolutist course of action would, “the resources the victim needs in order to survive generally do not belong to the agent” (1999, 290). In our case, however, we do possess the resources to do so, which puts the moral burden squarely on our shoulders. The idea that to let someone die or “preventing people from being saved” (Hanser 1999, 277) is passive while ‘killing’ is active is therefore a false dichotomy. Inaction is action. Even if the agent were not to be held morally accountable to the same standard for their inaction as for their action, they would at minimum be responsible for perpetuating the conditions under which rational human beings are unable to thrive – a Kantian tragedy.

Despite this, the absolutist may still argue that the right not to be killed (held by the child) and the right to be saved (held by humanity en masse) is a contest between negative and positive rights, and in such a contest the former must always prevail. Let us consider this rights-based argument, and see if it overwhelms the moral culpability of inaction I have proposed above.

Can the morality of our choosing or not choosing to kill the innocent child in order to bring about peace for others be framed in terms of the supposed superiority of negative rights over positive rights? Hanser, for instance, writes that because “the right not to be killed is negative, whereas the right to be saved is positive”, the right not to be killed is superior to the right to be saved (1999, 288), and so we are not morally permitted to kill our hypothetical innocent. Put in terms of a familiar scenario, in which a gunman, A, is about to kill innocent bystanders B, C, and D unless we intervene and kill A, although our agency puts us in a position where we could allow three to die or kill one directly we are not permitted to do so because it would be a violation of A’s negative right to not be harmed, while it is a positive duty towards B, C, and D on our part to save them.

I propose two objections to this line of reasoning. The first follows from the argument made above that inaction is a highly purposive form of action, resulting in moral culpability for the consequences thereof, especially when the agent in question possesses the resources to save the potential victim or prevent them from being saved. This does not bear repeating beyond saying that where one denies the creation of the conditions for the survival and thriving of rational beings when one is able to provide it, the Kantian absolutist by his own priorities shoots himself in the foot by arguing for the prevention of the lesser evil. For the reasonable Kantian (that is, one which would not refuse to infringe upon any negative right even to prevent, for instance, such an extreme event as a genocide), “the rights protecting the greater balance of good should, ceterus paribus, prevail” (Quinn 1989, 308). The second, and more fundamental objection, is that this whole notion of having to choose between the violation of negative and positive rights is yet another false dichotomy; rather, we are choosing between the violation of a single person’s negative rights and the violation of every other person’s negative rights. Were we to have conceived of the criteria for our acting in the affirmative in other terms, the distinction might indeed have rung true, however we did not, making the elimination of violence and outwardly-inflicted misery from the world our markers for ‘peace’ and ‘happiness’. The choice is not between killing an innocent to bring smiles to more people’s faces, or to aid them in some way which is beneficial but not integral to their humanity; the choice is between killing an innocent and protecting the negative rights of all other innocents from being violated. Our action is no more a defence of others’ positive rights than is a government’s outlawing of murder to serve as a guarantor of the negative right against having your life threatened by violence. ‘Happiness’ alone is a positive right, or perhaps even merely a privilege – the integrity of your personhood is an explicitly negative one. This being the case, it is morally absurd to argue that it is more permissible to fail to prevent millions of negative rights violations from occurring (many as extreme as the one we are to commit, such as murder, rape, and indeed, torture) than to inflict a single one.

Torture, Innocence, and the Dirty Hands Objection

Lastly, there is the objection that employing torture as a means to secure peace and happiness, especially against a defenceless and innocent victim, would morally stain society henceforth – what Michael Walzer calls the “problem of dirty hands” (1973). It is a priori too great an evil to be used as a means towards any ends. Does this objection hold in our case?

There seems to be four common objections to the utilitarian use of torture. Firstly, that torture is wrong as in action in itself, that is, it is deontologically impermissible (P.F. Brownsey, commentary in Jones 1980, 14); Secondly, that while it is certain that harms will result from torture – to the victim, the torturer, and the society which condones it – the likelihood of good effects resulting from it is low (Sussman 2005, 12); Thirdly, that using torture as an extraordinary measure in extraordinary circumstances creates a slippery slope for its use to become more normalized, as in the ‘War on Terror’; and Fourthly, that there is no certainty that torture will accomplish what it sets out to do – for instance, in the ‘ticking bomb’ scenario, there is no guarantee that the individual we torture will tell the truth, or that they even possess the relevant information. To these objections, three replies can be made which negate them in our specific scenario. We may respond to the first one in the same way that we did in the section on ‘The Deontological Objection’. To the second objection, we can say that because of the way in which the hypothetical question posed to us is constructed, we know that the desired goods will come to pass, and according to the conditions we established. And to the third and fourth objections, because our action would create an environment wherein the further use of torture would not exist, we do not have to worry about opening the door to future abuses; there is no slope to slip on. We are, as Machiavelli phrased it, doing ‘all our evils at once’ (1996, 30) and thereby preventing any further ones from occurring. Admittedly, this is a luxury afforded to us by the hypothetical nature of the question we seek to answer. So it seems there is no difference in our moral position premised upon the idea of torture as a distinctive form of violence.

Does the innocence of our victim, then, morally burden our decision? Writers concerned with the deontological morality of killing argue that it does, because there is a Kantian obligation to justify the application of harm to individuals on a one-by-one basis. Michael Otsuka, for instance, proposes a “numbers skeptic” approach, where when one can save a single person or multiple persons in mutual exclusivity, one must consider the claim of each individual rather than automatically saving the greater number (2004, 414-415). We can in turn mount two replies to this. Supposing that this is a moral mode of calculation assumes an indefinite value on innocence in itself which is inflexible to circumstance, in the same way that a “very strong thesis of precedence” of negative rights over positive rights rules out any violation of negative rights to save “any number of others even from death or torture” (Quinn 1989, 307). This manner of Kantian absolutism again turns upon itself in that its respect for autonomy might in fact require “a conscientious Kantian moral agent…to sacrifice the innocent because it will promote the good” (Cummiskey 587). Our situation is not a simple trade-off between one life and another – it is a trade-off between one life and the creation of the conditions for all others to be free from harm, the provision of which is a Kantian obligation. This leads to our second reply, that our choice is not between not killing an innocent and not killing a non-innocent, but between not killing an innocent and not saving countless other innocents. The innocence of the child to be killed, then, is only relevant if actively taking the life of an innocent produces moral culpability while failing to save other innocents does not. Because, as has been argued, the latter is indeed an act of agency and therefore of moral culpability, the question of innocence is for our purposes largely irrelevant, because innocence is at stake through both our action and inaction. If anything, we are compelled to preserve the greater innocence, which undoubtedly permits us to make our sacrifice.

To dispel further doubt, let us engage in the opposite of the question of innocence: what if the intended torture victim is a terrorist, and therefore, presumably, ‘evil’? Would this make our action more morally permissible? I think not, for two reasons. First, and simply, if torture is morally reprehensible, it is always so, regardless of the victim. To seek a differentiation between types of people based on the rightness or wrongness of applying torture to them is in effect to ask, ‘who deserves to be tortured more’, the answer being that nobody can deserve to be tortured. But more fundamentally, it does not matter what ‘type’ of person we are torturing, be they a saint or a mass murderer, because their moral status has no effect on the kind of outcome produced. Torturing an ‘evil’ person will take our peace no further, just as torturing an ‘innocent’ person would not make our sought-after world less peaceful. It is not as though there would be one less terrorist in our peaceful world; there would be none regardless. So while torturing a terrorist to death to achieve our goal may be intuitively more appealing than torturing to death an innocent child, we see that it in fact makes no difference to the morality of our decision.

The Decision

Having emphasized the fallibility of the arguments for uncompromising absolutism, we are able to confidently choose our course of action free from our intuitive moral objections to it. This is not say that the decision to torture and kill anybody, innocent or otherwise, can ever be free from moral constraints. In fact, my position is quite the opposite – we must apply and meet extremely stringent and challenging criteria if such a thing is even to be considered, especially if it is then to be given the label of a moral action. Let us review our case.

Our prerequisite criteria of ‘proportionality’, just cause’, ‘chance of success’, and ‘last resort’ are met by the phrasing of the question posed to us, with the exception of the precise nature of what is meant by ‘peace and rest at last’, which we have, in minimal terms, outlined. We have established that the deontological objection is morally untenable and contradictory because it leads to situations in which the very evils it seeks to limit (violence and violations of rational personhood) can be allowed to flourish when there are no exceptions made to its policy of not causing harm. Neither is the intentionality and rights-based objection to answering ‘yes’ to Dostoevsky’s question valid, because by not acting we are in fact making a very conscious decision to allow harm to come to innumerable persons based upon a false dichotomy between the defence of negative and positive rights. Finally, the objection to torture on account of the innocence of our subject also falls short, because the especial evils of torture are not applicable as our assured result precludes them, and also because that very innocence does not increase our moral burden. Would I, then, consent to being to being the architect on these conditions? I am morally compelled to answer ‘yes’.

Citations & Bibliography

 

Tagged Absolutism, Consequentialism, Kant, Morality, Torture