Here is a puzzle: It is sometimes permissible for an agent to perform an action but impermissible for that agent to demand something in exchange for not performing it. How can it be? This notoriously perplexing question, often presented as “the paradox of blackmail,” has prompted numerous scholarly attempts to explain why a blackmailer’s threat to act permissibly is wrongful. A special subgroup of suggested solutions consists of revisionist, coercion-based theories, which advocate unraveling the puzzle by denying its very own presuppositions. They hold, in other words, that upon reflection all cases of wrongful blackmail actually involve a threat to act in an impermissible way. Thus, blackmail is not more puzzling than regular cases of extortion: it is, in fact, a threat to act impermissibly. What is puzzling about blackmail is only the fact that the threatened act seems to be permissible at first glance.
A revisionist approach thus eliminates the paradox, rather than resolving it. It aims at uncovering the feature that makes the threatened act seem permissible, though it is actually impermissible. One major strategy of these theories is to claim that the threatened act, while being legally permissible, is morally impermissible. Thus, for example, a case where A threatens to expose B’s sexual orientation unless he is paid hush money is wrongful, since exposing this secret is morally impermissible. A conditional threat to act impermissibly (morally speaking) is wrong for the same reasons that general cases of extortion are wrongful.Footnote 1
Yet there are cases in which this approach does not seem to fit. There are, in other words, cases of threats to do things which seem to be perfectly permissible, even from a moral point of view. For example, revealing one’s adulterous behavior to one’s spouse might be permissible. Why would a conditioned threat to do so be wrongful? This kind of case pulls in the direction of a special kind of coercion-based theories, which build on a subjective notion of right and wrong. They hold that while blackmail involves a threat to do something which is objectively permissible, some special character regarding the agent who performs the act (or threatens to do so) renders this act impermissible. Such theories, which constitute a subset of coercion-based theories, are “subjectivist theories of blackmail,” of which Mitchell Berman’s is the most prominent representative.Footnote 2
This paper starts by elucidating and defending an improved version of a subjectivist theory of blackmail. As a coercion-based theory, it holds that blackmail, like extortion, involves a threat to act in a (morally) impermissible way. The impermissibility of the threatened action is based on the fact that its harmful consequences are intended by the blackmailer. The paper demonstrates how this illicit intention can be inferred from the structure of the proposal. It therefore first offers a clearer background for the existing subjectivist theory, as well as an improvement of its evidentiary component.
However, the claim that intentions are relevant to permissibility has been substantively challenged in modern moral philosophy. In response, I show why even those who reject the general relevance-of-intentions thesis can still accept the subjectivist theory. While some cases are easily resolved by having them hinge on the predictive value of intentions, a full picture is revealed only when we appreciate the role of permissibility and responsibility within the scheme of coercion claims. I therefore conclude with an improved subjectivist theory of blackmail which endures even if intention is not genuinely relevant to permissibility.
1. “The Paradox of Blackmail” 101
Dozens of articles thus far have dealt with the puzzle of blackmail.Footnote 3 Most of the discussion tends to focus on the case of “informational blackmail”: A is threatening to reveal embarrassing information regarding B (paradigmatic examples include B’s sexual behavior), unless B meets A’s financial demand. The puzzle arises from the deeply intuitive sense that something is wrong with A’s act coupled with the difficulty of explaining why.Footnote 4 Moreover, as a matter of doctrine, blackmail is usually criminally prohibited.Footnote 5
Like other profound theoretical problems, a major part of the challenge is not only to solve the puzzle, but also to properly articulate and specify the challenge it poses.Footnote 6 At least four major different articulations of the puzzle can be listed:
1. The “Synergetic Miracle”: Since the threatened act is permissible (and hence, it would seem, nothing can be wrong with expressing the intention of doing it),Footnote 7 and since the demand to pay, in and of itself, is permissible and legitimate, it is unclear why the combination of these two permissible components should produce a forbidden act.Footnote 8 Of course, in many cases a wrong may emerge from a combination of permissible acts, with no mystery at all: driving is permissible, and drinking alcohol is permissible (at least for the sake of this argument), but the combination is wrong.Footnote 9 It seems better, therefore, to rephrase this articulation by replacing the “why” with a “how.” Unlike the driving example (drinking makes driving more dangerous), in the blackmail case it is harder to demonstrate the connection between the components that produces this “normative synergy” of two rights that make a wrong.
In the case of blackmail, this connection takes the form of a condition: the performance of the act is conditioned on not fulfilling the demand. This emphasizes the centrality and importance of the choice made by the victim of blackmail, and the presence of an exchange.
2. Blackmail and Bribery: Another formulation, which has earned the title “the second paradox of blackmail,”Footnote 10 focuses on the disparity between so-called blackmail and what seems to be the symmetric transaction of bribery initiated by the other side. If the wrongfulness lies in the exchange, why should there be a difference between the two directions in which this exchange is conducted?Footnote 11 This formulation brings to mind the importance of the identity of the act’s initiator. It hints, therefore, that what makes the blackmail wrong is not only the content of the transaction, and that we must pay attention to the agent rather than solely to the blackmailer’s act.
3. The Vague Borders of Hard Bargaining:Footnote 12 According to this formulation, the so-called blackmail cannot be distinguished from other transactions which are conceived as legitimate, both legally and morally, such as threats to enforce one’s rights through legal proceedings, to disconnect commercial relations,Footnote 13 or to refrain from selling products in great demand. Again, we encounter a structural similarity to other permitted exchanges, but now the similarity includes even the identity of the initiator!
To solve this problem it is necessary to distinguish, not in an ad-hoc manner, between these groups of cases. This can be done in two ways: the first is to show that some negative feature of “blackmail” transactions does not appear in the other permitted transactions. The second way is to admit the similar wrongfulness of both of these kinds of transactions, prima facie, but to nevertheless point out the special advantage or benefit unique to the permitted transactions, which makes them, all things considered, tolerable or even desirable.
4. Expanding the Victim’s Options: Coercion seems to be wrong since it limits the opportunities open to the coerced party. This formulation of the paradox stresses that blackmail transactions are characterized by an expansion of the victim’s options.Footnote 14 The conditioned threat gives victims another option, which they did not have before: to buy out the threat.Footnote 15 This option is usually better for the victim (after all, blackmail often works). Hence, it is argued, nothing can be coercive here.
Surely, though, the same argument can be applied to the regular case of extortion: threatening a forbidden act (e.g., “if you don’t pay me, I’ll kill you”). In this case also, the mere offer expands the victim’s possibilities, and is built on the expected preference of the victim to pay rather than die. The only difference is that in the case of extortion, the victim is entitled to both his money and his immunity against violation of his rights, whereas in blackmail cases there is an apparently legitimate transaction (since the victim is not entitled to the threatening party’s abstention from fulfilling the threat).Footnote 16 This formulation of the problem, then, seems to collapse into the previous formulation, which hinges on the nature of the exchange between the parties.
However, blackmail expands the victim’s options only on the premise that the threatened act would have been executed absent any connection to negotiation with the victim. In these cases, presenting the option of buying immunity indeed benefits the victim. Yet as long as the threat evolves solely from the possibility of negotiation, with no other reason or intent to execute the threat independently, there is no expansion of the victim’s possibilities.Footnote 17 The additional challenge of this formulation is, therefore, to distinguish between two kinds of threatened acts: those which would have been executed anyway, and those which were born for the sake of negotiation.
To sum up, the first articulation (“the synergetic miracle”) focuses on the connection between the two components of blackmail that make it, somehow, impermissible. The second (“blackmail vs. bribery”) concentrates on the centrality of the actor, noting that the act’s permissibility seems to depend on which of the parties initiated it. The third articulation points out the parallels between blackmail and other legitimate practices, showing that something else, beyond the mere structure of this transaction, is needed in order to account for the problem of blackmail. And the fourth, which elaborates on the fact that the blackmailer’s proposal actually expands the victim’s options, highlights the premise that the blackmailer is not expected to execute the threat and only makes it due to the chances present of gaining something from it by means of blackmail.
The problem with blackmail is multifaceted. An adequate theory of blackmail should respond to all of these worries. This is where the coercion-based and the subjectivist theories of blackmail come in.
2. Revisionist Theories: Blackmail, Extortion and Coercion
A revisionist theory, which denies the permissibility of the threatened act, holds that blackmail is as wrongful as ordinary extortion. Al, who threatens to burn down Ben’s business unless paid, commits wrongful extortion. The wrongfulness does not stem from the threatened forbidden act itself, which may never be performed. It stems from the way the threat affects the recipient’s volition, leading Ben to succumb to the demand, in a context where proper consent is required.Footnote 18
In other words, the problem with extortion is that it involves obtaining something through coercion. According to Nozick’s classic formulation, coercion is a case where the coercer (C), who wants one’s victim (V) to do some act (Φ), claims that C will bring about an outcome which shall turn V’s decision not to Φ less preferable to V than Φing.Footnote 19 The claim is coercive if (and only if) it tries to induce V’s choice by proposing to inflict loss on V, if there is no surrender to C’s demand.
Surely, loss and gain are defined in relation to a certain baseline.Footnote 20 A widely accepted view, which I will not try to defend here, holds that this baseline should be moral.Footnote 21 According to the moralized baseline view, a conditional proposal is a coercive threat where the proposal indicates that if the recipient denies the demand, the proposer will make the recipient worse off than the recipient ought to be, or is entitled to be. In simpler words, a conditional proposal is a coercive threat only if it would be wrongful for the agent to carry out the threatened act.Footnote 22 Indeed, the impermissibility of the threatened act is the cornerstone of common legal doctrines of extortion or duress.Footnote 23
Note further that it is moral rather than legal impermissibility which is of importance. While the most grave moral wrongdoings are often also forbidden by law, a conditional threat of murder would be coercive even in a pre-legal condition, or if the legal prohibition is abolished for some reason. On this suggestion, coercion, as a moral concept, consists of conditionally threatening a morally impermissible act. Extortion, as a legal concept, consists of acquiring something through coercion, where proper consent is required (as, e.g., in the context of transferring property, sexual relations, etc.). There is no need for the threatened act to be illegal for the conditional threat to be coercive, and its use for appropriation of property (or other protected interests) criminalized. Consequentially, if blackmail involves a threat to act in a morally impermissible way, it is on par with regular extortion, and should be legally banned, as a wrongful way of gaining property.Footnote 24
Much more could be said here, of course, and much has been said elsewhere, in defending this view.Footnote 25 Moreover, some issues are still left unresolved.Footnote 26 For the purposes of this paper, though, I will assume—in line with this general picture—that conditional threats to commit morally impermissible acts are indeed properly criminalized as extortion.
However, this falls short of solving the puzzle of blackmail, since some paradigmatic cases of blackmail involve threatened acts that seem to be perfectly permissible, even from a moral point of view.Footnote 27 For example, revealing embarrassing but true and relevant information about a public figure might be—at least in some circumstances—morally permissible or even required. Nevertheless, anyone who attempts to extract money from this figure by threatening to reveal the information unless paid is widely considered to be committing the crime of blackmail, not to mention acting wrongfully.Footnote 28 Should we abandon the revisionist, coercion-based theories? This is where the subjectivist theory comes into play.
3. Motives, Intentions and Rightness: The Subjectivist Theory and its Limits
This section explains how, through the lens of a subjective notion of rightness, typical blackmail scenarios are nonetheless revealed to involve a proposal to act impermissibly, despite the apparent permissibility of the threatened act. I start by briefly surveying existing subjectivist theories, namely theories that relate the wrong in blackmail to subjective features of the blackmailing agent, and explain why a new theory is needed. Relying on the distinction between intended and merely foreseen harm and its alleged normative ramifications, I then show how the blackmailer’s bi-conditional proposal indicates that the harmful component of the proposed act is intended by the blackmailer, and explain when such a finding should suffice for this act to be deemed impermissible, and the proposal to commit it—coercive. This section then goes on to explore challenges to the view that intentions are relevant to permissibility, which cast doubt on this argument. Section 4 is devoted to dealing with those challenges.
a. Background: Intentions and (Im)Permissibility
The second and fourth formulations of the blackmail puzzle draw attention to the agents who make the blackmailing proposal, and to their (explanatory) reasons for doing so. This has led a few scholars to ascribe the wrongfulness of blackmail to the blackmailer’s intentions. Wendy Gordon claims that “[t]he deontologic case against blackmail seems clear: One person deliberately seeks to harm another to serve her own ends—to exact money or other advantage—and does so in a context where she has no conceivable justification for her act,” thus making it an “unjustified intentional infliction of harm on another to benefit one’s self.”Footnote 29 Likewise, Levy stressed the fact that blackmail involves “deliberately inflicted harm.”Footnote 30 Grant Lamond has also proposed viewing those motives as the basis of blackmail’s coerciveness.Footnote 31
Mitchell Berman was the first to integrate the motive into a coercion-based revisionist theory, by highlighting the motive blackmailers would likely have for doing as they threaten, rather than the motives for issuing the threat.Footnote 32 According to Berman, the wicked motive (or some other close mental state)Footnote 33 might make the threatened act wrongful, even if it would not be wrongful in the absence of this mental state [“the subjectivist hypothesis”].Footnote 34 It will thus render the conditional threat coercive, as a threat to act impermissibly. This hypothesis rests on a non-consequentialist comprehension of right and wrong, according to which the right act is defined not only by the consequences that it will bring about,Footnote 35 but also by other elements which characterize the action. The theory’s second component—the one which indeed gives Berman’s theory its name—is the claim that the setup of the blackmailing proposal can serve as sufficient evidence for the existence of this wicked motive behind the threatened action (were it to take placeFootnote 36) [“the evidentiary hypothesis”].
Packing it all together, the evidentiary hypothesis shows that a wicked motive accompanies the threatened act in blackmail scenarios; the subjectivist hypothesis holds that such a motive renders this act impermissible; and finally—the revisionist coercion-based theory holds that a conditional threat to act impermissibly is coercive, hence a demand backed by such a threat is on a par with ordinary extortion. Therefore, blackmail is properly criminalized, since like extortion it is a form of acquiring something through coercion, in a context where voluntary consent is required.Footnote 37 The suggested solution reveals how what at first glance appears to be a case of proposing to do a permissible act, is actually a threat to act impermissibly, once one notices other factors that impact the act’s deontic status. Thus, the paradox of blackmail is not only resolved—actually it is explained away, by denying its premise regarding the permissibility of the threatened act.
According to Berman’s own new version of the subjective hypothesis, “an actor wrongs another if he knowingly causes him harm without reasonably believing that producing that harm is consistent with the balance of undefeated moral reasons under the circumstances,” “based on reasonable and appropriately thorough moral evaluation.”Footnote 38 This view affects the content of what needs to be demonstrated at the evidentiary stage, and complicates it. As others have noted, the focus on beliefs also raises several difficulties regarding the relevant notion of permissibility.Footnote 39
I therefore wish to concentrate on a version of a subjectivist theory which is closer, but not identical, to Berman’s original view. This version is grounded in a more widespread notion of permissibility: one that rests on the distinction between intended and mere foreseen harm, a common distinction in intuitive judgment.Footnote 40 This subjective criterion, I argue, is also less expansive in its “evidential” assumptions. In what follows I briefly present this subjective criterion and the accompanying evidentiary hypothesis. I then move on to discussing the challenge this view faces in contemporary moral theory.
i. The Relevance View: Intending and Foreseeing Harm
Is the agent’s mental state relevant to the moral permissibility of the action? A widespread view (which I shall refer to as “the relevance view”Footnote 41) holds that it is, by distinguishing between bringing about intended harm and bringing about (mere) foreseen harm as a side-effect. This view is not only part of commonsense morality: it has been argued that such a view is required in order to accommodate the doing-allowing divide,Footnote 42 and maybe even for the coherence of any deontological theory.Footnote 43 A related view lies at the root of (at least one version of) what is known as the Doctrine of Double Effect (DDE).Footnote 44 According to this reading of DDE,Footnote 45 two acts, which are alike in their non-mental characteristics, differ in terms of permissibility simply by virtue of differences in the agent’s intentions. Intentions are thus intrinsically relevant to permissibility.Footnote 46
Note that the intending-foreseeing distinction should not be formulated in terms of an absolute dichotomy between forbidden versus permissible acts. It is only necessary to assert that it is morally worse (namely harder to justify) for an agent to perform an action while intending its resultant harm than to perform the same act while merely foreseeing that same harm, and intending other legitimate results.Footnote 47 Hence, acts which are “marginal” from a moral perspective, namely that come close to the impermissibility threshold, might be either permissible or impermissible, depending on the intentions of the agent who commits them.
Returning to blackmail scenarios, the argument goes as follows: the threatened act, which has some merits, involves harmful results to the victim. Generally, when these results are not intended, the act is permissible (due to its merits), albeit only marginally. However, if it could be shown that blackmailers intend to inflict harm on their victim by their act, this intention might be pivotal in making the threatened action impermissible after all.Footnote 48 As a result, the conditioned proposal becomes a threat to commit an impermissible act, hence coercive.
At first glance, if the subjectivist theory of blackmail could be constructed along these lines, it would be theoretically attractive. All that is left to show is that blackmail scenarios indeed contain an intended harm. This is where the “evidentiary hypothesis” comes in.
ii. Threatening as Intending Harm: The Evidentiary Hypothesis
Suppose that the relevance view is true, and that the permissibility of the threatened act in blackmail scenarios is indeed marginal, making the question of the agent’s intention pivotal for the act’s permissibility. Why should we think that the threatened act in blackmail involves an intention to harm the victim? Berman’s evidentiary theory offers a quite straightforward answer. First, the very fact of the conditioned threat indicates the threatening party’s belief in the harmful potential of the threatened act.Footnote 49 Otherwise, there is no basis for expecting the threatened party’s acquiescence to the accompanying demand.
Second, the blackmailer’s willingness to accept money in order to refrain from acting indicates that the motivating reasons behind the threatened act could not have beenFootnote 50 the kind of reasons which justify it, morally speaking. According to Berman, “A reasonable factfinder could suspect that, had any of these [good] interests motivated D [the threatening party], he would not have offered to sell H [the threatened party] his silence.”Footnote 51 Indeed, this argument is about “inferring bad motives, not deducing them,”Footnote 52 and hence establishes only a rebuttable presumption. However, in the absence of such rebutting evidence, it has strong probative value, hence the conclusion that the threatened act is meant to rest on bad motives still stands.
A distinction should be drawn between two different lines of thought that are present in Berman’s argument. According to the first, revealing the secret is morally problematic pro tanto, due to its negative effect on the secret-owner. However, generally it is nevertheless permissible, all things considered, due to its positive value, be it in terms of exercising free speechFootnote 53 or of caring about the interests of others who might gain from knowing the concealed information. The financial demand serves, according to this line of thought, as evidence for the lack of right motivating reasons behind this secret-telling, which—presumably—is incompatible with such a demand. Thus, there is nothing to outweigh the pro tanto wrongfulness.
However, this inference seems to me as unconvincing: a person might genuinely care about the interests of the other, but nevertheless have a stronger preference for personal gain. Acknowledging that “every man has his price” does not entail an absolute lack of moral reasoning. The blackmailer might be motivated to reveal the secret by the right reasons, yet nevertheless find the opportunity to gain some money to outweigh those reasons. One might even suggest that a higher financial demand indicates a strong “good” motivation in favor of the act, which requires substantial compensation in order for one to abandon it. Standing alone, then, the financial demand seems to have poor evidentiary power: the blackmailer might still be motivated (also) by the right reasons.Footnote 54
Therefore, I suggest another line of reasoning, according to which the evidence focuses not on the lack of reasons in favor of the act, but rather on reasons against it, which emerge from the fact that its harm is intended. An example may help: a blackmailer (B) who threatens to reveal the victim’s (V) adulterous behavior to V’s spouse might be motivated also by his belief that V’s spouse deserves to know the truth. Yet if V agrees to the revelation, or even volunteers to confess, it will foil B’s plan. The fact that V will be harmed by the revelation is required for B to be able to extract money from V. V’s harm is not a regrettable side-effect of telling the secret. Rather it is means to B’s end in threatening to do so.Footnote 55 The act B is threatening to commit is, therefore, an act of causing intended harm. According to this line of reasoning, the evidence does not preclude the good motive which gives the action its positive value, but rather indicates a harmful intention which gives the action its negative value. It derives from the inner logic of the blackmailing proposal, rather than from a guess regarding the blackmailer’s psychology.Footnote 56 It thus offers a more stable basis for the subjectivist theory.
Note, however, that even within the new framework, the pro tanto problematic nature of the threatened act has a role in the theory. After all, any commercial offer (which includes, at least in implied manner, a statement like “unless you pay me [5$], I will not give you [these donuts]”) is based on a similar plan to utilize the buyer’s interest in the merchandise. Hence, the donut seller offers to intentionally frustrate the customer’s interest if she does not pay, and will similarly be disappointed to find out that the customer has no such interest. How does the case of blackmail differ (recall the third formulation of the paradox!)? The answer lies in the initial nature of the act and the principle of marginality. Even without much detail about the harm-causing, suffice it to suppose that while refraining from selling the donut is—under normal circumstances—legitimate, blackmail cases involve acts which are more problematic pro tanto. Hence, the difference might be just a matter of being pivotal: while the additional component of intending to frustrate the interests of the victim is sufficient to render the disclosure of a secret impermissible, it is not sufficient to turn not distributing one’s donuts into a morally impermissible action. As long as the act is morally problematic and marginal in the first place, it’s possible for the intention-related component to be pivotal in making it, all things considered, impermissible.Footnote 57
To conclude, blackmail cases are cases of threatening to perform an act which is morally problematic due to its substantial negative effects on someone else. In general, the balance of considerations renders this action (marginally) permissible, despite its problematic pro tanto features. However, when these negative effects are intended, and not merely foreseen, the overall balance changes, and the action becomes impermissible. The accompanying demand—maybe together with other circumstantial evidence that may support or rebut this inference—indicates that those negative effects are intended. Thus, a proposal to act in this way is a proposal to act in an impermissible way, i.e., a coercive threat. A demand backed by such a threat is an attempt to achieve something through coercion, in a context where voluntary exchange is required. This is why it is wrong (and properly criminalized).
The argument can be presented as resting on the following four claims:
I. The threat (or basic coercion) principle: A proposal to Φ is a coercive threat, hence (prima facie) wrongful, if Φ is impermissible.
II. The intending-foreseeing principle: For any Φ which entails harmful results, if those results are intended by the agent who performed Φ, then Φ is morally worse than a parallel Φ where those results were merely foreseen.
III. The evidentiary hypothesis: Financial demands of the sort that prevail in blackmail scenarios are proper evidence for the existence of such intent.
IV. The marginality hypothesis: Blackmail scenarios involve proposals to act in a morally marginal (or borderline) way, therefore the question whether Φ is permissible or not might depend on whether their harm is intended or not.
This is an appealing picture. It explains blackmail as a form of basic coercion, and thus as parallel to extortion; hence, it accounts for the customary design of the criminal code.Footnote 58 It clarifies the connection between the two components of blackmail—the threatened act and the conditioned demand—and explains how together they generate the result of impermissibility. It reveals why a bribery case, which has much less evidentiary force,Footnote 59 is different from blackmail; and it shows why blackmail differs from other sorts of legitimate bargaining (due to the marginality hypothesis). Furthermore, it captures the intuitive force of some other central suggestions, which are based on the way blackmailers would have acted had they not met their victim;Footnote 60 and it fits common pre-theoretical intuitions about the various blackmail-like scenarios, as was lately shown in experimental work.Footnote 61 May we, then, stop worrying about the puzzle of blackmail?
b. The Flaws of Intentions-Based Theory
While the view that the intentions of agents are relevant to the permissibility of their actions was once widespread and influential, in recent decades it has been more often doubted than defended.Footnote 62 By now, supporters of the relevance view seem to be a minority among moral philosophers.Footnote 63 What matters for permissibility, according to the alternative view, are only the objective (or justifying) reasons for action, rather than the motivating (or explanatory) reasons which actually guided the agent’s behavior. If genuine reasons justify an action, it is permissible. The reasons upon which the agent happened to act are irrelevant to the moral status of the act itself.
For example, diverting the infamous trolley from its original track (saving five and killing one) is permissible in the given circumstances,Footnote 64 and this conclusion does not change even if the agent who diverts the trolley actually wants to kill the person who is standing on the other track, and would divert the trolley even if the original track was vacant.Footnote 65 Thomas Scanlon and others have suggested that opposing intuitions derive from failure to distinguish between moral judgments that relate to actions and judgments that relate to agents.Footnote 66 The agent who acts with illicit intentions is rightly criticized and condemned, but the deontic status of the act remains untouched.
Furthermore, other claims target the conceptual divide between intending and foreseeing harm as incoherent.Footnote 67 After all, even in the paradigmatic cases of intended harm, the harm can be presented as a mere side-effect. For instance, in the famous transplant scenario, one can argue that the agent intends only to harvest the patient’s organs, while the patient’s death is only a regrettable side-effect, which the agent would have been happy to avoid if it were possible. Put more generally, as long as one’s ultimate end requires the harmful means only causally and not logically, the harm can be regarded as unintended.Footnote 68
This kind of difficulty may motivate a fundamental alternative view of the DDE. This view does not focus on the subjective mental state of the agent, but rather on the objective causal structure of the action.Footnote 69 It examines whether the action’s harmful consequences were part of the causal chain on the road to the desired end. It is thus a Patient/Causal-Centered Version (“CCV”). According to this formulation, transplant is wrong not because the physician intended to kill the patient, but because the patient’s death occurred along the causal chain of saving the other five patients, whereas in the classic trolley case it does not. According to CCV, an action which maximizes the good might nevertheless be impermissible if this maximization is achieved through harming someone else.Footnote 70
While the two formulations of DDE largely converge, there could be a difference between them in various cases. Thus, again, according to CCV, killing the one and saving the five in trolley is permissible even if the agent wishes to kill the one and intends to do so (and even if the agent would have flipped the switch in order to kill this person anyway).Footnote 71 Note that what is of importance is not the causal chain towards the realization of the agent’s final end,Footnote 72 but the causal chain toward the good result which made the action prima facie permissible.Footnote 73 Since in this example the death of the one does not lead causally to the saving of the five, flipping the switch does not violate this one’s rights. Permissibility thus remains a matter of objective fact, with no relation to the agent’s mental state.
We can now return to our tentative reliance on intended harm as a component of the subjectivist theory of blackmail. The theory is based on the mental state of the agent, more specifically on the agent/mental-state-centered version of DDE. As described above, the harm to the threatened party is indeed a constitutive component of the blackmailer’s plan, which focuses on the possibility of extracting money from the victim. However, the evidentiary thesis seems to fail once one tries to formulate it in the Patient/Causal-Centered Version, since the harm to the threatened party has no role in the relevant causal chain.Footnote 74
Another look at the informational blackmail example will clarify this point. In this case, as described above, disclosing V’s secret is permissible due to, let’s say, the interest of V’s spouse in knowing the truth. While the harm to V serves the blackmailer’s final end, it is not part of the causal chain that leads to what justifies the act, namely the spouse’s knowledge. Hence, V’s harm cannot be seen as (merely) the means to maximize the good. In this state of affairs, according to CCV there is no deontological constraint against disclosing the information. Therefore, the blackmailer’s proposal is a proposal to act permissibly, and hence it involves no coercion (at least not under the simple account of coercion, as presented above).
To sum up, the revised subjectivist theory hinges on the intending-foreseeing divide. Alas, this divide and its normative significance are roundly criticized. Some tend to completely disregard this distinction. Others formulate this (or a close enough) distinction not in terms of mental states, but in terms of causal chains. Either way, the subjectivist theory doesn’t work, since even assuming that the blackmailer’s proposal indeed indicates one’s intent to harm the blackmailed party, this does not affect the permissibility of the proposed act. Hence, if this act is generally permissible, proposing to commit it unless paid is not coercive.
The subjectivist theory works, then, only for those who still uphold the relevance-of-intentions view.Footnote 75 While philosophical arguments should not be decided on the basis of head-counting, it sounds unpromising to base one’s theory of blackmail on a view which is constantly losing support in the general field, and to undertake the liability of unattractive positions (for independent reasons) in this field.Footnote 76 On the other hand, those who have abandoned the relevance view might see it as the false source of intuitions regarding blackmail, thus leading them to advocate the decriminalization of blackmail.Footnote 77
Before arriving at any of these conclusions, let us try to rescue the subjectivist theory by investigating whether it can survive the failure of the relevance-of-intent principle. In other words, we have to reconstruct the subjectivist theory without resorting to this hypothesis. The following section suggests two such strategies.
4. Rescuing the Intention-Based Theory
Let us recall where we stand now: the coercion-based solution to the blackmail puzzle is rooted in insisting that blackmailers are not proposing to do something they are entitled to do. Rather, upon closer examination it was revealed that they are proposing to act impermissibly. The heart of the theory lies in the claim that what seems to be, at first glance, permissible (though only marginally), turns out to be impermissible when performed out of a specific mental state (such as intending harm). Since the conditioned threat indicates the existence of this mental state, blackmail cases would seem to be threats to act impermissibly. However, in the final account doubt was cast on this argument, since the view that intentions are relevant to permissibility has been challenged. In this section, I will attempt to suggest two strategies for rescuing the subjectivist theory, while denying the relevance of intentions. The first focuses on the indirect relevance of intentions. The second, which is indeed more ambitious, suggests a minor correction to the general theory of simple coercion, thus making intentions relevant to coercion even without being relevant to permissibility.
a. A Double-Evidentiary Thesis: Intentions as Prediction or Cause
So far, the discussion has been devoted to assessing the intrinsic role of intention in evaluations of permissibility. Those who object to the relevance view deny that an agent’s intention, by itself, generates an important difference in the normative properties of the action performed by this agent, all other things held equal. In other words, the debate over intentions is not whether intention has any sort of ramifications for the permissibility of an action, but whether it has any such intrinsic ramifications.
However, leaving aside the cleaned-up hypothetical examples, in real life no one denies that one’s intentions might affect the permissibility of one’s action, albeit only instrumentally.Footnote 78 First and foremost, intentions have what Scanlon termed “predictive significance” as to the foreseen consequences of the act.Footnote 79 For example, while the interesting philosophical debate about terror bombing (which intends the killing of civilians) versus strategic bombing (which merely foresees it) focuses on cases of equivalent acts and consequences, in reality the difference in intention is expected to result in different ways the bombing is planned and carried out, hence in its consequences as well.
In the same fashion, revealing V’s secret in order to damage V may plausibly be predicted to be carried out in a way that will cause graver damage to V than a revelation motivated by care for V’s spouse. It might, for example, affect the extent of publication (whether C reveals the secret to V’s coworkers too), or its form (whether C presents the data in a gentle or shocking way). The intention thus affects the predicted consequences of the action. The threatened act in blackmail scenarios is predicted to bring about graver harm to the victim than regular cases of revealing secrets. This difference might suffice in order to render this originally marginal act impermissible.
Moreover, sometimes the agent’s intentions have not only predictive or indicative value, but also causal value, meaning that the very existence of the intention leads to a bad consequence. If, for example, people tend to experience an intended harm in a more severe manner than they experience mere unintended harm, then the intention affects the permissibility of the action derivatively, by expanding the scope of the expected harm.Footnote 80 Along these lines, the victims of blackmail might experience the harm in the revelation of their secret as more serious when they know it was revealed in order to cause them pain. Thus, again, intentions are relevant to the intended act’s permissibility, not intrinsically but through their effect on the consequences of this action.
This tentative conclusion opens the door for a double-evidentiary theory of blackmail: the conditioned threat indicates that the threatened act will (hypothetically) be executed with the intention of bringing about harm to the victim. This harmful intention, in turn, indicates that the magnitude of the harm is expected to be more serious than in the parallel no-intention case. It is this latter factor that turns the act from being generally and ordinarily permissible into being wrongful when done intentionally. Thus, as a matter of evidence, blackmail is a threat to do what is (expected to be) impermissible.
However, this is not enough. While intended harm usually indicates more severe consequences, this is not necessarily the case. One can imagine cases of blackmail where the threatened act has the exact same consequences as in the non-intended case. If the theory cannot explain their wrongfulness, it remains under-inclusive. A failure to account for some cases of blackmail also implies that it probably fails to trace the root of the problem.Footnote 81 Therefore, we should still look further, for another possible strategy.
b. Beyond Impermissibility: Coercion, Consent and Culpability
So far, I have discussed the way the criminalization of blackmail can be reconciled with no more than the ordinary account of simple coercion, which consists of conditionally threatening to act impermissibly. Yet given the nuanced normative evaluation of acts and actors (as stressed above), perhaps the traditional characterization of coercion is inaccurate, neglecting the difference between the act and the agent to the same extent as holders of the relevance view do. In other words, perhaps the distinction between coercive threats and permissible offers can be reformulated in terms of the agent’s culpability (or something close to it) rather than in terms of permissibility. According to this amended view, a conditional proposal is a coercive threat if the agent that will carry out the threatened act would be culpable for doing so. If intentions are relevant to the assessment of culpability,Footnote 82 and culpability—rather than only impermissibilityFootnote 83—affects coerciveness, then the subjectivist theory can survive the rejection of the relevance view regarding permissibility, but still explain the wrong in blackmail in terms of coerciveness, and hold that blackmail is just a subcase of ordinary extortion.Footnote 84
Why should we think that culpability, even without impermissibility, affects the coerciveness of proposals? To answer this question it is necessary to develop no less than a comprehensive theory of coercion, a task I cannot undertake within the limits of the present paper. Instead, I shall provide an initial outline. I first claim that the wrong in extortion is about acquiring something without proper consent. I then analyze the special role coercion claims have in the context of evaluating consent, and how it differs from other contexts, such as excusing someone from responsibility or placing blame on the coercer. I explain why the threatening party’s culpability might be relevant to the extent that the threatener is entitled to rely on the threatee’s consent. Finally, I demonstrate how these claims fit into the classic blackmail scenario.
i. What is Wrong with Extortion?
Revisionist views on blackmail hold that there is nothing special about the criminalization of blackmail. Rather, it should be regarded as a general case of extortion, with the subjectivist theory explaining its coercive component. Both Berman and Shaw, the prominent advocates of earlier subjectivist theories, focus on the wrong in coercion, which consists of attaching an improper sanction to one’s deliberation, or placing improper pressure on one’s choice, in a way that violates one’s freedom or autonomy.Footnote 85 If the wrong of blackmail derives solely from the wrong of coercion, it should presumably be limited only to cases where the coercee is entitled to act free from this pressure, meaning presumably cases in which the coercer is not entitled to act as threatened.
In contrast, I have insisted earlier that blackmail and extortion should be understood as an instance of theft by coercion.Footnote 86 In other words, it is not the coercion per se which constitutes the wrong, but the acquiring of something of value through the use of coercion as a means to this end. At the heart of the wrong in extortion and blackmail is the fact that the victim’s consent is defective, due to being obtained in a dubious way. In this way, the moral wrong of coercion turns to be the legal wrong of acquiring the other’s property without the owner’s proper consent. Focusing on consent and the way it is invalidated by coercion prepares the ground for a more careful examination of the role coercion has in the wrong of blackmail (and extortion in general).
In a seemingly unrelated paper, Mitchel Berman famously distinguished between two different functions of coercion claims: an offensive function, suggesting when an agent is wrongful in coercing and a defensive function, which serves as a call for excusing a coercee from blame for what had been done.Footnote 87 As Berman persuasively argued, these are totally different (and logically independent) functions. Berman himself focuses his discussion of blackmail on the offensive meaning only, since apparently what is at stake is the criminalization of blackmail, which involves the offensive function. I wish to suggest, however, that having it hinge on something which is closer to the defensive function is the key to a full understanding of the wrong in extortion, hence also in blackmail. Let me elaborate.
ii. The Third Function of Coercion Claim: Coercion and Consent
If the wrong in blackmail derives directly from its coerciveness, then one should indeed focus only on simple coercion, which depends—as presented earlier—on the permissibility of the threatened act. In contrast, to the extent that the wrongfulness is rooted in taking someone’s property without proper consent, through the use of coercion as a tool for extracting consent, agreement or a promise, it is possible to recognize a third category of functions of coercion claims. This category resembles—yet is not identical to—the defensive function that relates to the conditions in which a coercee is excused for its deeds.Footnote 88 Appreciating this point requires a closer look at the nature of consent.
Consent is morally transformative: it turns an otherwise illegitimate act into a legitimate one by waiving one’s objection to the transgression of one’s moral boundaries.Footnote 89 Thus, the conditions in which the victim’s consent should not be held to be transformative, or—relatedly—this promise should not be deemed obligatory, are not identical to the conditions in which the victim would not be responsible for one’s physical actions, or would be excused for violating any duties or transgressing another’s rights. If V—a victim of coercion—did something wrong (e.g., committed theft) under a coercive threat, V might be excused for acting wrongfully.Footnote 90 In contrast, when V breaks a promise extracted through coercion, or insists on the right surrendered under coercion, V does nothing wrong to be excused for. The promise was not binding and the consent was not valid in the first place. In other words, there are two different levels of coercion here: the level of coercion which excuses someone from being responsible for acting wrongfully is different from the level of coercion which denies one’s consent of its transformative nature, or precludes one’s promise from being obligatory.Footnote 91 A given level of pressure might not excuse from wrongdoing but still suffice to render the promise nonbinding.
Moreover, the difference between excuse-related coercion and consent-related coercion is not only a matter of the degree of required “freedom,” or the extent to which one’s choice is constrained. It is rooted in another feature of promising or consenting: the fact that both are dyadic-relational actions, performed between two agents. For that reason, as long as the coercing party is the same party who gains the consent or is the promisee, there is room for a comparative-relational account of consenting and promising. Under this model, in order to determine the moral implications of one’s consent it is necessary to evaluate the behavior of both of the parties, rather than only the constraints faced by the consenting party. What is called for is comparative evaluation, whether or not the consent was efficacious regarding the rights and duties of the concrete other party; or whether a promise was binding towards the concrete promisee. It is unfair for the coercing party to benefit from consent that was obtained by one’s own faulty deeds. Thus, the consent should not be effective as to this party.Footnote 92
This comparative notion of coercion is actually also a distributional one. It asks which of the two parties—the coercer or the victim—has to bear the cost of the retreat from what was seemingly (the manifestation of) V’s consent or promise. While generally persons have to bear the costs of their consent or promise even if they do not stand behind it anymore, sometimes it is justified to shift the costs of this retreat (or gap between intention and manifestation) onto the other party, who induced this consent in dubious ways.Footnote 93 Focusing on the distributional feature of such a claim opens the door for an evaluation of the parties vis-à-vis the consent, rather than focusing merely on the impermissibility of their acts. In other words, if one of the two parties has to suffer the consequences of the consent, it should be the more culpable one.
iii. Back to Blackmail
How does this all tie in with the case of blackmail? In the basic blackmail scenario, C authors V’s distress by threatening to reveal V’s secret, thus inducing V’s consent to give to C V’s money. In doing so, C is acting maliciously, since even if intending V’s distress does not affect the permissibility of the act (i.e., revealing the secret), it surely does affect the evaluation of the actor, namely the culpability of V vis-à-vis that action. Therefore, under the distributional model of coercion, V’s consent (to give C money) should not be regarded as transformative as to C’s rights and duties, nor should V’s promises be binding against C.Footnote 94 C, the blackmailer, is not entitled to earn from his malicious behavior, and so V’s consent is not transformative as to C’s rights in V’s money. Consequentially, by taking V’s money C is taking money without its owner’s proper consent. By doing so, C commits theft by coercion, an act which is properly criminalized as an instance of theft,Footnote 95 even if the act V is threatening to do is permissible after all.
Note that this suggestion is still part of what I earlier termed a “revisionist theory,” namely one that sees blackmail as basically just a form of extortion. Extortion involves coercion, and so does blackmail. Extortion consists of taking one’s money without one’s proper consent, and so does blackmail. What is special in the case of blackmail is only that it is more complicated to demonstrate the coercion, namely, to explain why the victim’s consent should not be regarded as properly transformative as to the blackmailer’s rights in the money extracted. For that purpose I introduced the distributional account of consent, which strives to justly allocate the costs of consent that was given out of distress, when the beneficiary is responsible for this distress.
To fully appreciate this view, let me briefly examine it against two other examples. First, let us take Thomson’s famous vengeful doctor example,Footnote 96 where the doctor injects the suicide-assisting drug out of vengeful motives, intending to cause the patient’s death as an end. Thomson has argued that while the doctor’s character is morally stained, the act is permissible. Imagine now that the doctor is acting within the framework of a contract, willing to come only if paid. In other words, she proposes to the patient: “I will inject the drug only if you pay me a sum of money.” Should my account lead to the conclusion that this contract was extortionist, due to the malicious behavior of the doctor? I believe it should not. Indeed, the doctor acts in a malicious way, and is relatively the wicked party to the agreement. Yet the doctor’s maliciousness is not the cause of the patient’s distress (the pain is). Moreover, what the doctor is threatening to do if not paid—i.e., not to inject the drug—is not the behavior that manifests the maliciousness. Therefore, the patient’s consent to pay is not the result of an intended distress, caused maliciously by the doctor. Under my account, then, vengeful physicians may continue to get paid for assisting suicide.Footnote 97
What about the donut seller? Earlier I discussed the case of the donut shopkeeper, who says that unless money is paid, she shall not sell her donuts. The seller, I argued, intends to cause (some level of) distress in refusing to give the donut, in order to induce the consumer to pay for the donut. Here again, however, there is no real challenge. First, the shopkeeper is not responsible for the consumer’s hunger or sugar craving.Footnote 98 Second, the level of distress caused by being denied a donut is—at least under normal circumstances—not of the sort that should affect the moral power of one’s consent. Lastly, abstaining from distributing your donuts for free is much less morally suspicious than revealing someone’s embarrassing secret, and so the relative-distributional consideration here is much weaker. It is safe to conclude, then, that my suggested theory of coercion and blackmail will not lead to excessive over-inclusiveness, due to making every case of bargaining a case of extortion.
In summary, under the view suggested in this section, in order for a transaction to be extortionist it is not necessary for someone to threaten to act impermissibly. While such a threat is surely the paradigmatic case of coercion, since it is a deliberate and malicious way of causing distress and exploiting that distress for one’s own profit, there are some peripheral instances that share these characteristics without involving an impermissible act. The distributional understanding of consent requires that a consent that originated in someone’s distress not enrich whoever deliberately created that distress in order to induce that consent, as long as it was created in a morally condemnable way. While in most cases deliberately creating distress in a morally-condemnable way is also impermissible, this is not a necessary requirement for the consent to be deemed ineffective as regards the condemnable party. Thus, extracting money (or other things of value) in this way through reliance on someone’s manifestation of consent is a way of taking their property without their proper consent, namely a sort of theft by coercion. Theft is properly criminalized as a way of protecting one’s right to property.
Conclusion
Coercion consists of threatening to execute an impermissible act. This explains well why extortion is wrong, and why some cases of threatening to act legally but immorally are wrongful too. Yet why do threats to execute permissible acts also seem to be coercive? Subjectivist theories suggest that in evaluating the permissibility of an act factors that relate to the agent’s mental state while doing it should be taken into account. In this way, acts which are generally permissible might be found to be impermissible after all.
In this paper, I have suggested an improvement on existing subjectivist theories, by relating to the difference between intending the harmful results of the act and merely foreseeing them, a difference that might negatively affect the moral evaluation of this act. Thus, the same physical act with the very same consequences might be either permissible or impermissible, depending on the reasons which motivated the agent. I demonstrated why the structure of the blackmail transaction indicates that blackmailers intend the harm to their victim. For some actions, which are morally suspicious to begin with, this factor suffices to render them impermissible. Thus, we can conclude that blackmail actually contains a threat to act impermissibly, based on a firmer evidentiary basis and clearer normative comprehension.
This suggestion, however, revealed the theory’s vulnerability to attacks on the relevance of intentions to moral permissibility. If the harm being intended does not affect the threatened act’s permissibility, then blackmail scenarios cannot be coercive. I therefore offered, in the last section, suggestions as to how the theory can avoid this conclusion. First, I explored the way in which the uncontroversial instrumental role of intentions in evaluating permissibility can be integrated into the subjectivist theory. Thus, the agent’s intentions might frequently indicate the seriousness of the act’s consequences, or even have a causal role in creating those consequences.
Second, I investigated the possibility of grounding the phenomenon of coercion not in the permissibility of the act which constrains the choice of the victim, but rather in the relative culpability of the parties regarding this constraint. Relying on a distributional view of coercion and consent, I argued that where someone deliberatively and maliciously creates distress in order to induce the other’s consent, such consent should not be regarded as transforming the rights and duties between those two parties in a way that will benefit the malicious party. If the consent is not transformative, taking the victim’s money is not permissible and should be dealt with as a sort of theft, just as regular extortion should. Thus, even if intentions are not relevant to permissibility, they are relevant for the consent-inducing function of coercion claims. Subjectivity matters, then, although in a more convoluted way than previously argued.
