Jackson Cases

The second problem with the Fact-Relative View is that one might think it gets the wrong verdict in some cases.²⁰⁸ Consider,

Jackson’s Case. All the evidence at Doctor’s disposal indicates, in keeping with the facts, that giving Patient drug A would cure her partially and giv-ing her no drug would render her permanently incurable. However, the evidence leaves it completely open whether it is giving her drug B or drug

208 Another way to put the problem is that, even if the rights correspond to what one ought to do in the fact-relative sense, Jackson’s Case shows that what one ought to do in the fact-fact-relative sense is not particularly important; and so, if rights correspond to the fact-relative ought, rights are not particularly important.

175

C that would cure her completely, and whether it is giving her drug B or drug C that would kill her.²⁰⁹

Suppose that it is drug B that would completely cure Patient, and so drug C that would kill Patient. Given the Fact-Relative View, Patient will have a right that Doctor give her drug B. Yet, we might find this counterintuitive. In fact, we might think that Patient has a right, given all the available evidence, that Doctor give her drug A (as well as rights that Doctor not give her either drugs B or C).

Given that, on the Evidence-Relative View, X’s right that Y Φ depends on the best avail-able evidence to Y, Patient will have a right that Doctor give her drug A (as well as rights that Doctor not give her drugs B and C). We can say that this holds because giving drug A is the expectably best option. So, it seems that cases like Jackson’s Case speak in favour of the Evidence-Relative View, and against the Fact-Relative View.

There are a few options available to the defender of the Fact-Relative View. First, we might argue that the Fact-Relative View can recognise that Patient has a right to drug A.

I make some remarks to this effect in the conclusion to this chapter. However, I also raise doubts with this way of proceeding. So, here I would like to examine other ways we may go.

Second, we might bite the bullet and suggest that, though the Fact-Relative View says that Patient has a right to drug B, this is not as counterintuitive as one might think. For example, we might say it is just unclear which treatment Patient has a right to. We will presumably have a second-best ought to deal with these sorts of case. However, one might press, do we have second-best rights that adhere to this second-best ought? Further, what is the relationship between these second-best rights and the Evidence-Relative View?

I am going to take a third option and suggest that the Evidence-Relative View is suscep-tible to structurally analogous counterexamples like Jackson’s Case. Thereby, cases like

209 This case is taken from (Zimmerman 2014, 30) with some slight alterations, originating in (Jackson 1991, 462–63). See the Miners case also, discussed in (Parfit 2011a, 159–61; Tadros 2011b, 222). Tadros says he is not sure who invented the much-discussed Miners case; since there is a fact-relative ought, he really ought to know.

176

Jackson’s Case give us no reason to prefer the Evidence-Relative View over the Fact-Rela-tive View. Since I have suggested we have reason to prefer the Fact-RelaFact-Rela-tive View over the Evidence-Relative View in the previous section, this means we have most reason to endorse the Fact-Relative View.

Before that, note that even if I am wrong, and so if cases like Jackson’s Case do give us reason to prefer the Evidence-Relative View, we still have countervailing reason to prefer the Fact-Relative View from the previous subsection. On the balance of reasons, I prefer the Fact-Relative View. The argument I make below that Jackson’s Case gives us no reason to prefer the Evidence-Relative View simply makes this verdict more secure.

The problem underlying the Fact-Relative View can be stated as follows.²¹⁰ We have two standards of what X has a right to, Standard 1 and Standard 2. We have at least three acts that the potential duty-bearer could perform. We then have a case in which the po-tential duty-bearer is unsure of what X has a right to on Standard 1, but knows that it has no chance of being act 1. And yet, because the other options could go so badly, we think that the verdict of Standard 2 is correct, and that X has a right to act 1. In Jackson’s Case, we have Standard 1 (Fact-Relative View) and Standard 2 (Evidence-Relative View). Doc-tor is unsure whether Patient has a right to drugs B or C on the Fact-Relative View, but knows both could go badly. So, we think Patient has a right to drug A, as the Evidence-Relative View suggests.

Yet, we can run cases of this structure against the Evidence-Relative View. This is because the best available evidence to duty-bearers—that which fixes what rights obtain on the Evidence-Relative View—is not always going to be luminous: we are not always going to know what the Evidence-Relative View says that we have a right to. And when we do not know, there are some cases where we ought to adhere to some more risk adverse standard than the Evidence-Relative View. Consider,

Smith’s Case. Patienthas some condition. Doctorhas three available op-tions: drugs D, E, or F. Suppose that Doctor has to give some treatment.

210 This formulation of the problem owes a lot to (Smith 2011, 5–7; Littlejohn 2009, 238–41).

177

Doctor knows that giving each of them may cure Patient. However, Doc-tor also knows that giving drug D may leave Patient badly off (e.g., with a bad rash). Giving drug E may leave Patient extremely badly off (e.g., with the loss of a foot). And, giving drug F may leave Patient incredibly badly off (e.g., with the loss of a hand). There is no evidence available to Doctor as to the likelihood of each treatment’s success or failure. Doctor calls Col-league to ask her what to do. ColCol-league says that, “Drug D is not best. As to which of E or F is best…” and then the phone goes dead. Doctor tries to call back, but there’s no luck. Doctor needs to give a treatment now.²¹¹

Suppose Colleague is both Doctor’s epistemic superior and a very reliable testifier. Doctor has no evidence of the respective probabilities of the treatments, so the evidence does not support any of D, E, or F. However, she has excellent evidence that drug D is not best.

So, Doctor knows that Patient will not have a right that she give her drug D. It is just that Doctor does not know whether it is drug E or F that is best.²¹² And yet, because either treatment could go so badly, intuitively Patient has a right to drug D—this is true, even though Doctor knows, relative to the best available evidence to her, that option has no chance of being the one Patient has a right to on the Evidence-Relative View.

(Here is another example (Littlejohn 2009, 238–41). Suppose there is evidence available to you as to what treatment is best, though you are not sure whether you will respond to that evidence correctly due to other pressures. Suppose you know that D is not best but are not sure whether E or F is best—were you to have a little more time, you could figure

211 This case is based on (Smith 2011, 5), with some details changed. For discussion, see (Zimmerman 2014, 69–76).

212 What does Patient have a right to in Smith’s Case given the Evidence-Relative View? With a further as-sumption, the Evidence-Relative View arrives at the verdict that Patient has a right that Doctor give her drug E. This is because the Evidence-Relative View needs to say something about what rights obtain when potential duty-bearers have no probabilities (or insufficiently robust probabilities) concerning what to do.

For example, suppose that Colleague never answered the phone in the first place. If Doctor has to prescribe some drug, Patient has a right that Doctor give her drug D, and rights against Doctor giving her drugs E and F. When faced with no probabilities, Zimmerman plausibly suggests that we ought to apply a principle of indifference, on which all of the options should be assigned equal probabilities. Given that Doctor has good evidence that drug D is not best, drug D is off the table. We assign equal arbitrary probabilities to the remaining options, and so drug E comes out as best, and so the drug Patient has a right to.

178

it out with more confidence. Knowing that if you get it wrong, things could go disas-trously, you ought to take the conservative option that you know has no chance of being best relative to your evidence.)

Just as Jackson’s Case presents us with an example in which the Fact-Relative View tells us that Patient has a right that Doctor give her drug B where, intuitively, Patient has a right that Doctor give her drug A, Smith’s Case presents us with an example in which the Evi-dence-Relative View tells us that Patient has a right that Doctor give her drug E, where, intuitively, Patient has a right that Doctor give her drug D. If Jackson’s Case and Smith’s Case are analogous in all morally relevant ways, Jackson’s Case gives us no reason to prefer the Evidence-Relative View over the Fact-Relative View.

Now, the defender of the Evidence-Relative View cannot appeal to just any old solution to Smith’s Case. This is because the defender of the Fact-Relative View could then use that solution to solve Jackson’s Case. What the defender of the Evidence-Relative View needs is a solution only to Smith’s Case. They need to find a disanalogy between Jackson’s Case and Smith’s Case and draw on that disanalogy to solve Smith’s Case.

The most obvious disanalogy between the cases is that, in Smith’s Case, Doctor receives evidence that D is not best through testimony, whereas there is no testimony in Jackson’s Case. Further, the testimony is not of probabilities on which she can base her judgement.

Rather, it is testimony that drug D is not best. A defender of the Evidence-Relative View could say that testimony of what is best does not give you evidence of which option is best.

So, Colleague’s testimony that drug D is not best does not give Doctor evidence that drug D is not best.

However, this move has unacceptable consequences. Let us amend Smith’s Case such that the phone does not cut out, and Colleague says “Drug D will not be best. As to whether to give either drug E or F, drug E is best.” Colleague is Doctor’s epistemic superior. She is a reliable testifier. It is very implausible that, on the Evidence-Relative View, this does

179

not make it the case that Patient has a right that Doctor give her drug E. So, testimony of what option is best must give Doctor evidence.²¹³

To conclude this subsection, Jackson’s Case gives us no reason to prefer the Evidence-Rel-ative View over the Fact-RelEvidence-Rel-ative View, as the Evidence-RelEvidence-Rel-ative View is susceptible to structurally analogous counterexamples. The task for the defender of the Evidence-Rela-tive View is to find a solution to Smith’s Case that both (i) does not lead to implausible results in other cases and (ii) is not available to a defender of the Fact-Relative View. I am not sure such an answer is available. Again, even if I am mistaken and Jackson’s Case does speak in favour of the Evidence-Relative View, we have countervailing reasons to be scep-tical of the Evidence-Relative View.

5. Conclusion

This chapter has questioned whether rights depend on duty-bearers’ beliefs, the evidence available to them, or the facts. If rights do not depend on the facts, we have our solution to the Problem of Pure Risk. This is because, when it is intuitively plausible that others have rights that we not subject them to risk, often our beliefs and the evidence supports others’ wellbeing being sufficiently weighty to place us under duties. I have argued that rights depend on the facts, and so we cannot solve the Problem of Pure Risk in this way.

Against the Belief-Relative View, I argued that it is unclear why the Moral Belief-Relative View would not be true in place of the Belief-Relative View, but that we have good reason to reject the Moral Belief-Relative View. Further, the Belief Relative View implausibly

213 Zimmerman makes a similar move to solve Smith’s Case, distinguishing between ‘evidence available to someone and the evidence of which that person in fact avails himself of’ (Zimmerman 2014, 72). Whereas he used to formulate his view in terms of evidence available to agents as we have been doing (Zimmerman 2008), he has revised his view so that it responds only to evidence that agents have availed themselves of.

Yet, first, Doctor has availed herself of the evidence available to her in terms of Colleague’s testimony.

Notwithstanding this, second, Zimmerman’s view is going to get our revised version of Smith’s Case wrong.

Third, it gets Real Gun wrong too, for Non-Believer has not availed herself of the evidence as to whether the gun is real.

180

over and undergenerates rights. Finally, more generally, I argued it is implausible that our rights depend on others’ beliefs about the world.

Against the Evidence-Relative View, I argued that it is implausible that our rights depends on the evidence available to others and others’ ability to respond to evidence. I also argued that the Evidence-Relative View implies, when new evidence becomes available, new rights come into existence rather than us gaining better evidence about which rights ob-tain; but this seems mistaken. And, finally, I argued that the Evidence-Relative View im-plausibly undergenerates rights.

All of this gives us good reason to endorse the Fact-Relative View. Yet, there remains the question what to say about Jackson’s and Smith’s Case. What I hope to have shown above is that Jackson’s Case gives us no reason to prefer the Evidence-Relative View over the Fact-Relative View.

During discussion of whether the fact-relative ought can reach the verdict that Doctor ought give drug A in Jackson’s Case, Littlejohn considers what Doctor may think to herself.

He says, ‘if I know I don’t know whether it is drug [B] or C that is best and know that guessing could be disastrous, I ought to give drug A’ (Littlejohn 2009, 238). He builds on this, saying we might actually have a conditional, fact-relative ought to give drug A. On this view, we can see facts about uncertainty entering the picture. Similar to this, we might think that Patient has a conditional right to drug A, conditional on Doctor not knowing which of drugs B and C will fully cure Patient. However, I am unsure whether we are going to reconcile deficient evidence in cases like Jackson’s Case with the deficient evidence of the soldiers in Duped Soldiers. For example, might not the soldiers think to themselves, “Given that I don’t know that those people over the border whom we might attack are innocent, and given I know that not attacking could be disastrous (after all, I’ve great evidence they are terror-ists, planning a terrible attack), we ought to attack them.” I think further discussion of cases like Real Gun, Duped Soldiers, and Jackson’s Case may point to a gap between what rights obtain and what one ought to do, where that ought has a more practical, decision-theoretic flavour. This is something I look to pursue moving forward. Before that, let us see whether we can hold rights against risk on a Fact-Relative View given the Safety Con-dition.

181