I will be brief for most of the questions. And, as announced, I won’t be posting the questions since you were allowed to take them out of the exam hall (so please don’t email me to ask for them).

I will be discarding Question 2 (see my explanation below). Consequently, the final exam will be upon 29 marks, for 40% of your final grade. The highest score was a 27/29. The mean is at 18.06. The 75th, 50th, and 25th percentile scores are 21, 18 and 16 respectively. These are good scores: good job everyone!

Click through to see…


  • Question 1

Option C. Almost all (95%) got this. See L02 Slide #9 (and L01 Slide #33, 35).


  • Question 2 (discarded)

Option D. Almost all didn’t get this. See L02 Slide #11. Option C is wrong because a norm against intentional harming is entirely relevant in this case—Bern’s action didn’t violate it since she dropped the machine unintentionally, and that’s how–despite the fact that she did do something that caused harm–we can conclude that she didn’t do anything morally wrong. But let me expand on the reasoning a bit: so Bern dropped an extremely heavy object on Gene’s foot, causing him great pain. Under normal conditions, this is exactly the sort of thing that would run her afoul of the stated norm. But she has a proper excuse/exculpating factor: yes, she did do something that harmed Gene, but she didn’t do it intentionally. It was an accident. The norm is relevant in the sense that it is the one the consideration of which was triggered, even though the conclusion was that Bern passed the test and was declared as not having done anything morally wrong. Now imagine another (putatively Deontological) norm, say, against intentional deception. Bern also didn’t run afoul of that norm. But consideration of that norm wasn’t even triggered since she didn’t do anything that might have been considered a violation. It’s not a norm relevant to the situation.

(I can see that students are thrown off by this one. On hindsight, I was counting on too subtle a distinction, and one that wasn’t directly covered by the lectures. This question will thus be discarded.)


  • Question 3

Option A. A majority (61%) got this. The thing to keep in mind is that Lloyd could be a consistent Deontologist too—e.g., one who happened to hold to the ‘right’ combination of norms such that his actions come out permissible. There isn’t enough information to rule out such a possibility. One student wrote in to ask if Option D should be accepted since, for all we know, Lloyd is neither a consistent Deontologist, a consistent Consequentialist, nor a consistent Utilitarian. Unfortunately, the logic of the option rules that out. What the options says is basically that Lloyd could not be a consistent Deontologist, or a consistent Consequentialist, or a consistent Utilitarian.


  • Question 4

Option C. A majority (60%) got this. Norcross’ Puppy Argument makes no assumptions about either Consequentialism or Utilitarianism, while the Utilitarian argument doesn’t required the soundness of the Puppy Argument. Carmen’s scenario is not covered by the Utilitarian Argument since the purchase and consumption of the cured veal will not support the institution of factory farming (see L03 Slide #14). Dave is also right—for obvious reasons, a consistent Deontologist should not accept Premise 3 of the Utilitarian Argument. However, her Deontology does not tell against her accepting the soundness of the Puppy Argument.


  • Question 5

Option B. Most (76%) got this. Abe is right since the Puppy Argument is about consuming factory-farmed products, not going full vegetarian. Dave is wrong since the Puppy argument takes exactly the supposed causal impotence of our actions into account. (With the above, you already have the answer.) Will is also right as Premise 1 of the Puppy Argument, if it is accepted, tends to be accepted because of the perceived pain caused to puppies.


  • Question 6

Option D. A large minority (44%) got this. So my decision to abstain from eating animal products will not have any impact on animal product suppliers. Under such circumstances, my not buying or consuming meat products be neither necessary nor sufficient for reducing the suffering of animals due to our consuming animal products. Not sufficient because my actions won’t do anything. Not necessary because–for all I know, someone else’s action would have made the impact without any further help from me. Many (46%) thought that it was necessary and so chose Option B.


  • Question 7

Option A. Only a plurality (33%) got it. Abe is wrong because Singer wasn’t arguing for the idea that we ought to save the poor child drowning in the pond—that was a premise of his argument. Betty is wrong because her version of the conclusion would entail that even the poor people in the famine struck countries are also obligated to donate to famine relief. Carmen is wrong because Singer wasn’t contesting the definition of a supererogatory act as one that is praiseworthy but not morally required. He is arguing that many of the things we consider supererogatory are in fact obligated. Update: Conceivably, you might even think that what Abe said is an intermediate conclusion of some part of his larger argument. But even so–just ask yourself. What’s the larger point that Singer is trying to argue for in his paper. What is “the conclusion that Singer is attempting to draw from his arguments“?–as the question statement asks. That we ought to save a child drowning in a pond? Or–as the title of the paper itself makes very clear–that people in rich countries ought to do more to help with famine relief in poor countries? From that larger perspective, to call the bit about the child the conclusion he is attempting to draw from his arguments would be very inaccurate.


  • Question 8

Option D. Most (69%) got this. Option A is wrong since the total amount of happiness involved is the same for the two courses of action (given the conditions described in the option), Milton will not be acting in a morally wrong way according to Utilitarianism. Option B is also wrong. Utilitarianism says that the morally correct act is the one that generates more overall happiness, rather than any particular person’s happiness. The information given in the option doesn’t imply that Milton will be choosing a course of action that will generate less overall happiness.


  • Question 9

Option B. Almost all (91%) got this. Option A is wrong since, even though the loss of overall happiness may be small, it is not nothing. Option B, on the other hand, expresses exactly an Utilitarian calculation.


  • Question 10

Option D. Most (85%) got this. Adson is wrong since there are three, not four kinds: Actual-Explicit, Actual-Implicit, and Hypothetical. Bernice is wrong as that’s not the relevant difference between Actual vs. Hypothetical contracts. Actual contracts are grounded in the agents’ actual consent. Hypothetical contracts are grounded in what would have been reasonable for the agents to consent to. Calso is the only one with a right answer.

Judging from the emails, I see that a number of you are confused between two things. A case where someone commits to (actually) consenting, but only if certain conditions are met. This is what Bernice was talking about–imagine her saying “I will (actually) consent—but only if the leaders are able to perform.” This is not Hypothetical Consent, which contains no actual consent at all. For Hypothetical Consent, even though the person did not actually consent, and in some cases, couldn’t have actually consented or come to consent (because dead, in a coma, etc.), consenting is what would have been reasonable for the person to do—had the option been available. See L05 Slide #36 and this. The difference is like that between saying: “He loves her as long as she loves him back” vs. “Grandma—who is dead—would have loved him”. The important thing to note is this. As long as it is true that “consenting would have been the reasonable thing to do”, then, hypothetical consent already exist–even though the person hasn’t actually consented, or maybe will never be able to consent. The first part of what Bernice said, however, implies a situation where plausibly consent doesn’t obtain when the conditions are not met.


  • Question 11

Option A. Most (79%) got this. If coercion is essential to government, then, it will follow that if there isn’t any coercion around, there won’t be any government around, which also means that if there isn’t any morally acceptable coercion around, there won’t be any morally acceptable governments; consequently, governments are morally acceptable only if at least some coercion is morally acceptable, hence, Option A. But is it the case that if at least some coercion is morally acceptable, then the government retains the monopoly over the use of force? No, hence, not Option B. Option C is also wrong—since nothing said so far has ruled out the possibility of morally acceptable coercion.


  • Question 12

Option A. Only a minority (20%) got this. This one bears a bit more explaining. If Huemer’s argument is unsound, then what this means is that either its premises are false, or it is invalid. Let’s grant that it is valid—so that leaves us with the two premises. So, for Huemer’s argument to be unsound, at least one of its premises needs to be false. If Tony’s actions are morally acceptable, the first premise would be false. But the first premise (in particular) being false is not a necessary condition for the argument to be unsound. Hence, Betty is wrong. On the other hand, if there is a morally relevant difference between coercion by private individuals and coercion by the government then the second premise would be false, and this is a sufficient condition for the argument to be unsound. Hence, Carmen is right. What Dave said is simply neither here nor there. Even if it is the case that only the modern welfare state can efficiently and effectively undertake welfare transfers from the rich to the poor, it won’t follow that either of the premises would be false. Hence, Dave is wrong.

Abe is the harder one. To see why he is right, you need to think hard about what is required for either premise of Huemer’s argument to be false. There are three possibilities: (a) both the first and second premises are false; (b) the first premise is false, but the second one is true; and (c) the first premise is true, but the second one is false. If the first premise is false, then, Tony’s actions are morally acceptable—so some coercion is morally acceptable, and this covers (a) and (b). Possibility (c) is where Tony’s actions remain morally unacceptable, but (contrary to the second premise), there is a morally relevant difference between coercion by private individuals and coercion by the government. It will follow that at least some coercion by the government is morally acceptable, hence, some coercion is morally acceptable. So, Huemer’s argument is unsound only if some coercion is morally acceptable.

(Yes, this was a harder one. But if you managed to get that Carmen is right and the other two are wrong, your options do narrow down quickly even if you couldn’t quite figure Abe out. Most students (67%) chose Option D: “Betty and Carmen”–so it’s important for you to understand why Betty’s statement is wrong.)


  • Question 13

Option C. Most (82%) got this. All Abe managed to say is that given the conditions, we can still feel morally responsible; but that doesn’t say that we are morally responsible. Betty’s statement is basically the position of Classical Compatibilism (see L06 Slide #29). Carmen’s statement is basically the position of what I called “Semi-Compatibilism” (see L06 Slide #32). By the way, it’s not that you need to know these positions by heart to solve this question. What you do need is to be able to see that Betty and Carmen have each proposed a way for us to be morally responsible in a deterministic universe.


  • Question 14

Option A. A majority (56%) got this. Abe is right in that both arguments are neutral with respect to the truth of Determinism. Indeed, the Regress Argument has a more radical conclusion. But if moral responsibility is an incoherent idea (what the Regress Argument wants to show), it will follow that we aren’t morally responsible for our actions. Betty is wrong in that if Determinism is compatible with free will, then one of the premises in the Standard Argument will be false. This is not the case for the Regress Argument. Carmen is wrong in both what she said about the Standard Argument and the Regress Argument. A good number (29%) chose Option D, probably distracted by the fact that the Regress Argument does conclude more ambitiously than the Standard Argument, missing the fact that even so, both, if sound, entail the claim stated by Abe.


  • Question 15

Option C. Only a plurality (44%) got this. If Determinism is required for moral responsibility, then Dave can’t also agree with what Abe said. It follows from what Dave believes, that Determinism rules out free will. This means if he thinks that we are morally responsible for our actions, then he also thinks that free will is not needed for moral responsibility, hence, Betty is right. If Dave believes that we have free will, it will also follow, given his other beliefs, that the universe is indeterministic. But since he also thinks that Determinism is required for moral responsibility, he should also think that no one is morally responsible for what they do, hence Carmen is also right. The wrong answers were more or less evenly split between the remaining three Options.


  • Question 16

Option C. A majority (58%) got this. Basically, Dave accepts PSR(a) but not PSR(b). This means that he can’t accept the Leibniz/Clarke version of the Cosmological Argument. But he can accept the Kalam version, since with PSR(a), he can derive the weaker principle that there must be a causal explanation for the fact of a thing’s coming into being. See also the explanation for Quiz 6B.


  • Question 17

Option B. Most (82%) got this. Abe’s argument has the right form for committing the fallacy of composition. Betty’s statement, on the other hand, does not. In fact, it doesn’t even present an argument, as much as it just explains what the prof meant by his words.


  • Question 18

Option C. Only a plurality (40%) got this. Both arguments are meant to be arguments in natural (a)theology and so aim to appeal to facts accessible to all—whether theists, agnostics, or atheists, believers who just go on faith even against reason, and believers who go on reason. Abe and Carmen are wrong because they rule out part of the audience that was meant to be covered. See L08 Slides #14-16. Many (37%) were distracted by Option B.


  • Question 19

Option D. A good plurality (41%) got this. This one wasn’t meant to be hard, but many seemed to have been puzzled by it during the exam. Some of you asked when this conversation took place. Well, given the statements, it has to take place after the semester is over–otherwise, the statements don’t make sense. Did all of the funding for Damien’s holiday come from what he stole? No–you can’t deduce that from “Damien had been stealing the membership fees to fund his holidays”–which is compatible with him chipping in some of his own cash. So there is no inconsistency between the four statements or subset thereof. For the case where all four statements are taken on board, as long as you realize that the shortfall of 5 gold pieces could have come from anywhere. Another student wrote in to say that maybe the Gold Pieces are only specific to the Philosophy Interest Cult. But why would you assume that? Since there’s no information either way, it’s important to draw your conclusion on basis of that ignorance. And in any case, those GPs can’t be restricted to the Cult—the Cult had “income” in GPs, which meant that they have to get them from somewhere. And someone is paying for goods and services provided by the outside in GPs too. Yes, these GPs are specific to the Philosophy Interest Cult storyline; but that’s not the same as saying that they are specific to only the Cult within that storyline.

Update: One student wrote to say that he and a few others changed the answer from D to C after I reminded everyone not to overthink the question. Sigh! My apologies then. The reason why I said that you shouldn’t overthink the question was because we received a lot of queries premised upon rather unnatural (even outrageous) readings of the question (“Are these all the gold pieces in the world?”). What I meant was that you should take the natural meanings of the statements. If someone told you that so and so “embezzled money from his company to fund his holidays”, would you assume that all of the money for his holidays come from the embezzlement? You shouldn’t. In one sense, this was a simple math question plus just one additional thing in the background. Ok, so the numbers don’t add up. But is there enough information for me to infer that the numbers must add up exactly such that if they don’t, there has to be an inconsistency? Is there enough information for me to conclude that these are all the numbers, that there are no unstated numbers? No there isn’t, therefore, I can’t conclude that there is an inconsistency. Or consider this–very real life example. There were exactly 430 students in the exam hall at our finals. After we went around to collect your forms, we counted 429. After recounting, we still have 429. Is there an inconsistency in the story as told so far? (As it turned out, the missing form was found on a student’s table…)


  • Question 20

Option D. Most (66%) got this. Adson is wrong since, as explained in detail in the lecture, Mackie is careful to say that two other quasi-logical rules are needed for the contradiction to be generated. Bernice and Calso are wrong because rejecting the existence of evil, or denying the Principle of Non-Contradiction, are each not an only possible solution even though, conceivably, each might be a possible solution.


  • Question 21

Option C. A majority (50%) got this. If Adson and Bernice are convinced by Calso’s argument, then they have a good reason to think that a Self-Existent being exists. But this doesn’t imply that Eugen’s argument is unsound. On the other hand, if they are convinced by Eugen’s argument (i.e., Mackie’s LPOE Argument), then they have a good reason to think that an Omnipotent and Omnibenevolent being doesn’t exist (since evil exists); but this doesn’t imply that Calso’s argument is unsound. Hence, Option A is wrong. For similar reasons, since it is possible for both arguments to be sound and both conclusions true, Option B is also wrong. Finally, if Adson and Bernice are convinced by Eugen’s argument, then, accordingly, they have a good reason to think that an Omnipotent and Omnibenevolent being doesn’t exist (since evil exists). And they do believe that evil exist—it was Damien’s evil doings that got them into the discussion in the first place (see Question 20). Hence, if they believe that God–which is supposed to be an Omnipotent and Omnibenevolent being–exists, they have to do so as a matter of faith against reason.

Update: A couple of students wrote in to say that they didn’t think that they were supposed to take the information from Question 20 on board. This is fair enough. But on further consideration, it turns out that it doesn’t matter. Option C has to stay. This is because stated fully, Mackie’s LPOE was meant to argue against believing that the Classical Theistic God exists (see the opening paragraph of the reading; see also L08 Slide #16). This means that if Adson and Bernice were convinced by that argument–i.e., they agree with Eugen that it is sound–then, they are admitting that they have a very good reason to believe that the Classical Theistic God doesn’t exist. And as I have reminded everyone at the exams itself, you are supposed to take it for granted that “God” always refers to the Classical Theistic God–something also mentioned in the lectures. Once you have this set up, only Option C remains as the correct answer. I’ve also edited the explanation to make this point more clearly.


  • Question 22

Option D. Almost all (91%) got this. Option A is wrong as it is a case of a (possibly) justified but untrue belief, so, no knowledge. Option B is wrong since from the fact that something is false, you can’t infer that someone would be unjustified to believe it. Option C is wrong as it is the case of a true but unjustified belief, so, no knowledge. Option D is right because truth and justification aren’t the same thing and they can come apart.


  • Question 23

Option D. Only a minority (11%) got this. Option C (“If being firmly convinced of something, by itself, counts as being justified in one’s belief, then everyone can be a teacher, which is absurd”) is wrong despite being apparently correct. That is, the statement in Option C looks like something that Zhuangzi said, but is in fact not the same. Recall this from the reading:

#6 If a made-up mind counts as a teacher, then who doesn’t have a teacher? Why should it just be the self-chosen experts on the order of things who have them? Stupid people would have them, too. But to have right and wrong before you’ve made up your mind—that’s like leaving for Yue today and getting there yesterday! (Zhuangzi, Ch. 2)

As explained in both the reading and the lecture (L09 Slide #20), “S learned P from a teacher” = “S is justified to believe that P” (this being a background cultural assumption). In other words, the point that Zhuangzi was making isn’t–if we accept that anyone who is firmly convinced of something is thereby justified in his belief, then he (literally) has a teacher. (In fact, even if you take him literally, you still can’t conclude that everyone (literally) can be a teacher. From “everyone has a mother”, you can’t conclude “everyone is a mother”.) What he is saying is–if we accept that anyone who is firmly convinced of something is thereby justified in his belief, then anyone (no matter wise or foolish) would be equally justified in their beliefs as long as they are firmly convinced. (This is what happens when you are in a haste to respond to students, a whole lot of them at that, and doing it for 1.5hrs straight: you get confused…) The majority of students (65%) were distracted by Option C; but after careful consideration, this one can’t change.


  • Question 24

Option B. Most (69%) got it. But there’s a complication that bears explaining. Option A, C and D wouldn’t imply that Zhuangzi’s argument is unsound, since these are either not proper cases of disputation, or cases where the disputants exited disputation and so are not longer covered by Zhuangzi’s argument. Option B is the best answer as it is basically the nutshell of “Response 5” (see L09 Slide #33-35).

But after the exam, one especially perceptive student pointed out to me that, as I explained in the lecture (see Slides #36-37), the response doesn’t show that the Argument from Disputation is unsound, as much as it allows us to tame the conclusion. In the fluster of the moment (dealing with aftermath of a 434-person exam and all that), I was garbled in my response and the complication completely slipped my mind. This is what the option says:

It is at least possible that, under suitable conditions, different and equally viable standard of assessment may end up justifying opposing views in a disputation.

To be strict about it, this is only the second part of Response 5–the part corresponding to “Response (5.2)”, or “Permissivism” (specifically, Slide #35). It doesn’t cover the first part, which correspond to “Response (5.1)”, or “Retrenchment”, which introduces a relative, rather than absolute conception of justification. If Retrenchment is taken on board, then it is indeed the case that the overall response will not show that the Argument from Disputation is unsound. But if we only have Permissivism on its own, then in principle, we are still dealing only with an absolute conception of justification. This means that if we do accept the statement in Option B as true, we can show that the Argument from Disputation is unsound. Specifically, the part of Premise 1 saying that if the disputation is interminable, then neither disputant is justified is rendered false. (Why then introduce Retrenchment in the lecture? Mainly because on its own, Permissivism is more controversial than Retrenchment plus Permissivism.) (Thanks to A/P Tang Weng Hong  for reminding me on our way back to the office.)


  • Question 25

Option A. A majority (65%) got it. Bern is wrong since even full-fledged physicalists can think that some things have minds—though the fact that something has a mind is fully reducible to the fact that it has a bunch of physical characteristics. Gene is wrong since nothing in full-fledged physicalism says that our mental characteristics (fully reducible to physical characteristics) are determined fully by what we eat or drink. Tess is the only one who got it right.


  • Question 26

Option C. A majority (64%) got it. Bern is wrong since Nagel does not come out firmly for dualism. Tess is wrong because for Nagel, even if brain does taste like chocolate, we still shouldn’t believe that the experience of tasting chocolate is nothing but a complicated physical event in our brain, as physicalism insists. Only Gene is right.


  • Question 27

Option D. A majority (63%) got it. The scenario of the show implies that the presence of consciousness in another can be conclusively detected. It presents no proof on whether any of the three theories are true.


  • Question 28

Option A. Most (83%) got it. From the information given, they should assign an equal probability to Tess being at each of the three locations. So, it follows that Jane is right and what she said a proper application of the principle of indifference. Neither Bern nor Gene are applying the principle at all. (Someone asked if we should assume that Tess is physically chatting with Lena, as opposed to doing that, e.g., over her phone, possibly in Greenhouse #5 or the Machine Shed. But Phil did say three distinct locations are involved…) Incidentally, do read what Bern said carefully–her point is that Tess should be indifferent between two course of action, i.e., she shouldn’t prefer doing one to doing the other. This is not an application of the POI, which is about what you should believe.


  • Question 29 (explanation expanded)

Option C. A majority (54%) got it. Carmen and Abe are the easy one–the first is correct while the second is wrong. See the final paragraph of the Bostrom reading (both points were also discussed in the lecture). The important thing is that these are the two clear points. This means that any option that excludes Carmen, or that includes Abe, has to be wrong. And there is only one option that satisfies both conditions—Option C.

What about Betty then? That’s the tricky one. Now, given the math in Bostrom’s argument, it is unlikely for all three propositions to be true at the same time. But we can also deduce this answer more informally. Suppose:

(1) The fraction of all human-level technological civilizations that become Posthuman Civilizations ≈ 0.

(2) The fraction of Posthuman Civilizations that are interested in running ancestor-simulations (or that contain at least some individuals who are interested in that and have sufficient resources to run a significant number of such simulations) ≈ 0.

It will follow from the above that:

The faction of all human-level technological civilizations that will go on to create ancestor simulations ≈ 0.

This suggests that the odds are heavily against there being enough ancestor simulations around such that:

(3) The fraction of all observers with human-type experiences that are, in fact, Sims ≈ 1.

But here’s the catch: the reasoning doesn’t quite show that it is impossible for all three of the argument’s main propositions to be simultaneously true. After all, there is still a very slim chance that the very few human-level technological civilizations that do become Posthuman Civilizations and are interested to create ancestor simulations and are thus doing just that ended up creating such a massive lot of them that, well, (3) turns out true a. It’s extremely unlikely, but not impossible. (Bonus: Can a similar reasoning show that Abe is correct? No. That’s because of the way Bostrom defined a Post-Human Civilization as one that can run massively many ancestor-simulations. So, if both (1) and (2) are false, it will follow that there are massively many ancestor simulations for each Post-human civilization out there such that (3) has to be true.) Hence, Betty is also right. That is, her claim is at least compatible with the soundness of Bostrom’s argument.

Now, yes, I recall asking everyone to consider whether it’s possible for all three of Bostrom’s propositions to be true. (In fact, I even used to think that they can’t all be true—before I redid the math myself.) Also, I can see how the right judgment about Betty’s statement will come more naturally to you if you think in terms of the math, rather than purely informally. But you can get the answer even without the math. Nevertheless, whatever might be the conclusion about what Betty said, it’s definitely more speculative when compared to what Abe and Carmen said. Conversely and more importantly if you managed to see that what Carmen said is certainly correct, you can already rule out options B and D, leaving only A and C. And if you also see that Abe is wrong, you are left with only Option C.

The most common (42%) wrong answer was Option D, which is a non-starter since picking this option implies that you are taking Carmen to be saying something incompatible with the soundness of Bostrom’s argument. More generally, To give a mark to anyone who picked Option A, B or D is to unfairly reward a failure to take into account clear statements from the reading and the lecture, which would be unfair to the majority of students who did chose the right answer.


  • Question 30

Option D. Most (72%) got this. Abe is wrong. For all we know, all those resources may come to naught. Betty is wrong since the scenario is just as compatible with Helena having serious issues, rather than her being the avatar of a simulator living her virtual life among us. Carmen is also wrong since only Artificial Intelligence is needed to make sense of her scenario. We don’t know if the car has Artificial Consciousness.