This issue rests on the larger question of what grounds obligations towards children in general.  Tim Bayne and Avery Kolers have plausibly suggested that any account of parental obligation will be pluralistic – we should recognize a number of distinct sufficient conditions for parental obligation, reflecting the varied structure contemporary families may take.  With this in mind, I will discuss four options to ground parental obligations: communal, consensual, causal and biological.   Only two of them – biological and causal – could sensibly justify the obligation of unwilling fathers to their children, but each faces particular problems of plausibility and counterintuitive implications.  These do not exhaust all possibilities, but they are some of the most prominent in the literature.  In the end, some degree of bullet-biting will be necessary – something will have to give.

Four models

Communal obligations towards children are a cornerstone of the welfare state.  As children are members of society, liberal societies will extend to them the general provisions of welfare, including education, healthcare, and direct financial support.  Indeed, such obligations are stronger than general welfare duties insofar as children cannot be said to be responsible for their predicament.  Unwilling fathers in a welfare state will then owe obligations to children via the taxes they pay, a portion of which will go towards welfare programs.  But this obligation is very indirect, not singling out fathers per se, nor their children in particular.  One might wonder whether unwilling fathers should bear a disproportionate burden of welfare support for their children, but such a duty would have to be grounded in reasons other than general communal duties.

Consensual obligations certainly make sense as a sufficient condition for parental obligation.  Entering into voluntary contracts is an exercise of individual autonomy, and having children in particular is a central case of exercising reproductive autonomy.  Demanding that parents fulfill obligations they agreed to is quite respectful of this autonomy, and assuming the conditions of consent were appropriate, would be perfectly fair.  The consent model (unlike the causal and biological model) also helps explain why we treat adoptive parents have similar obligations as biological parents.  But again, this won’t help a defender of financial obligation of unwilling fathers, as unwilling fathers did not consent to have a child – hence their unwillingness.  One could argue that they provided tacit consent by having sexual intercourse, but it is not at all clear that such an action could come anywhere near the usual conditions of information and voluntariness that are needed to make consent morally transformative.  In addition, it would be especially hard to argue for consent in cases where the father agreed to the use of contraceptives – in such a case, the father was if anything explicitly conveying an unwillingness to become a father.  It is this lack of consent that Shrage (as well as, a decade earlier, Elizabeth Brake) relies on for her case against the financial obligations of unwilling fathers.

Causal obligation, by contrast, could be used to justify the obligations of unwilling fathers – though it is a dubious general model, even as a purely sufficient condition on parental responsiblity.  James Lindemann Nelson has defended this model against the consensual model by drawing an analogy to the responsibility that comes from causing unintentional harm to others.  While conceiving a child is not a harm, a related principle that causal contribution is sufficient for responsibility would explain why unwilling fathers bear financial obligations to their children: they are a proximate cause of the creation of the child.  This view has a few odd implications, however, indicating that it proves too much. As Nelson accepts, it indicates that sperm and egg donors can be held financially liable for the well-being of resultant children, even if the non-biological parents consent to waive such responsibility (the duty is to the children, not the non-biological parents, and so the non-biological parents don’t have the right to waive the duty).  Perhaps more disturbingly, it implies that fertility doctors involved in in vitro fertilization are financially responsible for the well-being of the resultant child.  This is because doctors did proximately cause the existence of the child, and their actions can, in conjunction with the contributions of the mother and father, be considered roughly sufficient for the creation of the child.  But these implications appear somewhat absurd, at least as radical and potentially unjust as Shrage’s proposal to obviate paternal financial duties.

A natural move at this point is to advert to biological relation as the basis of obligation.  It may not necessarily release sperm and egg donors from duty (at least in its general form), but it at least excludes fertility doctors.  And it has strong intuitive support, aligning well with related familial duties to (even long-lost) parents and siblings. This may be the most intuitively attractive option, but it faces a larger problem of philosophical justification.  What, exactly, is it about biological relation that can plausibly be said to matter morally?  Mere genetic similarity is itself intuitively implausible, as that would imply one would have strong parental obligations in cases of freak genetic coincidences where unrelated individuals have a very similar genetic profile.  Moreover, the fact of genetic similarity does not itself seem very morally important – perhaps it is relevant in evolutionary terms, but it is not clear why we should care about it. One could rely on a more specific kind of biological process – participation in sexual intercourse resulting in impregnation and childbirth – but again, it is not clear what makes that morally relevant independent of general causal considerations.

Upshots

             Each option for defending the financial obligation of unwilling fathers, then, faces significant philosophical difficulty.  Communal and consensual accounts could not justify the obligation, while causal and biological stories either are counterintuitive or lack convincing grounding themselves.  At the same time, Shrage’s proposal is itself counterintuitive.  Something has to give.  So where do we go from here?  If one has a strongly intuitionistic approach, the biological model (where participation in a process of sexual intercourse resulting in pregnancy and childbirth is sufficient for obligation) is probably the best option.  While it might lack in philosophical grounding, there are relatively few counterintuitive implications (at least that I can think of).  This would support the obligations of unwilling fathers, and personally it is the option I am most attracted to.  But a more theoretical approach may well support Shrage’s account, insofar as the consensual and communal models enjoy stronger theoretical support based on autonomy and political theory, compared with causal and biological models that rely more on intuitions about cases.

            Whatever direction one takes, one should be upfront about the intuitive and philosophical costs of each approach, which may well have implications beyond the question of whether unwilling fathers should be held financially responsible for their children.

At the moment, prosthetic devices like Meyer’s are used to restore normal human functions among those who lack them. Yet as such devices become ever more sophisticated, to the point that they eventually outperform “natural” limbs in terms of speed, strength, executive control etc., “will it become the norm to have one of these”, Meyer asks? Also, as the author of the Observer editorial worries, “what happens when these technologies and machines get so smart that humans can be written out of the equation altogether?”. For instance, what if we could simply turn to our iPhone rather than a human doctor to get a diagnosis for our ailments as well as appropriate treatment recommendations? Such suggestions can elicit fears of a dystopian future where humans are pressured to become “cyborgs”, whether they like it or not, if they are to remain competitive on the job market (including competitive sports) and in other contexts; or where they are increasingly made obsolete by more effective machines, and real-life human interaction is reduced (machines replacing staff at supermarket checkouts, but also GPs, etc.), and becomes less accessible than it is now (think of having to pay a significant premium to see a human doctor).

Critics of enhancement technologies have correctly emphasized such concerns, yet they have also tended on that basis to endorse a general anti-enhancement stance, at complete odds with transhumanist thinkers. In light of the potentially immense benefits of these technologies (including when it comes to preserving the existing goods that we, including enhancement critics, properly value), such a radical stance strikes me as untenable. Suppose for instance that machines eventually proved able to diagnose illnesses more accurately, or to conduct medical research more effectively, than even the most competent humans can. Were this so, isn’t there a case to be made that we should then step aside and let them do the relevant work, even if this meant losing to them certain intrinsically valuable occupations? (This wouldn’t apply to occupations in relation to which achieving successful outcomes as efficiently as possible doesn’t seem to be a moral imperative; furthermore, professions like doctor also involve a social function which it isn’t clear we should want to delegate to machines.)

Secondly, let us assume that, in contrast to some transhumanists, we prefer our current fleshly envelope to a cybernetic form of embodiment. This assumption is still compatible with supporting various forms of enhancement, such as radical life extension, to the extent that it preserves our fleshly envelope when decrepitude and death would have destroyed it. Also, preferring forms of life extension that preserve our current bodies over those involving cyborgization is compatible with choosing cyborgization if the only alternative is destruction. Finally, such preferences largely seem to depend on aesthetic considerations that will no longer warrant them if we consider enhancements that do not affect the person’s appearance, or the phenomenology of her experiences, in a negative way (or at all). Some of us may not like the idea of replacing our natural hands with bionic ones, but why should we feel the same about getting an artificial “enhanced” heart were this to become possible, if the only consequence of doing so were to prolong our healthy lifespan? Concerns about fair access, mentioned by both Bostrom and Miah, would then still apply, yet it seems that these could in principle be addressed without refraining from developing the technology, or trying to enforce a ban on its use.

It is to be hoped that events like FutureFest will help move the enhancement debate beyond the overly simple, radical opposition between “pros-” and “anti-” towards a more nuanced and constructive assessment of those technologies that takes into account both their great promise and potential perils – ideally resulting in suggestions as to how to fulfill that promise while minimizing the perils.

—

Alex Erler is a research associate at the Oxford Uehiro Centre for Practical Ethics. You can find his academic page here and his Twitter page here.

As it happens, Swansea is my hometown, and I’ve been working on a project related to vaccine ethics for the past few months. So, the ethical issues raised by the measles epidemic in Swansea are especially interesting for me. In particular, I’ve been addressing the following two issues. The first issue is whether it is impermissible for parents to refuse the MMR vaccine for their children. The second is whether the parents who refused the MMR vaccine should be blamed for the refusal.

The first issue might seem uninteresting. After all, it is uncontroversial to hold that, unless there are other weighty considerations to the contrary, it is impermissible for a parent to allow their child to be harmed. Are there any plausible weighty considerations to the contrary? This is tricky. First off, let’s set aside cases where the child has an allergy to the MMR, and assume that it is permissible to refuse the vaccine in such cases.

One possible consideration that could render an MMR refusal permissible is the value of autonomy. Although there are many accounts of what the value of autonomy entails, broadly speaking, it is the idea that it is good to make and act upon significant decisions about one’s own life, according to one’s own beliefs about what is valuable. Given that a relationship with one’s child is a significant part of a person’s life, it could be argued that the value of autonomy means that it is permissible for parents to refuse autonomously the MMR vaccine for their child, even though this refusal is bad for their child.

However, it seems unlikely that autonomy could render this refusal permissible for a number of reasons. One reason is that it is unclear why the importance of autonomy should outweigh concerns for the health of the child. After all, even if what brings significant meaning to my life involves autonomously planning to punch a stranger in the street, it seems implausible to claim that the value of autonomy renders the action that causes this harm to the stranger permissible. Yet refusing the MMR vaccine could result in far greater harm than my punch (there are, of course, a number of responses to this).

A further possible consideration that could render a parent’s MMR refusal permissible is the value of the parent-child relationship. As Brighouse and Swift have argued, this value makes a significant contribution to the lives of parents. Moreover, it is argued to be a unique source of flourishing. Nevertheless, it is difficult to see how a parent refusing the MMR vaccine for his/her child could have any bearing on the value of the parent-child relationship. Perhaps it could have some bearing in cases where a parent holds a religious view that forbids vaccines. But even in these cases, it is difficult to see why the importance of realizing this value would outweigh the value of the child’s health.

So, I think that neither considerations of autonomy nor the value of the parent-child relationship are convincing grounds to claim that it is permissible for parents to refuse the MMR vaccine for their child. There may be other, more persuasive considerations that I haven’t mentioned, and if you have any suggestions, let me know.

It could be objected that all of these points miss the most important issue about the parents’ refusal. It could be claimed that it is permissible for parents to refuse the MMR vaccine for their child if the parents believe that the refusal will be good for the child, even when the refusal will be bad for the child. It is very likely that in the cases where parents refused the MMR vaccine, they did so because they believed that the refusal will be good for their child. There are good grounds for doubting that a person’s beliefs affect the permissibility of his actions, but I don’t have space to get into this here (but if anyone knows a compelling argument why beliefs can render otherwise impermissible actions permissible, let me know). However, an agent’s beliefs are important when we assess whether a person is blameworthy for their impermissible actions, and it is not clear whether the parents who refused the MMR vaccine are blameworthy.

Whether we consider someone to be blameworthy for an impermissible action is important. For example, many philosophers think that it is unjust to punish someone who is not blameworthy for an impermissible action. A sufficient condition for a person not to be blameworthy for an impermissible action is an excuse. If an impermissible action is fully excused, then we should not blame the person who acted impermissibly.

Did the parents who refused an MMR vaccine for their child have an excuse? There are a number of conditions that determine whether a person is excused for an impermissible action. Probably the most relevant excusatory condition in this case is ignorance. Suppose that a person does not know that his action has harmful consequences. It is commonly held that this ignorance can excuse this harmful action.

So, given that (at least some of) the parents who refused the MMR vaccine did not know that their MMR refusal would result in harm in their child, does this meet the excusatory condition of ignorance? It seems plausible to think that they do meet this condition, and so (at least some of) the parents should be excused and not blamed for refusing the MMR vaccine. (However given the widespread publicity surrounding this epidemic, it is unlikely that future impermissible refusals would meet this excusatory condition). Nevertheless, it could be objected that parents should have known that refusing the MMR vaccine for their child could result in the harm of their child. But this does not seem clear. After all, the parents were faced with a variety of conflicting information concerning the safety of the MMR vaccine, both in a highly respected academic journal, and in their local newspaper. Sadly, there are a number of further responses and objections that I don’t have the space here to address, but let me know if you have any thoughts. I would greatly appreciate them!

[updated version]

Sarah Murnaghan is a 10-year-old from Pennsylvania. Suffering from cystic fibrosis, she was likely to die without a lung transplant. Her situation was deteriorating. But because of a rule that says that children under the age of 12 have the lowest priority for adult donor lungs, Sarah would have to wait for another child’s lungs to become available, a much rarer occurrence.

Sarah’s parents sprang into action. They got the attention of members of congress and the media. They shared Sarah’s story on social networking sites, showing pictures of their daughter in the hospital bed. They said that the “Under 12” rule was discriminatory against children, and got a federal judge to agree. So, with the help of a court order temporarily preventing the enforcement of the Under 12 rule, Sarah got a second chance at life. An adult lung match became available, and Sarah is now recovering from transplant surgery.

It’s a story with a happy ending—depending upon how you tell it. Certainly the news is good for Sarah. Yet as Sarah’s mother acknowledged in a post on Facebook, “We … know our good news is another family’s tragedy.”

But who are those families? What are their stories? What are the names of those who will die—or who have already died—without a lung transplant of their own?

What this case illustrates is something we might call “pushethics” – a way of pushing one’s own story, or that of one’s family member, into the moral spotlight. Since ordinary human beings—from news anchors to congressmen to federal judges—are more likely to feel empathy for known individuals with compelling narratives of suffering, they can become motivated to bend the rules in favor of those specific individuals whose stories best capture their attention.

The basic psychology involved is well-known. As Peter Singer famously argued with his case about a child drowning in a pond, no decent person could fail to rescue someone who was close at hand—whose need was evident, and whose suffering could be relieved by a minimum of well-directed effort. But we’re much less likely to take actions that could save the life of a child far away, even if the cost of doing so would be minimal—such as by donating a small amount of money to an effective charity—because that child’s suffering is not as salient to our emotions. We didn’t evolve in an environment in which our actions could bring benefit over long distances (or stretches of time) and so our intrinsic moral motivations are constricted. By the same token, anonymous people pull at our heartstrings much less powerfully than do named individuals.

(In fact, one way that charities have learned to increase donations from the public is to accompany their requests with photographs of specific children who are suffering from, e.g., malnutrition. This is different from pushethics, however, in that it’s not done to advantage any particular child, but rather all of the children who might be helped by an increase in charitable donations.)

But is there anything wrong with engaging in pushethics? Wouldn’t anyone with a dying loved one want to shout that person’s story from the rooftops, doing anything they could to secure help? And aren’t the friends and parents of those “anonymous” people just as free to harness public empathy to their side?

The ethics of pushethics comes down to a couple of key factors. First, there is nothing wrong per se with reaching out to the news media, or to congress, or to Facebook and Twitter to bring attention to the suffering of someone you love. And if some injustice—some discriminatory rule—is standing between your child and a life-saving operation, then to fail to challenge this rule would be condemnable. But what if the rule is fair? The problem with pushethics, when all is said and done, is that it unfairly disadvantages someone else who is equally in need. The problem is with zero-sum games.

As Kathleen Sebelius, the U.S. Secretary for Health and Human Services, stated at a congressional hearing on the Under 12 rule: “Unfortunately there are about 40 very seriously ill Pennsylvanians over the age of 12 also waiting for a lung transplant.” If an adult set of lungs goes to a 10-year old child, then it can’t go to an older patient who would have been a more secure match.

Of course, it may be that the Under 12 rule places too much weight on maximizing prognosis. It may be also be too arbitrary, needlessly precluding case-by-case analysis. And, as Sarah’s parents argued in court, the rule may indeed be unfair to children. This sort of debate needs to happen. If the rule does turn out to be faulty, however, it needs to be replaced with a fair procedure that will determine who will live and who will die—no special pleading. There are, sadly, not enough organs to keep everyone alive. A fair procedure will mitigate the force of pushethics—allocation according to attention, priority to the emotionally endearing or the better-known.

Another problem with pushethics is that it doesn’t cure the underlying disease—in this case, the basic shortage of organs. We consistently fail to fix this problem because we don’t personally see the people in need actually dying before our noses. So we choose to burn or bury our life-saving organs (after our own or our children’s death) because it is too painful, too time consuming, or too bothersome to think about the suffering of anonymous people—even as they gasp for breath, or slowly turn blue and then black.

Unless we see them on the TV.

If it weren’t for our profound moral and psychological limitations—the ones that generate as well as react to the deceptive allure of pushethics—there would be no need for a fair allocation procedure. For this reason, we need ethical policy. People should donate their organs. Burying them in the ground or burning them up—when they could save a person’s life—is not only unconscionable, it is as good as murder.

Implicit racial bias is an evolved, unconscious tendency to feel more positively towards members of one’s own race (one’s ‘in-group’) than towards members of a different race (members of an ‘out-group’). The bias can be (and was in this study) measured using a version of the implicit association test, which requires participants to quickly catagorise faces (black or white) and words (positive or negative) into groups. Implicit bias is calculated from the differences in speed and accuracy between categorising (white faces, positive words) and (black faces, negative words) compared to (black faces, positive words) and (white faces, negative words). Crucially, implicit racial bias has been shown to be uncorrelated with explicit racial bias – self-reports of negative racial stereotypes. This means that even those who are not consciously averse to people from other racial groups often demonstrate a deep-seated bias against them as an evolutionary hangover. Hearteningly, the authors of the study started from the idea that encoding people by race may be a reversible by-product of human evolution used to detect coalitional alliances. What their study confirmed is that immersive virtual reality provides a powerful tool for placing people into a different race ‘coalition’ by changing their body representation and consequently reducing their implicit aversion to the racial characteristics there represented.

Clearly, reducing the existence and associated behavioural effects of the implicit racial bias would be a good thing. Especially where people who exhibit it (which, (no)thanks to evolution, is most of us) are tasked with making objective and fair decisions about things like employment and justice. However, whether we should classify such an intervention as a moral enhancement will be up for discussion. Whilst the effects are clearly good for society – reduced existence/evidence of deep-seated racism – the mechanism by which this is achieved might be seen to preclude conceptualisation of the outcome as an instance of moral enhancement.

The first concern might be that the intervention changes behaviour via sub-conscious processes: the reduction in evidence of the bias is not achieved by, for example, the individual spending time reflecting on the equality of all people and the importance of treating everyone fairly – it is, too a large extent, out of his control. Two contemporary philosophers disagree on the importance of cognition and conscious reasoning to moral enhancement.

Tom Douglas (2011: 162) understands moral enhancements to be ‘interventions that will expectably leave an individual with more moral (viz ., morally better) motives or behaviour than she would otherwise have had’. Douglas thus allows that moral enhancement could consist in the moral improvement of behavior even where there is no moral improvement in motives.

In response, John Harris (2012: 172) has argued ‘that moral enhancement, properly so called, must not only make the doing of good or right actions more probable and the doing of bad ones less likely, but must also include the understanding of what constitutes right and wrong action…if, “once the enhancement has been initiated, there is no further need for cognition”, then the morally enhanced action is effectively automatic, unconscious and therefore unintended, entirely outside the realm of moral responsibility.

The sub-conscious effects of the immersive virtual reality intervention expectably leave an individual with morally better behavior (so qualify as moral enhancement on Douglas’ account) but these effects are achieved via automatic, unconscious processes (which Harris believes to be inadequate for moral enhancement, as the individual is not responsible for them). Whilst we might think that reductions in discriminatory behavior are more praiseworthy if they result from conscious effort, the problem with the implicit racial bias is that it seems to be immune to attenuation by reason. This is suggested by the data which shows that explicit racial bias is not correlated with implicit racial bias: people who would be genuinely upset by evidence of their implicit bias – who would disown it – nonetheless often still demonstrate it. If the implicit racial bias is not a product of beliefs or reason, it is unlikely that it can easily be attenuated by meditating on one’s believes. So we may – particularly where evolutionary biases are concerned – need sub-conscious interventions to enable us to act more in accordance with our understanding of what constitutes right and wrong action. In such cases, it is not our motives that need improving but our ability to act in consistency with these motives.

However, there may be a further worry to do with the precise nature of the outcome of the intervention.  Whilst the outcome looks like an increase in acceptance and positivity towards others of a different race – something we would see as a virtue – an alternative hypothesis might be that the intervention simply extends one’s concern for oneself – something less virtuous. The psychologists report that they anticipated achieving the effect by inducing a ‘body-ownership illusion in a differently raced avatar’. The implication of perceived ‘body-ownership’ might be that I extend the conception of myself and, correspondingly, my ‘in-group’ so that I also respond positively to my illusory characteristics. If so, it is still a bias towards myself and those similar to me – I’ve just been deceived into thinking that dark-skinned people are similar to me.

But perhaps this is exactly what racial acceptance should be: a blurring of the boundaries of groups. Rather than insisting – for moral virtue – on some persisting conception of ‘out group’ or ‘others’ so that I learn to accept these others despite their difference from myself, I just stop seeing them as different. Given the nature of the implicit racial bias and its frequent incongruity with the racial attitudes and beliefs individuals consciously hold, this sort of intervention may be the only way to attenuate it. The possibility that it does so by causing me to expand the ‘in-group’ that I favor is a step in the right direction even if this necessarily involves some degree of narcissism. Our moral psychology – replete with biases and limitations – is a product of evolution, after all. Not that this is in any way an excuse; rather, it generates a duty to try to overcome these limitations through reflecting on our attitudes, but maybe also though other, less deliberative means. Interventions that reduce implicit racial bias should be pursued if they are shown to have effects persistent enough to make real differences to rates of discrimination. It would certainly be interesting to see what effect a ten-minute-a-day skin swap for judges would have on racial discrimination in our courts.

References

Douglas, T. (2011), ‘Moral enhancement via direct modulation: a reply to John Harris’, Bioethics, 27 (3): 160–168.

Harris, J. (2012), ‘Ethics is for Bad Guys! Putting the “Moral” into Moral Enhancement’, Bioethics, 27 (3): 169–173.

The allegation is that consciously, or sub-consciously, the decision to prosecute Africans is driven by racial prejudice.  Now, this may or may not be true.  There may be all sorts of other reasons why almost all the indictees have been African.  However, let us suppose that it is true – that racism explains why Africans have been the almost-exclusive focus of the ICC.  Certainly many in Africa believe this to be the case.  And many take the further step of believing that therefore the ICC should be dismantled.

 A comparable issue arises in other contexts.  Thus, it is often said that the movement to boycott citizens and products from Israel is driven by anti-Semitism, since – critics of a boycott point out – there are countries with far worse human rights records than Israel, and these countries are not the target of boycott campaigns.  The additional claim is then made that this is a reason to oppose any boycott of Israel. 

 But is it? 

Let’s return to the ICC example.  Our motives for action are usually multifaceted.  Even if we accept that there’s some sort of bias in the ICC, it’s implausible to believe that it’s the only (or indeed main) motivation.  At least part of the motivation is the admirable quest for justice.  Everybody (or virtually everybody) believes that those who have committed war crimes should be caught and tried and punished.  So an honourable and a dishonourable motivation is combining to produce an outcome we approve of – that at least some tyrants are held to account. 

Nonetheless, moral inconsistencies in the actions of the ICC, or in boycott calls, may highlight prejudice and in so far as they do so, they stain the organization or movement in which the prejudice is exposed.  

The best remedy is the application of consistency.  Thus, with regard to the ICC there ought to be a non-prejudicial pursuit of human rights abusers in every corner of the world.  And were we to be convinced that a boycott against Israel was justified and would be effective in achieving certain worthy aims, then the response to critics who allege anti-Semitism would be a call for a boycott of other places where we believed this too would be legitimate and effective.

In the meantime, we have apparent inconsistencies.  One can imagine a world in which these were so glaring that we’d demand the ripping down of the whole edifice.  The ICC will only ever target a tiny number of people – those who are accused of the most heinous crimes.  We want mass murderers behind bars even if they are put behind bars partly because of their race.  But we couldn’t morally tolerate this for the entire criminal justice system, involving far more trivial crimes.  It seems to me that it would be better that no shoplifters were pursued by the courts than that only black shoplifters were. 

The ICC is not like this.  The balance is tipped by the seriousness of the crime.  It would not be better to have no war criminals behind bars than only Africans.     

That does not mean they are unimportant: we are at an important choice-point in regard how to handle mass surveillance. But the battle is not security versus freedom, but secrecy versus openness.

It is not obvious whether we should wish for better surveillance or less: it is not clear it is a force for good or evil in general. Surveillance can prevent or solve crimes, alert society to dangers, provide information for decision-making and so on. It can also distort our private lives, help crime, enable powerful authoritarian and totalitarian forces, and violate human rights. But there doesn’t seem to exist any knock-down argument that dominates the balance: in many cases messy empirical issues and the current societal context likely determines what we should wish for, did we know the full picture.

But what is obvious is that that unaccountable surveillance is much easier turned into a tool for evil than accountable surveillance: the key question is not who got what information about whom, or even security versus freedom, but whether there is appropriate oversight and safeguards for civil liberties. And the only answer that comes out of the Obama administration (or any other government) is “of course!”. Unfortunately this is not convincing, since it is what an untrustworthy or incompetent government would say too.

A NSA spokesperson told The Guardian:

The continued publication of these allegations about highly classified issues, and other information taken out of context, makes it impossible to conduct a reasonable discussion on the merits of these programs.

This is entirely true. Unfortunately a reasonable discussion on the merits of the programs requires the discussants to have relevant information. The secrecy surrounding them makes it impossible to have a public discussion. Either it is done among fully informed people privately, or it is done publicly and the secrecy will be broken. The private discussion does not inspire trust since there are valid concerns about regulatory capture. The public discussion will be incomplete and bound to be biased unless most information is revealed, negating secrecy. In both cases it will never be clear that all relevant information has been revealed unless it is possible to go into the secret domain and at least do spot checks.

In order to actually be convincing the oversight needs to demonstrate that it (1) doesn’t suffer regulatory capture, (2) can access the information needed, (3) actually uses it to make relevant decisions. The general perception is that many elected leaders knew and seem to have accepted widespread surveillance, yet – given the current denials and confusion – either do not know the full picture or are captured in various ways. Even if this is an untrue picture it undermines trust in the legitimacy of the oversight system: who watches those watchmen? In the past we tended to trust authorities by virtue of them being authorities or having excellent character, but this default trust has been eroded. We recognize how power corrupts and how incompetence is everywhere, and few government officials today have reputations as incorruptible – especially when dealing with very powerful vested interests we can at least suspect could apply significant pressure against perceived opponents.

The way out is to add transparency and verifiability. James Clapper of the NSA actually gestured along these lines by declassifying some documents to argue that the ‘the program operates “within the constraints of law” and “appropriately protect[s] privacy and civil liberties”.’ Whether a selective declassification is enough to convince is another matter.

The problem with the view that national security overrides all other considerations is that it makes itself impossible to criticise: evidence and procedure must be secret, why they must be secret is secret, and so on. We cannot know whether the tradeoff is right because we are not allowed to see the effectiveness.

Even in a perfect world this would block the openness of society: open societies work because citizens can criticise any part of the system, demanding accountability, and the system itself can be changed to accommodate this if there is enough support for it. This is how mistakes and corruption get exposed and corrected, this is how the society is reshaped to fit the citizens rather than to fit some minority plan. It might not be quick, neat or easy, but it is a self-repairing and self-modifying system. But if there are aspects of the society that cannot be criticised or changed, then those are excluded from these mechanisms. Since mistakes happens even when people are dedicated and competent, even in the ideal world closed parts of society run the risk of becoming faulty. Add the realistic components of people covering up embarrassment, the possibility of corruption, regulatory capture and the existence of individuals with problematic agendas, and the existence of closed parts of societies become much more problematic. If they are also strongly empowered – legally and technologically – they become potentially very dangerous, no matter how noble the initial intentions were.

It is not hard to imagine, for example, how a well-meaning institution might fall into the trap Janet Radcliffe-Richards described in her Uehiro lectures that characterises politically correct thinking: given that you aim at a morally good thing, you become averse to accepting empirical findings or arguments that disconfirm this, or even discourage attempts at investigate such things. This is an is/ought mistake, but it makes the institution regard attempts at investigating or curtailing its powers as attacks on its well-meaning intentions. Hence, in order to safeguard those, the attempts must be thwarted and the groups attempting them also become suspect: what could they have against these good intentions? With powerful surveillance it will not be hard to find evidence that supports such suspicions…

Intelligence analysts will no doubt dislike this caricature: those I have met are remarkably devoted to reducing bias and actually finding true facts to guide sane decisions. But my experience with the information ecology of large institutions have also shown that cognitive biases can thrive even where the individual members are trying to avoid them – especially if management structures are not de facto rewarded for truthfulness and neutrality but rather for bureaucratic survival skills. It is easy to be overconfident in the niceness of ones own organisation. This is yet another reason oversight and transparency is needed.

Closedness sometimes bites itself. The kafkaesque FISA requests to companies for surveillance data make them unable to discuss the requests. One side effect is that now nobody will believe the denials from tech companies: if it is illegal to disclose surveillance, then no amount of denials, no matter how plausible, will ever assuage our concerns. The irony is that real regulations actually make conspiracy theory logic look reasonable. Conspiracy theorists think the absence of evidence is evidence for a cover-up: here the do-not-discuss orders make any denial pointless (not quite the same thing, but still a strong reducer of trust).

Even from a security standpoint these technologies are double-edged: while they allow amassing massive information invisibly, that doesn’t mean they will only deliver it to the intended recipients. The Petraeus scandal demonstrated that government officials are also ensnared in the net, even when they try to avoid it. Foreign governments have no doubt exploited mandatory eavesdropping functionality in telecom systems against not just their own citizens but also the interests of governments that mandated the systems in the first place. And it is not implausible that they provide tempting targets for many non-governmental groups, who can untraceably gather information for their own agendas. Since their nature and use cannot be freely discussed and analysed the problems they cause cannot easily be corrected.

When the government says law biding citizens have nothing to fear from government surveillance, the rejoinder is of course that law abiding governments have nothing to fear from transparency. Expansion of surveillance power must be balanced by an equal or larger expansion of transparency and accountability. Expansion of secrecy must be balanced by even more accountability – secrecy is in many ways, no matter how useful it can be, a more dangerous tool than surveillance.

Over about 14 months, Harry Kakavas lost $20.5 million in a casino in Melbourne. It could have been worse. He put about $1.5 billion on the table. He sued the casino. It knew or should have known, he said, that he was a pathological gambler. It shouldn’t have continued to take his money. It should have protected him from himself. Nonsense, said the High Court of Australia.

Here’s why:

‘Even if, contrary to the findings of the primary judge, the appellant did suffer from a psychological impairment, the issue here is whether, in all the circumstances of the relationship between the appellant and Crown, it was sufficiently evident to Crown that the appellant was so beset by that difficulty that he was unable to make worthwhile decisions in his own interests while gambling at Crown’s casino. On the findings of fact made by the primary judge as to the course of dealings between the parties, the appellant did not show that his gambling losses were the product of the exploitation of a disability, special to the appellant, which was evident to Crown.

Equitable intervention to deprive a party of the benefit of its bargain on the basis that it was procured by unfair exploitation of the weakness of the other party requires proof of a predatory state of mind. Heedlessness of, or indifference to, the best interests of the other party is not sufficient for this purpose. The principle is not engaged by mere inadvertence, or even indifference, to the circumstances of the other party to an arm’s length commercial transaction. Inadvertence, or indifference, falls short of the victimisation or exploitation with which the principle is concerned.‘ (paras 160-161 of the judgment).

So it all turned on findings of fact (it wasn’t ‘sufficiently evident’ that his losses were the result of a disability, and if they were, they weren’t the product of a disability ‘special to the appellant.’)

That last criterion is interesting. The court seems to be implying that everyone who puts themselves in the position of losing large amounts of money in a casino is necessarily not quite right in the head. To establish liability you need a degree of vulnerability over and above that possessed by the ordinary punter. By accepting the trial judge’s finding that Kakavas did not suffer from a ‘psychological impairment’, the court was presumably saying: ‘Right: so Kakavas is weak and easily exploited: but that’s true of everyone who walks through the door, buys some chips and sits down at the table. That sort of weakness is within the general bell curve of human flabbiness. But Kakavas wasn’t particularly, dramatically, visibly weak.’

Now link this with the ‘sufficiently evident’ criterion. It’s hard to understand why it wasn’t ‘sufficiently evident to Crown ….that [Kakavas] was unable to make worthwhile decisions in his own interests while gambling….’ Didn’t Crown know how much he’d lost? Did it really think that it was cool calculation that kept him at the table? Didn’t it know that it had an addict on its hands – and an addict who was doing something that was plainly against his objective best interests?

No doubt there’s a thrill in doing something that is plainly against one’s objective interests. No doubt it’s partly that thrill which takes people through the doors of casinos. The figures say that you’re probably going to lose.

The law should be very slow to interfere with people’s right to do stupid things. It should be slow to interfere when the purported justification for truncating individual autonomy is the vindication of wider societal interests, and even slower when the justification is the protection of the individual against himself. And by and large it is. But sometimes it steps in – for instance in R v Brown ¹(to criminalise entirely consensual sado-masochism), or in the enforced Caesarean section cases (where part of the rationale is to give the woman something that, at some level of herself, she really wants). In Caesarean section cases the judge, having behaved in a way dubbed by the woman’s counsel in argument as oppressively paternalistic, often gets a gushingly grateful letter from the woman, enclosing a picture of the beaming baby.

It might be said, in the gambling context, that the fact that casinos are unashamed predators on human frailty should make the court more ready to intervene. I’m not sure that’s right. The predatory nature of the casino is a necessary element of the pleasure that the punter gets. If you can’t lose big time, you don’t get the buzz you want. And there should be a strong presumption in all liberal societies that that buzz is itself legitimate.

So: the solution isn’t to lambast the casino. The solution is to make the casino the trusted agent of a benevolently paternalistic society. Crown should have said: ‘Enough: the prospect of you recouping these plainly devastating losses is remote. To continue to play is pathological.’ It’s not necessary, to come to that conclusion, to characterise it as unusually pathological, even by big gamblers’ standards, to stay at the table.  

1. [1994] 1 AC 212. See too Laskey, Jaggard and Brown v United Kingdom (1997) 24 EHRR 39

The moral question here turns on the empirical question of whether having a pub in a motorway service station would lead to an increase in the number of motorists drinking in a manner that would compromise their ability to drive safely. Since this venture is the first of its kind, any answer to this question must necessarily be somewhat speculative. However, it seems that we can consider certain factors which might be thought to bear on the question of whether we could reasonably expect the presence of a pub in a motorway service station to increase drink-driving.

First, we might worry that the main market that Weatherspoon’s is targeting in opening a pub at a service station is drivers. However, this is not necessarily the case; service stations are also visited by a large number of non-drivers, in particular, by passengers on long haul coach journeys. As such, it is not the case that Weatherspoon’s are specifically targeting drivers in a manner that might give us cause for concern.

A related worry might be that the mere presence of a pub would serve to tempt drivers into drinking when they previously would not have done so. Again though, it is not clear why this must be the case. After all, there are already a large number of pubs near or on Britain’s many ‘A’ roads, yet we do not think that they present a source of irresistible temptation to motorists. Moreover, a recent change on law in the UK allows supermarkets to sell alcohol in motorway service stations.

Perhaps the most plausible worry about opening a pub in a service station is the message that it might be understood to convey about drink-driving. Here, there seems to be  an important disanalogy between having a pub in a service station and allowing a pub near to an ‘A’ road. It seems that the primary purpose of service stations is to provide motorists with a place to rest from driving; they are not intended as a place to visit in the same way that a country pub might. With this in mind, there might be some credence to the claim that allowing a pub in a service station serves to legitimise drinking as part of the motorist’s recuperation during long drives; if so, this does seem to represent a way in which the presence of a Wetherspoon’s pub might serve to incite more people to drink whilst they are driving.

Moreover, we might also point out that there is a disanalogy between having a pub in a service station and selling alcohol in a service station supermarket, in so far as making alcohol available in a ‘pub’ setting would normally seem to make the availability of alcohol far more salient to the motorist than merely selling it in a service station supermarket.

This is not intended as a knock-down argument against the development. As I mentioned at the outset, it is an empirical question as to whether the presence of a pub in a service station would increase the number of motorists drink-driving, and the devil will very much be in the detail. However, I believe that the above represents one of the more plausible reasons for why the presence of a pub in a service station might possibly increase the numbers of motorists drink driving.

Many of these issues have been discussed on this blog and elsewhere, but it is a nice comprehensive review of a number of issues brought up by the new technology. And while the machines do not yet have fully autonomous capabilities the distance to them is chillingly short: dismissing the issue as science fiction is myopic, especially given the slowness of actually reaching legal agreements. However, does it make sense to say that robots should not have the power of life and death over human beings?

It is an underlying assumption of most legal, moral and other codes that when the decision to take life or to subject people to other grave consequences is at stake, the decision-making power should be exercised by humans. The Hague Convention (IV) requires any combatant “to be commanded by a person”. The Martens Clause, a longstanding and binding rule of IHL, specifically demands the application of “the principle of humanity” in armed conflict.68 Taking humans out of the loop also risks taking humanity out of the loop.

Heyn’s point here seems twofold: there is a personhood aspect which links it to moral agency and responsibility (I doubt even a radical bioconservative would accept a weapon fired by signals from a fertilized human ovum, no matter if he thought the cell was a moral person), and a humanity aspect linked to being civilized and having a conscience.

Robots currently fail at both, the first due to their lack of intelligence and inability to change their goals, and the second by virtue of being non-human. Future robots might perhaps be able to fulfill both (there are of course some AI skeptics that think this is not possible even in principle, and intelligence alone  does not automatically resolve the moral agency or value factors) but would then essentially be like normal soldiers. Heyn follows Peter Asaro in arguing that non-human decision-making regarding the use of lethal force is inherently arbitrary, and all resulting deaths are hence arbitrary deprivations of life.

Machines lack morality and mortality, and should as a result not have life and death powers over humans. This is among the reasons landmines were banned.

Leaving aside the mortality issue (would Superman be banned from the battlefield?) the landmine case shows an interesting problem with this otherwise rhetorically convincing approach. Would landmines become better or worse if they tried to distinguish between combatants and non-combatants? It seems that adding this function would make them more acceptable, even if they don’t reach a threshold of being morally permissible. The extra autonomy reduces the arbitrariness and might improve the proportionality. Non-human decisionmaking can be non-arbitrary if implemented right: an autonomous car that refuses to run over a person in front of it is overall better than a car that just charges ahead. Machines can act as moral proxies of their programmers or users.

The main headache is where the moral buck stops. Drone warfare has already demonstrated problematic diffusion and separation of responsibility: an individual might be formally responsible for firing decisions, but they are embedded in a techno-social system that dilutes both moral intuitions about individual responsibility and the nature of combatants. Autonomous machine warfare moves things even further: now the responsibility might be diffused not only across the institution fielding the system but also the companies creating the technology. In many cases misbehavior will not be due to any discrete mistake in any part, just a confluence of behaviors and assumptions that produce an emergent undesirable result.

Can we run warfare without anybody being responsible? I do not claim to understand just war theory or the other doctrines of ethics of war. But as a computer scientist I do understand the risks of relying on systems that (1) nobody is truly responsible for, (2) cannot be properly investigated and corrected. Since presumably the internal software will be secret (since much of the military utility of autonomous systems will likely be due to their “smarts”) outside access or testing will be limited. The behavior of complex autonomous systems in contact with the real world can also be fundamentally unpredictable, which means that even perfectly self-documenting machines may not give us useful information to prevent future mis-behaviors.

Getting redress against a “mistake” appears far harder in the case of a drone killing a group of civilians than by a gunship crew; if the mistake was due to an autonomous system it is likely that the threshold will be even higher. Even from a pragmatic perspective of creating disincentives for sloppy warfare the remote and diffused responsibility insulates the prosecuting state. In fact, we are perhaps obsessing too much about the robot part and too little about the extrajudicial part of heavily automated modern warfare.

One interesting issue that has not been raised as far as I am aware is the problem of internal responsibility. If a soldier acts wrongly or contrary to orders he is held responsible; organized misbehavior is treated even more harshly. However, autonomous systems might misbehave in ways that cannot be assigned to a responsible party. Worse, diffusion of responsibility might occur inside military forces – if an autonomous weapon system decides to act, to what extent can its “superiors” be held responsible for its actions?

Since a commander can be held accountable for an autonomous human subordinate, holding a commander accountable for an autonomous robot subordinate may appear analogous. Yet traditional command responsibility is only implicated when the commander “knew or should have known that the individual planned to commit a crime yet he or she failed to take action to prevent it or did not punish the perpetrator after the fact.” 58 It will be important to establish, inter alia, whether military commanders will be in a position to understand the complex programming of LARs sufficiently well to warrant criminal liability

Perhaps the strongest reason to believe military forces will want to have a firm leash on their robots might simply be that otherwise they will risk their internal chain of responsibility.

In the case of humans we learn the “programming” and intentions by having shared human experience, by interacting with them, and by giving instructions that are expected to be understood. The problem with complex autonomous systems is that this is far less possible. Indeed, the humanity aspect mentioned at the beginning is crucial: it allows us to make reasonable inferences about what other agents are up to, but non-human systems will be quite alien and hence unpredictable.

One can make an analogy to use of attack dogs. (The Shakespeare quote alluded to in the title does actually not refer to real canines but to ordering soldiers to pillage and sow chaos, something that today would presumably be seen as a war crime). While dogs have been used militarily since antiquity, attack dogs have disappeared from normal warfare because they are vulnerable to modern firearms. But another reason is no doubt their autonomous behavior: erratic behavior made the Soviet anti-tank dog program largely a failure. The complex behavior of an animal has desirable components (e.g. skilled locomotion and perception) that cannot be separated from undesirable components (e.g. getting afraid, attacking one’s own side). Animal training serves to reduce undesirable behaviors but also to make the desirable behaviors more predictable – or, to return to Heyn and Asaro, make the decision-making less arbitrary. We hold dog owners liable for attacks their dogs do, typically by invoking negligence: they should know the likely behavior of the dog and likely risks of having them in a certain environment. Presumably a deterministic dog would have a more narrow field of liability, while an unknown and unpredictable dog would be regarded as something that should a priori be treated as dangerous.

Autonomous machines are similar in this respect to dogs. The more we can understand and “empathize” with their behavior, the better we can describe in what domains they are discriminating and non-arbitrary agents. We cannot be certain of their behavior (and it might be excessive to demand too much certainty), but if we have reasons to expect the probability of certain actions – even when stupid by human standards – the more clear the moral responsibilities of their handlers or suppliers become.

The problem is not that the machine lacks morality. The problem is that it is an imperfect proxy for somebody else’s choices, introducing enough noise and uncertainty in the chain of events that responsibility and rational response become disrupted. This implies two important limitations to the use of autonomous machines in war: the first is that their “psychology” can be inspected, tested and judged in a proper manner so that legal rules can be applied (and moral responsibility assigned in some agreed way). The second is that most participants in a theater of war – including civilians – must have enough knowledge about it that they can act accordingly. If the machine cannot distinguish a white flag from a target, or that certain patterns of behavior will trigger attacks, then that is information they should morally and legally have. If they cannot avoid acting in ways that trigger dangerous consequences the machine is an indiscriminate weapon.

Autonomy is not an end in itself, at least not for this kind of machine. It can reduce their reliability and utility. As Shakespeare wrote in Coriolanus: ”Do not cry havoc, where you should but hunt with modest warrant.”