Espionage, Ethics, and Law: From Philosophy to Practice

Abstract

In this paper, I respond to Lars Christie, David Omand and Stephen Ratner for their thoughtful comments on my book Spying through a Glass Darkly. In that book, I provide a philosophical defence of espionage and counter-intelligence activities. I have little to say about how best to implement the moral norms I defend so that they can help guide intelligence officers’ actions, in the world as we know it here and now. Relatedly, I have little if anything to say about whether domestic and international law should reflect and entrench those norms. These are the gaps which David Omand’s and Stephen Ratner’s contributions seek to fill. First, though, I consider Lars Christie’s probing objections to my views on the ethics of deception.

1 Introduction

I am very grateful to Lars Christie, David Omand, and Stephen Ratner for their thoughtful comments on my book Spying through a Glass Darkly.¹ While I stand by the book’s main normative conclusions, their responses have led me to revise and qualify some of my views.

Here is a brief summary of the book. In the context of foreign policy in general and war in particular, I define espionage as seeking to obtain information about third parties—typically, foreign actors—which the latter are trying to keep secret. Counter-intelligence consists in protecting one’s own secrets from those parties. Those activities raise the questions of what we—governments, political leaders, intelligence agencies, and citizens on whose behalf those people act—are morally allowed or obliged to do in order to obtain and protect secret information. I argue that espionage and counter-intelligence are morally justified, indeed sometimes morally mandatory, but only as a means to thwart violations of fundamental moral rights (construed as rights to the freedoms and resources necessary to lead a flourishing life.²) In the course of defending that thesis, I claim that deception, treason, manipulation, and computer hacking are morally justified means of procuring and protecting secrets, subject to the requirements of necessity, proportionality, and effectiveness.³ On my account, thus, some intelligence activities as carried out by political actors and their agencies are morally justified; many more are not, either because they serve unjust ends such as an unwarranted act of military aggression or the naked pursuit of economic interests, or because they inflict unwarranted, ineffective, or disproportionate harm on innocent parties.⁴

I speak of moral justifications: this is a philosophical book and the actual empirical cases which I use throughout have a largely illustrative role. I have little to say about how best to implement the moral norms I defend so that they can help guide intelligence officers’ actions, in the world as we know it here and now. Relatedly, I have little if anything to say about whether domestic and international law should reflect and entrench those norms. These are the gaps which David Omand’s and Stephen Ratner’s contributions seek to fill. First, though, I consider Lars Christie’s probing objections to my views on the ethics of deception.⁵

2 Espionage and Deception

The Oxford English Dictionary’s definition of the verb “to spy”, as “to watch (a person, etc.) in a secret or stealthy manner”, implies that espionage by definition involves deception: to do something in secret or by stealth involves hiding what one does from others, thereby deceiving them as to one’s intentions or aims. This is reflected in the law: the 1907 Hague Regulations, which outline the laws of war on land, define a spy as someone who “clandestinely or under false pretenses… obtains or endeavors to obtain information in the zone of operation of a belligerent”.⁶

Although my own definition of espionage and its counterpart, counter-intelligence, is neutral in that respect, the fact remains that much intelligence activity is deceitful: indeed, some of the best known and, truth be told, most admired intelligence operations in history, such as the Allies’ Double Cross system during WWII, have involved large scale deception of the enemy.

Therein lies the rub. To deceive another person is presumptively wrong. That is partly why Immanuel Kant, one of a handful of philosophers in the Western canon to have written about espionage, describes and condemns it as an “infernal art”.⁷ I do not agree with Kant. In Chapter 5 of the book, I attempt to justify deception in the service of intelligence activities. Here is a case where this seems straightforward:

Infiltration Green is locked in a conflict with Blue, a quasi-state organisation intent on conquering swathes of territory via a mixture of conventional and terroristic means. One of Green’s agents, Asset, strikes up relationships with Blue’s guerrilla fighters, concealing his identity and true allegiances. He collects information about a possible attack on Green, which he passes on to Green’s intelligence agencies. Green aims to eliminate those fighters before they can strike.

I assume that Green’s authorities are morally justified in thwarting the threats to fundamental rights posed by Blue’s fighters: the latter—I argue—have forfeited their right not to be deceived. As I also note, however, not all acts of deception are morally permissible, and some are harder to justify than others. Constraints on deception are grounded in the doctrine of double effect and the doctrine of acts and omissions. Christie rejects my account at both junctures: neither doctrine, he argues, supports my conclusions. I believe that he is partly correct.

2.1 Deception and the doctrine of double-effect

According to the doctrine of double effect, it is permissible to harm an innocent person in the course of pursuing morally justified ends so long as the harm thus inflicted is merely a foreseen, and not an intentional, effect of one’s acts—subject to the requirements of necessity, proportionality, and effectiveness (which I take as read throughout this section and the next). Contrast Infiltration with.

Infiltration* Green’s services have identified a high-level Blue commander who, they have good reasons to believe, is preparing an attack on Green’s soil. They encourage Asset to insert himself in the commander’s live-in entourage, including his innocent family members, with a view to spying on him, thereby also gathering information about his family.⁸

In the book, I argue that Asset is morally permitted to deceive the innocent relatives in Infiltration* so long as he does so merely as a side effect of deceiving Commander. Christie objects that contrary to what I claim, Commander’s relatives are not innocent bystanders: they are latent threats who, if they found out that Asset is not who he claims to be, would likely endanger him and pose a serious risk to his operation. Asset, thus, is morally permitted to deceive them intentionally.

As Christie notes, this presupposes that one may harm someone at time t1 in order to forestall the threat he might otherwise pose at time t2, even if one cannot be certain that he would pose such a threat. Not everyone accepts this point: on some restrictive views of permissible defensive harms, one must wait until the threat is imminent or has already materialised. As it happens, I agree that under certain conditions pre-emptive harming is justified, and I also agree with Christie’s verdict in this case: he has exposed a weakness in my argument and shed light into a range of complexities which I had overlooked.

Before I examine some of those complexities, it is worth noting that, contrary to what Christie seems to suggest, the doctrine of double effect can give us the right verdict in this case. One some accounts of deception, the latter is presumptively wrong in so far as it involves treating the deceived party as a means only to one’s ends. If the deceived party poses a threat which deception can parry, he is not wronged. Thus, Asset may intentionally deceive the relatives, thereby using them as a means to protect himself and his mission. We can also imagine cases in which the deceived party is not a latent threat, and in which the deception could thus be characterised as a merely foreseen and unintentional harm.

That aside, Christie advances two different justifications for Asset’s intentional deception of Commander’s relatives in Infiltration*. First, straightforwardly, Asset may justifiably eliminate the threat they might pose at t2 by deceiving them at t1. Eliminative harming is not the same as opportunistic harming yet is morally justified in this case.⁹

Second, and more interestingly, Asset has a paternalistic justification for deceiving them at t1. Here is why. If Asset does not deceive them (as well as Commander) at t1, there is a serious risk that they will pose a threat to him at t2 by disclosing his identity to Commander. At just the point at which they would be about to speak up, he would be justified in inflicting much more serious harm onto them—perhaps even killing them. Note that this is so (Christie argues) even if they are unaware of Commander’s role in the terroristic movement, but form the belief that he is up to no good without quite working out why. Asset, thus, is morally justified in intentionally deceiving them at t1 in order to protect them from becoming a threat and being subject to greater harm at t2. He acts paternalistically in so far as his justification is grounded in their interests in not being put in a situation where they would incur serious defensive harm and in not being led to commit the wrongdoing of threatening Asset’s morally justified mission.

I am not persuaded that Asset’s justification is paternalistic. On the most plausible account of paternalism, someone X acts paternalistically towards another party Y when φ-ing if and only if (a) X acts out of concern for Y’s interests and (b) Y explicitly withholds consent to X serving his interests by φ-ing, or X can reasonably be expected to presume that Y would withhold consent if he were in a position validly to do so. If Y justifiably withholds, or would justifiably withhold, consent to their interests being served by φ-ing, and if X nevertheless proceeds out of concern for him, his act of paternalism wrongs Y.

On this account, thus, to say as Christie does that Asset is morally permitted intentionally to deceive Commander’s relatives out of paternalistic concerns for their welfare is to imply, first, that the relatives do not or would not consent to being deceived (for his deception could not then be described, in morally neutral terms, as paternalistic),¹⁰ and, second, that their withholding consent to their interests being served by X φ-ing would not be justified (failing which his deception could be described as paternalistic but would be morally wrong.)

Let us take those points in reverse order. Ex hypothesi, Asset is morally justified in carrying out his mission, which if successful will save dozens of lives; he is also morally justified in protecting his own life from the relatives’ latent threat of disclosure at t2. Suppose that the relatives do or would withhold consent to being deceived as a means to protect them from a more serious threat at Asset’s hands at t2. It seems to me that they may justifiably do so. Of course, so long as they do pose such a threat and as Asset cannot be sure that he would be able to eliminate that threat then, they may not justifiably object to his deceiving them now. But were Asset to deceive them anyway on the grounds that their wellbeing is at stake, he would act paternalistically and unjustifiably so (even though, to repeat, he would have other justifications, grounded in the importance of his morally justified mission, for doing so.)

In any case, by Christie’s own lights, it seems that Asset is not acting paternalistically anyway (whether he does so justifiably or not.) For as Christie says, in so far as Asset’s mission is ex hypothesi justified, the relatives if apprised of the deception have “reason to welcome having been intentionally deceived to minimise the risk of becoming a wrongful threat to Asset”. In other words, they have reasons to consent; indeed Asset can reasonably presume as much.

None of this is to deny that Asset may deliberately deceive Commander’s relatives out of concern for their interests, and having reasonably presumed that they would consent. I thus agree with Christie’s verdict (albeit not his route to it), to whose account I would add the following three points. First, at t2, Commander’s relatives may be at risk not just at Asset’s hands, but at Commander’s. Suppose that Commander finds out that his relatives have had suspicions about Asset’s identity. It is not unreasonable to imagine that he might accuse them of betrayal, that he might not believe in their protestations of sincerity and loyalty, and that he would punish them accordingly. Asset is morally justified in intentionally deceiving them to protect them from him.

Second, and relatedly, we can imagine cases in which the relatives would face an extraordinarily difficult moral dilemma: expose Asset as soon as they discover who he really is to protect themselves and other possibly more vulnerable relatives such as their children, in the hope that Commander will take them at their word but at the risk of not being believed, with attendant horrific consequences; or keep Asset’s identity secret in the hope that his mission will succeed and that Commander will be killed, but at the even greater risk to their safety the longer their own act of deception goes on. I agree with Christie (and with Kimberley Brownlee, whose work he cites in support of his view on this point¹¹) that it is an important component of moral agents’ wellbeing that they should not be embroiled in, and a fortiori commit, wrongdoings. When we face a moral dilemma, we unavoidably commit a wrong: no matter what we do, there is a moral remainder for which we are accountable. If Asset is morally permitted intentionally to deceive innocent relatives out of concern for their material wellbeing, then he is morally permitted to do so out of concern for their moral wellbeing. In fact, in the light of my defence of mandatory deception in that chapter of the book, I would go as far as to say that he is under a moral duty to do so.

To recapitulate, I agree, then, that my initial account of Infiltration* was in part mistaken. But so long as there are cases in which some people are deceived who are not and will not become a threat to Asset, we may be able to justify his deception as a foreseen albeit unintentional side effect of his deceiving Commander; moreover, cases in which intelligence operatives are deceiving others intentionally are harder (though not impossible) to justify than cases in which they do so unintentionally albeit foreseeably.

2.2 Deception and the doctrine of acts and omissions

The doctrine of double effect, thus, remains of service. What about the doctrine of acts and omissions? Christie thinks not, at least in the way that I apply it.

Suppose that I pursue unjust ends, for the sake of which I can either kill some innocent person Y, or let someone else, Z, die, at the hands of another or due to some natural cause. Since my ends are unjust, I should do neither. According to the doctrine, other things equal, my killing Y is morally worse than my letting Z die. Suppose next that my ends are just. In this case, I may well be morally permitted to harm Y; but if I can let Z die instead, I ought to do so. Put differently, my killing Y is harder to justify than my letting Z die.

I appeal to the doctrine to support the conclusion that some deception operations are harder to justify than others. My examples are drawn from two well-known Allies operations during WWII. On the one hand, in order to hide from the Germans the fact that they had cracked their Enigma code, the British decided to let the German Navy sink a number of Allied military and merchant ships in the Atlantic, at the cost of thousands of lives. On the other hand, in Operation Crossbow, they misled the German high commands into thinking that their long-range rockets were hitting Central London, by supplying false information via various spy networks; as a result, the Germans shortened their range, sparing lives in Central London but at the expense of lives in South London. In both cases, the Allies engaged in deception. But in the Enigma case, the Allies allowed thousands of crew members to die, while in the second case, they contributed to the killing of civilian lives in South London. I say, in the book, that “Operation Crossbow stood in greater need of justification than the measures taken to protect the Enigma secret”.¹²

Christie rejects my account: “Enigma”, he writes, does not “stand in lesser need of justification than Crossbow merely in virtue of being accused of allowing rather than doing harm”. In fact, he thinks that the deception involved in Enigma is harder to justify than that involved in Operation Crossbow. This is because in Enigma, the sailors who die as a result of the Allies’ decision not to rescue them are made to play a part in the Allies’ plan, whereas the civilians of South London are not pressed into such service: they just happen to be there. Christie thinks that Enigma is relevantly similar to the following scenario, which he borrows with some modification from Warren Quinn:

Guinea Pig A doctor withholds life-saving medication from Jim in order to observe the fatal progression of his deadly but painless disease. In this way, she will gain the knowledge necessary to save five other patients with the same disease.¹³

Here too, some innocent agent, Jim, is made to play a part in another’s ends, here the doctor’s end of saving five lives. Many would regard the doctor’s omission as monstruous. The case has the merit of drawing our attention to the fact that the doctrine of double effect and the doctrine of acts and omissions cut across each other. If Christie is right that it is similar to Enigma, and if we also think that the Allies were morally permitted to carry out deception in that case albeit at the costs of thousands of lives lost at sea, it also invites us to ponder whether we might need to revise our judgement in Guinea Pig. The question, then, is not merely whether Operation Crossbow is harder to justify than Enigma: it is also whether the latter can be justified at all.

On the first count, I disagree with Christie’s point that the Allies involved sailors in their plans in Enigma (as the doctor does in Guinea Pig) but not so in Operation Crossbow. Rather, the sailors’ death is a foreseen side effect of the Allies’ decision to allow some of their ships to sink—as are civilian deaths in South London. Christie objects that this move occludes the fact that the British need the Germans to believe that they have not been discovered. I agree. But it is not true that they need the Germans to kill some of their sailors. Suppose a terrible lethal illness had wiped all sailors shortly before the point of impact with German U-boat rockets. The Germans would not have been the wiser as to the cause of deaths of Allied sailors, yet their mistaken belief that they were able to advance undiscovered would have been maintained. By contrast, in Guinea Pig, the doctor does need Jim’s illness to progress to its lethal conclusion if she is to develop the cure she needs to save five patients.

Where, then, does this leave us? It leaves us with the thought that in both Enigma and Operation Crossbow, the deaths of sailors and civilians is an unintentional albeit foreseen side effect of the Allies’ deception: at the bar of the doctrine of double effect, they are relevantly similar. But as per my initial conclusion, they differ at the bar of the doctrine of acts and omission, such that the latter is harder to justify than the former.

Suppose that I am wrong—that the failure to protect sailors in Enigma does involve pressing them into the service of the Allies’ ends. If one thinks that the doctor acts wrongly in Guinea Pig, is one committed to the view that the Allies acted impermissibly in Enigma?

There is a sense in which the analogy between Guinea Pig and Enigma is misleading, to do with the sheer number of lives sacrificed but also spared over the course of the war thanks to such operations. So here is a better analogy. Suppose that one of Green’s platoons has cracked Blue’s encrypted messages, and that they now know that Blue’s guerilla network is planning to attack a village where (Green also knows) there are five civilians left. The platoon commander must choose between two options: she can let the enemy kill an innocent civilian now, thereby protecting Green’s decryption secret, as a result of which her platoon will be able to protect the five civilians the next day; or she can order her platoon to rescue that one civilian, as a result of which the enemy will know that their communications are not secure, will change plans, and will attack another village, with five people left in it, but too soon for Green’s platoon to get there on time. I am tempted to think that she may justifiably do the former—similar to what the Allies did with Enigma. If Christie is correct that she is thereby enrolling Blue’s innocent victim into her plan of saving the five the next day, then it seems that the doctor is morally allowed to allow Jim to die in Guinea Pig.

Intuitively, this is very hard to accept. Christie rehearses other putative differences (than the one rooted in merely foreseen consequences) between Guinea Pig and Enigma which he does not in the end find convincing. I think he is somewhat too quick. I agree with him that both Jim in peacetime and innocent civilians, as well as just combatants, in wartime have a right of rescue—the former against his doctor, the latter against other just combatants and the military leadership tasked with protecting them. However, the civilians and combatants know that their country is at war and must realistically anticipate that, in this context, the military leadership will not always be able to protect them and may have to sacrifice some lives for the sake of saving others. Put differently, it is part of a military leader’s role not merely to protect lives and limbs, but also to make difficult decisions as to whose lives and limbs to protect. Contrastingly, when Jim consents to being placed in the care of this doctor, he does so in anticipation that, by the morally justified conventions of her profession, she may not sacrifice his life for the sake of furthering medical research. Unless he consents to sacrificing his life to save another five, her doing so is a grievously wrong abuse of trust.

Admittedly, if the delivery of health care in Jim’s community is organised in such a way that everyone knows that, when they suffer from a terminal illness, they are at risk of medics deciding to use them for research purposes, Jim will not be able to complain that his trust has been abused, since he would have no reason to trust in the first instance. In the absence of compelling reasons so to organise the delivery of health care, indeed in the presence of compelling reasons to enforce the convention that a doctor must not allow a patient to die for the sake of furthering medical research, Jim has a grievance which the innocent victim of the platoon commander’s decision not to intervene (as well as the innocent victims of the Allies’ decision to protect Enigma) lacks. This does not, note, amount to justifying that decision, but it does suggest that contrary to what Christie suggests, Guinea Pig is different, morally speaking, from the deception cases I examine in the book.

3 Espionage Ethics and the Limits of Philosophy

In the Acknowledgements section of the book, I mention the fact that some of the people who have provided extremely helpful comments on the manuscript used to work in intelligence; I also express the hope that the world I describe is not too far removed from their former professional home. So it is immensely reassuring as well as gratifying when someone of David Omand’s stature in that somewhat opaque world finds much on which to agree with me here. It is also unsurprising that he should find some of my claims problematic, though as I hope to show in this section, there is perhaps less that separates us than unites us.

Let me first outline areas of strong agreement. I am well aware that, to him and to many of his colleagues, what I am about to say will seem as if I am pushing through an open door. It is nonetheless worth stressing, for the sake of readers who, firmly ensconced as they are in academia, might have formed the belief that intelligence agents are not as concerned with the ethics of their role—or, at any rate, are perhaps all too ready to allow the raison d’état to override ethics.

We agree, then, that espionage and counter-intelligence activities are morally justified only if they are carried out for the sake of morally justified ends, if they are a necessary and proportionate means to achieve those ends, if they stand a reasonable chance of success (or are reasonably likely to be effective), and if they discriminate between those who are legitimate targets and those who are not. We also agree, crucially, that under those conditions, espionage and counter-intelligence are not merely morally justified: they are morally mandatory. On our respective accounts, the duty to spy is an instantiation of a general duty of care. There is perhaps a difference of emphasis between his account and mine: while I claim that Green’s duty is owed as much to Blue’s innocent civilians as it is owed to Green’s citizens and state officials, Omand focuses on the latter.¹⁴ Moreover, as I make it clear in the book, Green’s duty of care to its officials (not least its soldiers) is not merely a duty to protect them from material and physical harm resulting from faulty or inexistent intelligence: it is also a duty to protect them from the moral harm of taking decisions (where to deploy troops in the face of the enemy, on whom to impose economic sanctions, when to start peace negotiations, etc.) which might be objectively wrongful in the light of all relevant facts. Consider drone operators in Green who know and believe that they are morally as well as legally forbidden from targeting innocent civilians in Blue’s territory, thousands of miles away. Green’s leadership decide not to conduct any kind of intelligence gathering in this area of Blue’s territory, in the knowledge that, as a result, its drone operators will kill scores of innocent civilians. Pending good reasons for not spying on Blue, Green’s leaders do not merely wrong those Blue civilians: they also wrong their operators. I do not think that Omand would disagree, on the contrary. The point is worth emphasising, though, if only to bring attention to a facet of the duty of care, as owed in respect of our moral agency and wellbeing, which has been relatively neglected in the philosophical literature.

Finally, Omand draws a distinction between creating the capacity to gather intelligence, and using that capacity. He rightly notes that I do not take up this point in the book, so let me do so now. The claim that there is a duty to spy under certain conditions implies that there is a duty to set up the requisite infrastructure so that, if and when those conditions obtain, the relevant agents may fulfil that duty. In other words, I am no more an intelligence-agency “abolitionist” than I am a “military abolitionist”.¹⁵

At first sight, the aforementioned conditions (morally justified ends, necessity, proportionality, likely effectiveness, and discrimination) on justified intelligence activities seem to be borrowed, without modification, from the ethics of war. Indeed, Omand’s own works are well-regarded for applying the just war theory framework to intelligence.¹⁶ As he notes in his reply, I find that framework less useful than he does, as it risks occluding important differences between espionage and war—not least the fact that the former occurs both within war (which complicates its assessment, for each operation must be judged by appeal either to the specific military operation which it aims to assist, or by the ends of war in toto, or both), and in peacetime. I prefer to construe those principles, then, as norms which govern the imposition of harm in general.

With that relatively minor quibble out of the way, in the light of his exposition and defence of those norms, I should like to clarify an important point, correct a mistake (of mine), expand on an under-developed claim, and note an area of disagreement. I shall then tackle his most important and probing objection, namely that my account is not sufficiently action-guiding and thus exposes some important limits of philosophy.

The point of clarification is this. On my account of the conditions under which one may permissibly harm another person, some agent, G, is morally justified in harming another agent Blue if and only if, subject to Green’s course of action being necessary, effective, and proportionate, Blue either unjustifiably contributes to the violation of some agent’s fundamental rights, or Blue unjustifiably fails to protect some agent from such rights violations, or harming Blue brings about, or is a collateral cost of justifiably pursuing, morally weighty ends.¹⁷ Omand queries my claim that those conditions are singly necessary and jointly sufficient, on the grounds that Green may have evidence that Blue is preparing such a violation and need not wait until after he has committed it in order to harm him (in the context of this book, thus, to harm him by spying on him.) Now, I agree with him that Green may so act under those circumstances. However, this does not impugn my overall framework: of key importance here is a distinction between fact-based and evidence-based justification. Thus, G is evidence-justified in harming Blue if and only if he has evidence that Blue unjustifiably contributes to the violation of some agent’s fundamental rights, or unjustifiably fails to protect a third party from such violation, and so on, and if he has evidence that its response would be necessary, likely effective, and proportionate. I stress the point here for two reasons. First, the task of justifying espionage and counter-intelligence is, in large part, the task of ascertaining what intelligence agencies and their individual agents are morally permitted (and obliged) to do given that they lack the evidence they need. Second, the question naturally arises as to what evidence must G have, of Blue’s putative wrongdoings, in order justifiably to harm Blue.

In the book, I say in the context of discussing the collection and use of bulk data that the probable cause for Blue’s wrongdoings must lie in open sources and their proper analysis by Green.¹⁸ Only then (I go on to say), may Green use as yet non-open information (collected in bulk) about Blue. I now think, thanks to Omand, that this was a mistake. As he correctly notes, Green may have been passed on secret intelligence about Blue by an ally or, indeed, by an informant within Blue, which would constitute evidence of probable cause. That said, as I recurrently make clear in the book, not all intelligence is obtained by rightful means. It is an interesting and difficult question, and one which I do not address, whether Green may use morally tainted information passed on to its officers by an ally, or a source within Blue (most obviously, information obtained by torture) as evidence of probable cause for carrying out further espionage activities in Blue. I am tentatively inclined not to rule it out at the outset. Suppose that Green’s informant in one of Green’s supposed allies, Amber, supplies Green with information about Blue’s incoming unjust invasion of a neighbouring country. Suppose that Amber’s services obtained the information by sexually blackmailing a top military official in Blue, and that Green can corroborate that information by other sources which in themselves would not be probative. I am inclined to think that Green is morally permitted to use the information, tainted though it is, as evidence of Blue’s plans and as a reason to mount an invasive surveillance operation on that official—even if its own services adamantly refuse on moral grounds to engage in sexual blackmail.

Attentive readers will have noticed that I have not mentioned a familiar, sixth criterion for a just war and, for Omand, for justified intelligence activities, to wit, the criterion of right authority. It is key to Omand’s own account, though not to mine—at least, as my account is set out in the book. In the law and ethics of war, the criterion stipulates that a decision to go to war is morally justified only if it is taken by a political actor which has the legitimate authority to do so. On orthodox accounts of the ethics of war, only the constitutionally appropriate representatives of states are morally authorised to do so, on the grounds that only sovereign states are morally entitled to wage war. On less orthodox accounts, non-state actors are morally entitled to wage war; however, they may do so only if there is a plausible sense in which they have been authorised to do so by those on whose behalf they fight.¹⁹ As applied to intelligence activities, then, the criterion stipulates that they are justified only if they are carried out by actors who have been authorised, or could claim to be authorised, to do so.

The word “justified” as I use it here is ambiguous. For crucially, neither I nor Omand believes that right authority is a necessary condition for such activities to be fundamentally justified. Here is an example. In November 1939, a German scientist called Hans Ferdinand Mayer approached the British Embassy in Oslo, entirely of his own initiative, and gave them detailed information which he had obtained through his work for Siemens about Germany’s military use of electronics. He certainly had not been authorised by anyone so to act—yet it seems intuitively plausible that he was justified in so doing.²⁰ Here is another example. Suppose that rebel groups are fighting a totalitarian regime within their own borders, but that they have not yet formed political structures through which they could plausibly claim to be authorised to do so by the regime’s victims. They are no less clearly morally justified in spying on state officials, with a view to furthering their just ends of overthrowing the regime. As I argue in earlier work in respect of the resort to war, when victims of rights violations are not in a position to consent, directly or indirectly, “to the war itself or to institutions mandated with fighting the war, insurgents may take matters into their own hands, [so long as] they have good reasons to believe that their fellow community members would consent if they could, and they put in place as usual mechanisms whereby those for whose sake they fight can hold them to account once the war is over”.²¹

I think that Omand would agree with me. However, as a “practical institutional” matter (in his words), he thinks that right authority is crucial: “having right authority properly recorded allows for legal and Parliamentary accountability for decisions and oversight of activities where there may be legitimate grounds for questioning whether the actions should have been ethically permitted”.

In this particular respect, his paper explores an aspect of espionage ethics which I deliberately and self-avowedly do not examine, to wit, the degree to which the norms I defend can be operationalised both by legislators and by intelligence agencies in such a way as to guide their officers’ actions, and as to guard against abuses. I take Omand to raise four doubts about my account. First, it is not clear whether those norms are objectively true. Second, even if they are objectively true, it is not clear how we can reliably know that they are true in the face of persistent disagreement. Third, it would be very difficult for any intelligence officer operating in the field to know what to do, armed with my book, in the fog of the intelligence war. Fourth, those norms cannot be enforced outside the domestic context in which they apply. Under those conditions, then, it stands to reason that intelligence agencies should not be allowed to operate other than under the tutelage of a democratically accountable authority and on the basis of lawful and codified professional guidelines for their individual agents. Imperfect though those institutional mechanisms are, they are the best that we can hope to get.

I agree that, when and if we are in a position to regulate and codify intelligence activities along those lines, we ought to do so: the “ought”, here, is not merely prudential but rather is a moral duty. That said, while I am aware as a philosopher of the limits of my own discipline, it seems to me possible to resist some of the pressure which Omand applies on those limits—or perhaps, better put, to meet him halfway. Of particular concern to him, and for understandable reasons, is the plight of the intelligence officer who needs to decide, here and now, whether to mount an operation against some foreign nationals but who does not know whether the information she procures will be put to just ends. What can she learn, if anything, from the book? That she is morally justified in proceeding if, to the very best of her professional judgement, the evidence at her disposal suggests that the risk she would incur by not proceeding, in the form of failing to save lives and limbs, is greater than the risk she would incur by going ahead (in the form of furthering wrongful ends). Objectively speaking, she may get it wrong of course. What the book does not do, and should have done, is develop an account of what she is owed, by her superiors, her government, and us too on whose behalf she acts, namely the tools to develop her professional judgement and the concomitant epistemic and moral virtues of thoughtfulness, discernment, and strength of character. On this, Omand is absolutely right, but his is a lesson which my own discipline need and should not resist—on the contrary.

I am more sceptical about his other concerns. At one point in his article he describes the conditions for justified intelligence activities as “artificial constructs”. I take it that he means that they exist merely by dint of the fact that some people have articulated and defended them. This would imply that they have no grounding in anything other than people’s minds, and that the fact that they vary across time and space is something that we should simply accept, without judgement. But I am not sure he himself is wedded to that view. In one of the most interesting passages of the paper, he writes that the code of practice which, he claims, ought to regulate those activities should be “generally accepted”, by which he means that it is enforceable and enables Parliamentary oversight. Crucially, however, he is at pains to stress that “that is not to say that the code is justified because democratically accepted”. (His emphasis.) Why not, though? If those norms are but an artificial construct in general, and of our time and place in particular, it seems that the fact that they are democratically accepted justifies them.

We ought to distinguish the claim that we are morally justified, or under a duty, to do x, from the claim that, here and now, in the world as we know it, professional codes of conduct and their underpinning legal apparatus ought to permit or compel us to do x. The first claim is compatible with the view that the relevant moral norms are not artificial constructs; the second claim accepts that there may well be good reasons for professional codes of conduct and the law not to reflect the morally best norms, and that, in a different world, these might have to change. Philosophers such as me and, in fact, many of my colleagues working in the just war tradition can readily accept that the tasks of shaping soldiers’ and intelligence agents’ professional codes and of determining what the relevant law ought to be, here and now, is best left to professionals and legislators acting together. But this need not lead us to eschew philosophical reasoning altogether: for in just the same way as it behoves philosophers to acknowledge the limits of their inquiry, it behoves professionals and legislators to acknowledge the limits of their practical enterprise.

I hope that Omand would agree with me: there is less blue water between his position and mine than we might think. But there is one outstanding difficulty which I have not addressed, and to which he alludes, namely that even if we, citizens of a relatively stable liberal democracy, can agree on the principles which should guide our intelligence agencies’ activities in the world as we know it, we and our leaders are faced with other states which are likely to disagree with us. In the absence of a fully developed international legal framework for regulating espionage, what, then, are we to do? To make progress on this difficult question, we must turn to Steven Ratner’s piece.

4 From Ethics to International Law

Towards the end of his paper, Ratner describes the book as a work of “ideal theory, and deep structure… Not oriented toward improving the status quo”. He further writes: “Such theory’s value lies in forcing audiences attached one way or another to existing practices (even those ratified by law) to question their assumptions and grapple with what spying ought to be allowed in an ideal world (or at least an ideal world of states).” His descriptive diagnosis of the book is entirely apt; his evaluative comment is pleasing, coming as it does from a scholar whose work in international public law I have long admired. Nonetheless, he probes in various ways the degree to which my normative account of intelligence activities adequately captures and makes moral sense of the ways in which states carry out those activities, and can serve as a bedrock upon which to build an international legal framework. It is fair to say that he is sceptical. Yet, as I hope to show, we disagree less than might appear at first sight as to where the limits of philosophy lie and where the task of morally directed institutional design must begin. The key point of disagreement pertains to the overall normative framework for such design.

Let me start with his account of the status of espionage in international law. States are authorised to procure information which other states would rather keep secret, to help ensure compliance with international law in general and international treaties in particular; indeed, they are encouraged to do so, particularly by international organisations such as, no less, the United Nations; indeed, it is assumed that they will do so. This is just as well for, as a matter of fact, they routinely spy on one another, for three reasons: as already noted, to uncover evidence of lawful action; to uncover evidence that the target of their intelligence activities engages in dangerous, albeit lawful, conduct such as highly polluting emissions or weapons stockpiling; and out of naked self-interest. In the first two cases, Ratner argues, they “[advance] key global—even cosmopolitan—values”. Following Asaf Lubin’s important work on espionage in international law, we see that, in law, states have agreed more or less implicitly that they should all have the Hohfeldian liberty to spy on each other, with no expectation therefore that any of them should be under a legal duty not to do so.²²

I have no difficulty accepting that states are morally justified in gathering intelligence for the purposes of monitoring and enforcing compliance with international law, and of thwarting dangerous threats, including outside of war. Indeed, I say as much in the book.²³ I also accept that international organisations as well as NGOs (which Ratner also mentions) are morally justified in engaging in or assisting intelligence gathering: in fact, Ratner’s state-centric approach risks leaving a good deal of espionage practices out of the picture.

Nevertheless, according to Ratner, I permit far less intelligence gathering than occurs in practice, for two related reasons. First, whereas I ground ethical spying in the defence and protection of fundamental moral rights, the ends which states seek to further by spying on one another often (more often than not?) have nothing to do with those rights. Second, whereas my moral rights-based approach implies that states’ right to spy depends on the moral status of the ends they pursue, in practice and in law states enjoy equal sovereignty rights and liberties in general, and equal liberties to spy in particular.

The key issue here is how political and moral philosophers of my ilk and international public lawyers such as Ratner conceive of the relationship between moral norms—what Ratner elsewhere calls the “moral fiber of the world order”—and the law.²⁴ He outlines three possible conceptions of that relationship: recasting international law and practices in terms of fundamental moral norms; evaluating international law and practices at the bar of those norms and constructing additional second-best principles for institutional and legal design; foregrounding international law and practices so as to develop moral norms which are fit for purpose in the world as we know it. (Those labels are mine.)

Ratner reads my book as endorsing the first and second approaches, while he favours the third. I resist both his objections to the former approaches and his defence of the latter.

Consider recasting. Here is an example drawn from the book which Ratner regards as unpersuasive. On my account, spying on North Korea to ensure that it complies with international regimes in respect of nuclear non-proliferation is justified on the grounds that such regimes protect fundamental moral rights. On Ratner’s view, it is not clear that any individual in the world has a fundamental moral right that North Korea should not possess nuclear weapons: rather, the claim that it should not do so reflects a commitment to a “public order value”, by which he means (I assume) that nuclear proliferation is a danger to the global order. Now, I do not dispute that claim. What I do dispute is the thought, implicit in his objection, that one cannot coherently defend the moral imperative of protecting the global order by appealing to individuals’ interests, wherever they are in the world, in being protected from the fall-out of a nuclear war, wherever the latter may occur.

Perhaps I am choosing too easy an example. But even the law of the sea or environmental law can be construed, and in some of their provisions justified, by appeal to those interests, even if in practice they are not understood as such. Consider the law of the sea. Admittedly, it is not framed through the lenses of fundamental rights. But there is a sense in which it can be recast in those terms as follows: it confers rights and liberties on states in respect of spying on one another in international and territorial waters; those rights in turn can be defended in the light of moral norms which are reflected in other international legal regimes such as the UN Charter, and which are justified independently of that fact. So, in that light, a state’s legal right to spy at sea can be recast, and defended, as supervening in part on its individual citizens’ jointly-held moral right to shape their collective future.

Ratner may perhaps agree with me that the approach I have just outlined works in some cases. However, he will likely counter that even if the protection of fundamental moral rights ground a state’s right to self-defence, it offers a much weaker grounding for its right to spy. I remain unconvinced. For a start, states need information about putative enemies in order to make evidentially sound judgements as to whether or not their citizens’ rights are under threat and, if so, how best to defend those rights. Moreover, states also have the legal right to provide military assistance to other states. Such a right can be plausibly framed as, and is best grounded in, a moral right to help those other states’ citizens whose fundamental rights are under threat. To exercise their right of other-defence, states do need adequate intelligence. In addition, the provision of military assistance often includes, and in some cases justifiably so, supplying vital information about the common enemy—as, in fact, the United States and their allies have done for Ukraine since (and before) the 2022 invasion.²⁵

I am more optimistic than Ratner, then, about the recasting approach. Of course, this is not to say that international law does reflect all of our fundamental rights. For example, on my account, individuals who pose an unwarranted serious threat of harm to another party are liable to be met with some degree of force, subject to the aforementioned requirements of proportionality, necessity, and effectiveness, whereas those who so counter their threat are not liable to counter-retaliatory force. This is a claim which the laws of armed conflicts do not always recognise.²⁶ Ratner also notes that whereas my approach implies that whether a particular individual may be justifiably spied upon partly depends on his or her contribution to violations of fundamental rights, in practice states permit themselves to spy on vast numbers of innocent people in ways which are not easily captured by the law. Deceptive practices are rife, and could not easily be turned into legal breaches.

Enter the evaluative approach. On my account, some of those deceptive practices, albeit permitted by the law, are morally wrong. It does not follow that there is nothing we can do, to reform both laws and institutional practices, to get us closer to the demands of morality. On the contrary, what we need is a set of second-best norms which are sensitive to facts about human nature, the problem of uncertainty in the face of many states’ reluctance to set up proper oversight mechanisms for their intelligence activities, the problem of disagreement between relevant parties as to what first-best norms require, and the difficulties inherent in ascertaining, above and beyond clear cases of injustice, what constitutes a breach. These obstacles to the evaluative approach are listed by Ratner himself. I agree that they are formidable: as I noted in my response to Omand, there is only so much of morality that can be implemented and enforced. This is not unique to intelligence activities of course; the same goes for the morality of war, indeed the morality of intergenerational transfers of resources, the morality of taxation, the morality of immigration regimes, and so on. The best that philosophers can do, qua philosophers, in addition to bringing into light the moral fibre of the word order, is to reflect on the degree to which various proposals for reforms satisfy first-best moral norms, why they fail to do so, and whether the gap between philosophy and practice is morally defensible.

My approach to bridging the gap—a combination of the recasting and evaluative approaches—differs from Ratner’s favoured strategy, which foregrounds state practice and extant law to yield moral principles. As he puts it, “institutional moral reasoning demands that theory consider institutions—which includes international law—as both sources of political morality as well as constraints on certain accounts”.²⁷ It holds that the norms applicable to the political realm must be arrived at by the consent of the relevant parties if they are to be authoritative, and that feasibility constraints are part of the normative fibre of the world order as it is, and not as we unrealistically want it to be. Thus, the fact that non-democratic states and democratic states are on a footing of equality in international law is not to be regarded as a morally regrettable fact which political theory must accommodate: rather, it is to be treated as an essential component of any normative, action-guiding principle for regulating international relations. On this view, then, states are first-best morally justified in spying on one another for reasons which have little to do with the protection of fundamental moral rights, so long as they do not violate normative constraints which are immanent in extant practices and in the law.

At the level of justification, Ratner’s account of the norms by which states must abide somewhat differs from mine. But regarding specific permissions and prohibitions relating to espionage, it is much closer than one may suppose. I have already suggested that I can endorse much of what he says in the light of the recasting and the evaluative approaches. Indeed, his useful four-fold distinction between the targets, modes, goals, and consequences of spying, and the conclusions he reaches in each case, are consonant with my arguments in the book. Here is one way to frame both the difference and the similarity between Ratner’s account and mine. The prohibition on military aggression is set out in the UN Charter. When Russia invaded Ukraine in February 2022, not only did it commit the legal wrong of aggression. It also committed a grievous moral wrong. On Ratner’s view, Russia’s moral wrongdoing consisted in violating one of the most important constitutive norms of our current world order qua such norm. On my view, its moral wrongdoing consisted first and foremost in violating Ukrainians’ jointly held moral right to sovereignty over this particular territory. On both our views, Russia is not morally permitted to spy on Ukraine as a means to achieve its ends.

If we agree on this and relevantly similar points, then, does it matter that we arrive at those by different justificatory routes? I believe so. On the one hand, the kind of philosophical inquiry which my book seeks to exemplify does not have internal resources to bridge the gap between philosophy and practice: it must draw on and work with other disciplines such as political science, and in particular its rich body of work on institutional design, international relations, and international law. On the other hand, however, hewing closely to immanent norms makes it harder to justify contesting extant practices and law. To illustrate with a somewhat “brutal” example, the first recorded case of a military commander charged with and convicted of war crimes in general, and the crime of rape by soldiers under his command in particular, is the trial of the mercenary Peter von Hagenbach in 1474. Yet—needless to say—rape continued to be, in fact still is, routine in war (and in peace.) It took over five centuries, until the late 1990s, for the resort to rape as an instrument of war to be explicitly characterised in international law as a war crime and, in some circumstances, a crime against humanity. Had such a call been made in, say, 1700, it is not clear how Ratner’s view could have accounted for it.²⁸ By that token, suppose that in a hundred years from now, the mass surveillance of civilian populations is not only widely practiced by states but in addition becomes permitted under international law, and thus an immanent permission thereof: it is not clear by what moral standard, in the light of Ratner’s account, our successors might condemn it as a gross violation of their rights and might seek to resist it.

Irrespective of which approach one takes to bridging the gap between philosophy and practice, we (scholars, policy-makers, diplomats, and government officials) need to confront a seemingly insuperable difficulty, namely that much of the business of states’ foreign policy is done in secret. This, in fact, is one of the reasons why my book has a lengthy discussion of secrecy—the role it plays, the risks it poses, the extent to which it is justified. Ratner’s paper ends with a thoughtful discussion of this issue, with which I agree. I am also inclined to think that, amongst other things, secrecy presents reform-minded political leaders with a bootstrapping problem: in order to know which steps can feasibly be taken to bridge the gap between the philosophy and practice of foreign policy, those leaders may need access to information which their counterparts are determined to keep secret; in order to access such information, they may have to instruct their officials to engage in practices which the deep morality of espionage might permit but which extant practices and their immanent norms might not and, for that very reason, they might have to shroud their activities in secrecy. I aim to show that under certain conditions, they would do the right thing. Ratner, and I believe Omand too for that matter, highlights some of the resulting costs. Notwithstanding the points on which we disagree, there is an important sense in which our inquiries are complementary.

5 Conclusion

It is a testimony to the quality of the objections one receives on one’s work that one wishes to have heard them before finishing the book. Had I had the benefit of reading Christie’s, Omand’s, and Ratner’s incisive papers, I would have refined my account of deception, and attempted to develop an account of the relationship and the gap between the philosophy of espionage, its extant practices, and the body of international law which, explicitly or implicitly, frames it. The book would have been the better for it. As it is, I am thankful to all three commentators for giving me the opportunity to map out some of the terrain that still needs covering.