Wednesday, January 30, 2019

The Perverted Faculty Argument: A reply to Edward Feser.

The Perverted Faculty Argument: A reply to Edward Feser.

1. Introduction.

Edward Feser defends a perverted faculty argument [1] against contraception, masturbation, same-sex sex, and bestiality. In this essay, I will argue that Feser’s argument is unsound, and that his argumentation fails to provide support for his moral claims.

In the next section, I will make some general comments on the type of argument he makes.

In the third section, I will argue against the first and sixth premises of the perverted faculty argument.

In the fourth section, I will argue that some of the main tenets of the Aristotelian-Thomistic metaphysics on which Feser bases the argument are very probably false.

In the fifth section, I will address some of Feser’s claims about the natural end of our sexual faculties.

2. Individually probable premises and improbable conclusions.

Suppose someone makes an argument for an anti-natalist view. So, they come up with premises Q1,…,Qn, which together entail that deliberately having children is always immoral for any human being. In support of the premises, they give reasons intended to persuade that they are true – in other words, they argue for the premises, not in the sense of giving syllogisms that have them as conclusions, but in the sense of ‘arguing a case’.

As it turns out, each of the premises, when individually considered, appears more probable than not from my epistemic perspective, but not certain. Should I accept an argument like that, and conclude that the conclusion is true, or probably true?

It seems not. Even if each of the premises, individually considered, is more probable than not, that does not imply that their conjunction is. This is so even if the number of premises is just two. But moreover, when assessing the probability of a conclusion, I also should factor in other sources of information if I have them. It turns out that another source of information is precisely my own sense of right and wrong: when I contemplate different hypothetical scenarios involving humans deliberately having children, I can find plenty in which my sense of right and wrong says that the behavior is not immoral. The fact that the immorality of all of those behaviors is implied by premises that individually considered appear more probable than not but not extremely probable, is not good enough evidence to conclude that my sense of right and wrong is failing, even if they would weaken my confidence to some degree or another. So, it is not the case that I should accept the argument. In fact, I should not.

Let us now turn to Feser’s argument. I do not find each of the premises probable – quite the opposite for at least two of them, a matter that will become clear later on -, and some of the very acts that Feser’s argument condemns are clearly not immoral in all metaphysically possible situations, by my own lights. Purely for example, in the majority of real life situations in which humans masturbate, it is obvious to me that it is not immoral to masturbate.

Still, there are people who do not find those assessments so clear, and/or are inclined to assign a higher probability to Feser’s premises. So, I would give counterarguments with counterexamples not involving any of the sexual behaviors Feser’s argument intends to condemn. However, it is important to keep in mind that even an argument based on premises that individually appear probable is not enough to establish a conclusion, and moreover, that it is sometimes rational to reject an argument on the basis of a direct assessment that the conclusion is false, for example if one has a generally reliable source of information that yields the verdict that it is false, and no sufficient evidence that the source in question is failing in that particular case, and this is so even if one properly reckons that every individual premise is probably true.

3. The perverted faculty argument and common sense.

In this section, I will first argue that the sixth premise of Feser’s argument is false, and in fact it represents a radical rejection of common sense assessments of rationality. Then, I will argue that the first premise is also false ordepending on how it is interpreted – probably false and additionally, irrelevant to the matters under consideration.

3.1. The sixth premise.

The sixth premise in Feser’s perverted faculty argument states:


6. But it can be rational to engage in an act only if it is in some way good for us and never when it frustrates the realization of the good.

One first difficulty is information. More precisely, is that supposed to be the case under conditions of full information only?

Feser says earlier that “what is in fact good is the realization of the various ends inherent in human nature; and thus a rational and correctly informed person will perceive this and, accordingly, direct his actions towards the realization or fulfillment of those ends”

However, the conclusion of his argument does not seem to make exceptions, and neither does this premise. Perhaps, Feser meant to make them? But if the sixth premise is only meant for cases of full information, that alone would make the premise – and thus, the argument - inapplicable to many real-world situations, independently of other considerations.

In any event, and for the sake of thoroughness, I will give examples in which human beings behave rationally even when what they do is in no way good for them, due to limited or false information. Moreover, sometimes it would be irrational on their part not to act in ways that are very bad for them. I will later argueon different grounds – that even in cases in which there plenty of relevant information – even as much as one finds in nearly all real life cases -, the sixth premise has absurd implications. So, let us begin with the first scenario.

S0: Jack has been diagnosed with cancer, and radiation therapy plus chemotherapy is the treatment his oncologist, Joe, recommends. According to Joe, with that treatment, there is a significant chance of recovery, even though there are significant negative side effects and recovery is not certain, whereas without the treatment, the cancer will almost certainly be fatal.

Jack requests a second opinion, and there is agreement. Jack wants to get better, to stay alive, go on with his life, stay with his family, etc. For those reasons, he accepts and gets the prescribed treatment. Unfortunately, a very improbable mix-up happened, and the results belonged to another patient. Jack does not have cancer, and the radiation plus chemotherapy only hurts him, damaging the proper function of several of his organs.

So, Jack behaves rationally, even though he does something that is bad for him, and in no way good. It might be argued that it is in some way good because he feels happier for a short while because he believes that he’s going to get better. But if that is enough to meet the criterion of the first part of the premise, no problem: let us add an example in which not even that threshold is crossed.

S1: Every day, Jill drives from her home to work and back. There is no public transportation, and the distance is too much for walking: it would take her hours; she would not be home in time for dinner; her family would miss her, etc. A bicycle would be risky on some of those roads.

One day, when she’s at work, Sergei stealthily plants a bomb in her car. He does so because he is a professional assassin, and he believes that the car belongs to Olga – who has an almost identical car. Sergei has been paid good money to assassinate Olga, and he will get more when the job is successfully completed.

Jill has no reason – like nearly everyone else on the planet who has a car - to check her car for bombs – not that she would find it except with a very thorough inspection. So, she turns the ignition key. The bomb goes off, killing her.

Surely, turning the ignition key was very bad and in no way good for Jill, but she was rational in doing so. We might modify the scenario so that her family is in the car too, and her actions would remain rational – though it was even worse for her -, or construct other, worse scenarios, e. g., she survives for a few minutes, burned and bleeding, etc.

If one has a moral objection to using one’s own car every day to go to work and back (buy why, given the specific circumstances of Jill’s life?), we can change the scenario and stipulate she rides a bicycle home and there is no danger in most days, but one day a terrorist planted a bomb in the bicycle path, etc.

The upshot: it is apparent that there are metaphysically possible scenarios – and even actual scenarios – in which a human being rationally act in a way that is very bad and in no way good for her.

In addition, there are metaphysically possible – and even actual – scenarios in which it would be irrational for a human being to fail to do something that is very bad for her, and in no way good for her. For example, in the scenario above, arguably it would not have been rational on Jill’s part to fail to turn the key. Why shouldn’t she do it? Perhaps, she wanted to take a walk first? If so, we may further stipulate that Jill had promised her children she would be home early to watch a movie together, and taking a walk – as far as she knew – would only make her be there late. But she values keeping her promises to her children, and she reckons – epistemically rationally, given the information available to her – that the only realistic way not to be late is to drive home without delay. We may add more conditions if needed, but the point is that in some surely metaphysically possible situations, the only rational course or courses of action are not at all good and very bad for the agent who makes the choice as to how to act – and surely far worse than alternative, irrational courses of action would be -, due to limited information.

So, let us assume that cases of limited information like the ones above do not count, so they cannot falsify the premise – that would be too easy. But then, a relevant question is: What are the information requirements in this premise? Is it full information? We may interpret full information not in an absolute sense, but in the sense of all of the information required to decide whether an action is good or bad for A, to what degree, etc. Yet, even then, it is not the case that there is full information in all real life cases. In fact, at least there isn’t full information in plenty of real life cases, including many of those involving the frustration of the functions or ends of human faculties. Maybe the correct interpretation of Feser’s words is something else? But what might it be?

Regardless, I will assume from now on and for the rest of this subsection that Feser has a way out of the information objection, and I will consider other kinds of scenarios, focusing on the second part of the premise – namely, about the frustration of the realization of the good.

Note that the word “cannot” in the premise is about metaphysical possibility – as clearly indicated by the rest of the premises and the argumentation given by Feser in support of them -, so the premise entails something like:

6.2. It is metaphysically impossible for it to be rational for a human being to engage in an act when the act frustrates the realization of the good.

However, this is absurd, information aside. To see why, we begin with the examples of smoking and breastfeeding:


To be sure, smoking to excess clearly does frustrate the natural end of breathing, and refraining altogether from breastfeeding one’s children arguably frustrates the natural end of lactation, especially if we factor in the bonding between mother and child that is facilitated by nursing. But then, precisely for these reasons, people are inclined to raise at least a mild moral objection to smoking to excess, and even gently to recommend that it is, all things considered, better for mothers to breastfeed their children. In this way, common sense clearly tracks the “old” natural law theory’s insistence that there is a connection between what is good for us and what is consistent with the realization of the ends nature has set for us.

The reason many people seem to raise mild moral objections to smoking to excess is that there are predictable negative consequences for family members or coworkers or other third parties who are exposed to the smoke, or at most – though most controversially, and this probably varies widely with the person raising the mild moral objection in questiondue to the negative health consequences for the smoker. Moreover, the moral objections that people tend to raise would be immediately withdrawn – except, perhaps, for some people rejecting common sense – in some of the metaphysically possible scenarios in which the actions of smoking enough to frustrate the natural end of breathing or the complete refraining of breastfeeding are carried out in order to prevent something much worse. On that note, let us consider the following scenario:

S2: Jorge lives in a dictatorship, and is a political prisoner. He lives in an infamous prison, ruled with an iron fist by Raúl, the prison commander. Jorge’s wife and children live in somewhat better conditions in a nearby building, but also within the control of Raúl, who has proven particularly vicious, and tends to brutally punish any prisoner who fails to follow his commands – even on trivial matters. In the case of prisoners who have their families held there as well, the punishment often involves having the prisoner’s spouse and/or children beaten up badly, or raped, or both.

When Raúl has an inmate’s family tortured, he always forces that inmate and at least several others to watch, so after six months of imprisonment, Jorge – like all prisoners who have been there for about month or more knows about Raúl’s abject behavior.

As it turns out, Raúl is also a very heavy smoker, and takes himself to be an expert in history – though he has a very twisted interpretation of many historical events. Given that Jorge is a history professor, Raúl takes an interest, so he has Jorge brought to him, gives him a cigarette – Raúl is already smoking -, and tells Jorge – making it clear he won’t take ‘no’ for an answer - to smoke with him, and discuss history. Jorge has never smoked before, and knows it is bad for his health and even believes that it frustrates the natural end of breathing, but given the information available to him, there is a very credible and terrible threat to his family, and smoking – and talking history with Raúl, lying of course and being very careful not to anger him – is clearly the best available option by far to protect them.

In the end, Raúl ends up doing that at least three days a week, and he and Jorge smoke for about two hours a day, on average. Fearing for his family, Jorge never declines Raúl’s ‘invitation’ to smoke, though as the years go by, Jorge realizes that his breathing function is not only being frustrated, but is degrading over time as a result of the smoking.

Surely, Jorge did nothing irrational, even though smoking like that frustrates the end of breathing. The conclusion that he was even slightly irrational in his acts of smoking would fly on the face of common sense. So, it is clearly metaphysically possible for human beings to smoke to excess – frustrating the natural end of breathing, as Feser says[A] -, and behave rationally in doing so. Here is another scenario:

S3: A rogue AI with superhuman intelligence has taken over the world[B], and humans live under its rules. The AI enforces its rules publicly and punishes violations.

The AI – by means of the trillions of robots of different sizes it controls - provides sufficient food and water for all humans, and also makes sure that the human population does not fall below 10 millions. To that effect, it uses cloning, artificial insemination, artificial wombs, etc.[C] Also, nearly every living human has at least one living sibling, at least until they reach they age of 70 – an age most humans do reach, in relatively good health -; this is also the result of the AI’s actions.

Humans are allowed to choose their own names freely when they are old enough to make that choice, but they are also given an identification number by the AI. The numbers are sequential – 1, 2, 3, etc.

For the most part, the AI does not kill, torture – beyond its very oppressive rules, that is - or otherwise inflict pain on humans. In fact, it even provides for nearly all of them very advanced medical attention if and when they need it. But male humans with prime IDs are not given medical attention after they reach the age of 40, and are also subject to the following weird rule: each of them must frustrate his own breathing function by means of smoking at least 50 cigarettes a day (it’s a rogue AI, and it ended up with a weird value function because of some programmer’s mistake).

If one of them refuses to smoke the daily 50 cigarettes, then the following 200 adult humans (in ID order) are tortured to death in 20 different and horrific ways: 10 of them are fed to starving wolves, 10 are doused with oil and slowly burned to death, etc.

John is a 40 years old human male, and lives under the AI’s rule. John’s ID number is 47055833459, so the prime number rule described above applies to him. In order to save 200 people – including some of his friends and his brother - from a horrendous fate, he smokes 50 cigarettes every day, frustrating the end of breathing, even though he is aware of the effects of smoking on human health – as most people are, because the AI is very clear about that, and it is known to be truthful: it does not lie, and provides good evidence of its claims when asked, as well as good information about nearly any subject that humans might ask about, with only a few exceptions – where it refuses to answer, rather than lie.

John smokes 50 cigarettes a day because he values keeping his brother, some of his friends and the rest of those people safe from the horrific fate that awaits them if he does not smoke the 50 cigarettes more than he values his not suffering the negative consequences of smoking like that, and surely more than he values the not frustration of one of the functions of one of his organs or systems.

The scenario is obviously not realistic, but it is metaphysically possible[B][C]. It should be clear that John’s behavior is rational. This also shows that Premise 6 is false, and even that 6.2 is false, assuming of course that any instance of frustrating a natural end of one’s faculties is in fact an instance of frustrating the realization of the good. But if it is not the case that any instance of frustrating a natural end of one’s faculties is an instance of frustration of the realization of the good in the sense of Premise 6, then Feser’s argument seems to be a non-starter, since the premise could not be applied to reach the intended conclusion.

Incidentally, Feser’s argument intends not only to show that some behaviors are irrational, but also that they are immoral. While that is not stated in Premise 6, it does follow from the view Feser’s defends that behaviors involving the frustration of natural human ends are immoral. But that also flies on the face of moral common sense. John’s behavior is not immoral, regardless of whether he behaved immorally before by having children in a world like that – though one might get around that as well by stipulating, for instance, that the AI made the children by artificial means like cloning or artificial insemination + artificial wombs, or forcibly impregnated John’s wife Penny using John’s sperm, or by some other means.

Granted, all of the previous cases are unrealistic, but again, Feser’s sixth premise involves a claim about metaphysical necessity, so unrealistic yet metaphysically possible cases can be used to test it and falsify itthough I will consider realistic examples later.

The other case identified by Feser – namely, breastfeeding – also can be used to show that the premise under consideration is false, in an essentially similar manner.

S4: The general situation is as in Scenario S3. Here’s another AI rule: Each human female with an ID number that is a prime power of 3 must have at least 3 children if she can the AI will use

forced insemination in case of refusal -, and must deliberately frustrate the function or natural end of breastfeeding by refraining from doing so altogether. She is allowed to feed them with high-quality formula, and otherwise allowed to raise her children, bond with them, etc., as long as the rules are kept of course. But she must refrain from breastfeeding.

If one of the females in that situation breastfeeds one or more of her children, then the following 1000 adult humans (in ID order) are tortured to death in 20 different and horrific ways: 50 of them are fed to starving wolves, 50 are doused with oil and slowly burned to death, etc. Also, one of her siblings is tortured to death in one of those manners, if she has a living sibling – which nearly everyone has before they are 70.

Mary’s ID number is 94143178827, so the rule applies to her, and she chooses to refrain from breastfeeding altogether. She does care for her children a lot, and makes a huge effort to raise them well in that horrible world. But she does not breastfeed them, for the sake of her sister Jane and the other 1000 people who would otherwise face torture to death.

As in the smoking example, it should be clear that Mary behaves neither irrationally nor immorally. In particular, the rationality of her behavior in this metaphysically possible scenario falsifies Premise 6.

Of course, in addition to smoking or a total lack of breastfeeding, one can find other cases in which the function or natural end of another faculty is frustrated, but the behavior is not irrational.

It is true that in the previous examples, people are acting under terrible threats. But they are still making choices, and rational ones. Indeed, their ability to choose has not been taken away from them entirely, even the range of available alternatives is hugely constrained. But there are no exceptions for threats in Premise 6. It could be argued that Feser meant to exclude cases involving threats, but that seems unlikely: there is nothing in the paper suggesting that. Still, let us consider a weaker variant, excluding threats.

6.3. It is metaphysically impossible for it to be rational for a human being to engage in an act when the act frustrates the realization of the good and the human being in question is not facing a significant threat.

It turns out that 6.3 is enough to show that the premise is false, and in fact absurd. So, for our next example – or rather, our next family of similar examples - , let us consider something Feser says about eyeglasses and binoculars.


Nor does the premise entail that to use man-made devices is per se to frustrate the natural end of F. On the contrary, man-made devices can sometimes restore natural function (as with eyeglasses) or enhance it (as with binoculars).

Fair enough, in those cases, there is no frustration. However, we can consider cases in which there is frustration, and even deliberate frustration of a function - or natural end of a faculty, or whatever one calls it[A]. For example, a number of experiments using distortion goggles have been done, in order to study different aspects of our visual function, and how much it adapts – if at all - to some specific abnormal conditions that make it malfunction.

It might be objected that the eyes have a general purpose of seeing, which is allegedly not frustrated. On that note, Feser makes the following distinction:

Feser: A third point to keep in mind is that there are crucial differences between, on the one hand, an individual deliberate act of using a bodily faculty and, on the other, an ongoing and involuntary physiological process. Use of the sexual organs is an example of the former whereas hair growth, breathing, perspiring, and lactating are examples of the latter. Now the former has a specific end-state or climax, while the latter do not.

The use of the eyes for seeing might be more similar in this regard, at least most of the time, to hair growth, breathing, etc., than to the use of the sexual organs in the way under discussion – though this too is debatable; the visual function seems to be somewhere in between, again most of the time.

However, these experiments do not fall into the ‘at least most of the time’ category. In fact, in these experiments, the eyes are deliberately used in a way that frustrates the visual function in very specific manners, in order to study how and to what extent the visual function can adapt, what other consequences there are, etc.

Granted, the distortion goggles are not actually making the eyes malfunction, but only distorting the images by getting in the way of the light that goes through the eye. However, the visual function is being frustrated and deliberately made to malfunction, even if there is no organ damage. For that matter, using a condom during sex impairs the reproductive function without actually causing any sort of organ damage, at least in nearly all actual cases in which condoms are used.

So, this is a decisive counterexample. However, I suspect that a defender of the perverted faculties argument might insist that there is no frustration in the case of distortion goggles, for some reason, perhaps insisting on Feser’s distinction between individual deliberate acts, and ongoing, involuntary physiological processes. I already argued that this reply fails and this example does involve a deliberate individual act, but leaving this example aside, let us consider the case of the process of eating, which also resembles the case of the use of the sexual organs in the relevant senses:

Feser: By contrast, the process that begins with arousal and ends with ejaculation within the vagina is episodic rather than ongoing, and its outcome, which is a specific event, is frustrated by contraception, masturbation, and the like.

Similarly, there is a process that begins with hunger and ends with a person swallowing the food. Just as the process described by Feser involves deliberately placing a penis in a vagina and taking actions that result in greater arousal until the male has an orgasm and ejaculates in the female’s vagina – the sexual act doesn’t have to end there of course, but Feser stops there in his argumentation -, the eating process we are considering involves placing food in the mouth, chewing it if needed and as much as needed, and then swallowing the bolus, which is the specific outcome.

Granted, in reality the deliberate actions are usually repeated, and the person usually swallows more than one bolus. However, that is not a relevant dissimilarity, not only because the male sometimes also ejaculates more than once, or because sometimes there is only one bolus – the availability of food might be limited; the person might be eating a moth or another small thing, etc. -, but more importantly and directly, because the fact that the process or processes involve in many – most cases – repeated deliberate acts does not make them any less deliberate, and repeated acts of swallowing the bolus are also specific events – or an event, depending on how one counts.

So, how can we frustrate the specific outcome?

The first way would be to provoke vomit immediately afterwards. Surely, there are metaphysically possible situations in which that is not immoral or irrational. In fact, in the context of medical experiments – among others -, no significant threat is required for it not to be irrational or immoral.

It might be objected that, perhaps, that does not frustrate the outcome, since it does not prevent food from getting into the stomach – it just removes it immediately afterwards. But now consider emergency contraception: Emergency contraception does not frustrate the outcome of depositing semen in the vagina. It frustrates the involuntary process of fertilization that begins afterwards. Yet, Feser says that contraception frustrates its outcome. So, it appears that frustrating the involuntary process of fertilization that begins immediately after ejaculation in the vagina counts as a frustration in the relevant sense, in the context of the perverted faculty argument. But similarly, then, frustrating the process of digestion in the stomach that begins after the person swallows the food is also a case of frustrating the eating process described above.

When we consider the entire relevant sets of events side by side, the parallel is striking: In the sexual case, a female and a male deliberately take ordinary actions resulting in his ejaculation in her vagina - which puts semen in her vagina -, and afterwards, she frustrates the fertilization process by taking emergency contraception. In the eating case, a person deliberate takes ordinary actions resulting in his swallowing food – which puts food in his stomach - , and immediately after that, he frustrates the digestion process by provoking vomit.

Another way to frustrate the process would be not to provoke vomit immediately after eating, but rather, taking a drug before eating that will provoke vomit immediately after the person eats, or a few minutes later. In other words, in this case, the vomiting happens just the same, but the person takes action before eating, not immediately after that. In this case, the parallel would be with taking non-emergency oral contraceptives.

Leaving aside the previous family of counterexamples, we can find new ones by considering Feser’s point about damaging organs.

Feser: And if someone did so mutilate the ears or nose that their function was impaired, this would not be a counterexample to the perverted faculty argument but rather exactly the sort of thing the “old” natural law theory would condemn.

Actually, that is a counterexample to the “old” natural law theory, as well as to Premise 6. Obviously, there are metaphysically possible situations in which mutilating one’s own ears and nose so that their function is impaired would be irrational and/or immoral. But also obviously – barring rejection of common sense-, there are metaphysically possible situations in which such behaviors would be neither irrational nor immoral. Purely for example, one may construct scenarios along the lines of Scenario S2, or Scenario S3, etc. Alternatively, we can consider the case of donating a healthy kidney or a lung. That is not in all actual instances immoral or irrational – not even to a small extent. Yet, surely by removing a healthy kidney, or a healthy lung, etc., at least one of the human faculties is impaired. Granted, the goal is not to impair it, but for that matter, the goal or goals of masturbation or same-sex sex generally is not to impair or frustrate any human end or function.

That said, these latest examples do involve significant threats – in a broad sense of the word -, either to the person making the choice, or – in ordinary cases of organ donation – to some other person. So, if threats are not allowed in the interpretation of this premise, these particular examples fail to falsify it. To avoid that problem, we may consider the case of appendectomies that are not for the purpose of a donation. For example, doctors who spend the winter at Australian bases in Antarctica have their appendix removed.[2] Here, no one is at risk. There might be a risk later, if they do go to Antarctica, but they are not forced to go there. Does that particular surgery impair a function? It is not certain, but there is a significant chance that – at least - the immune function is impaired. [3]

To close this section, it is interesting to note that Feser says that “Sometimes one good can be sacrificed for the sake of a higher good, as when one sacrifices marriage and family for the sake of the priesthood or religious life.”

While I think he is mistaken about that particular example – in my assessment, priesthood or religious life would generally if not always be a case of irrational behavior, if the person is sincere about it -, Feser is clearly correct that one can sometimes rationally sacrifice a good for the sake of a higher good. However, he does not make that exception for cases in which a rational agent uses one of her faculties in a way that frustrates its function or natural end. On the contrary, he argues against there being any exceptions. This is a mistake. As some of the previous examples – and many others one can construct – show, some of the metaphysically possible and even some of the actual cases in which it is rational for a human being to sacrifice a good to obtain a greater good – or to bring about an evil to prevent a greater evil – are cases in which a human being rationally acts in a way that uses a human faculty against its function – or natural end, or whatever one calls it.[A]

3.2. The first premise.

The first premise in Feser’s perverted faculty argument states:


1. Where some faculty F is natural to a rational agent A and by nature exists for the sake of some end E (and exists in A precisely so that A might pursue E), then it is metaphysically impossible for it to be good for A to use F in a manner contrary to E.

Granting for now and for the sake of the argument that faculties exist “so that” ends are pursued, one question is how to interpret the expression “good for A” in this premise. Does it mean ‘overall good’ for A? Does it mean ‘good in every respect, with no negative side effect’? Does it mean something else?

Feser does not say, but when responding to objections, he states:


A genuine counterexample to the perverted faculty argument’s key premise would have to involve an action that both involved the active frustration of the natural end of a faculty and yet which was in no way contrary to what is good for us, not even in a minor respect. I submit that there are no such counterexamples, and that there could not be any given an Aristotelian-Thomistic metaphysics of the good.

This looks like a very odd requirement. If a genuine counterexample requires that, then it seems that the first premise should be understood as something like:

1.b. Where some faculty F is natural to a rational agent A and by nature exists for the sake of some end E (and exists in A precisely so that A might pursue E), then it is metaphysically impossible for it not to be in at least some way contrary to what is good for us – even in a minor respect - for A to use F in a manner contrary to E.

What can we say about that interpretation?

First, even under that interpretation, the first premise seems to be false. As I mentioned before, participating in distorting goggles experiments, even as a subject – i. e., wearing the glasses - does not have – at least in some actual cases, not to mention metaphysically possible ones any negative side effects. The fact that our faculty to see is temporarily impaired and made to malfunction does not need to be negative or bad for a person in any way, just as closing one’s eyes for a while does not need to have negative or bad for a person in any way.

Second, even assuming for the sake of the argument that the distortion goggles example fails for one reason or another, under this interpretation the premise would not set apart acts in which A uses F in a manner contrary to E from plenty of other cases in which behaviors that are in some way contrary to what is good for us are overall good for us, and are rational – and this is of course even leaving aside cases of limited information.

Take, for example, major surgery to prevent death. There are plenty of times when there is good information, and surgery is rational, and even cases in which it would be irrational not to have surgery. Yet, surgery is in some way against what is good for the patient: the patient is cut, bleeds, etc., in other words, she suffers wounds that can be very serious – and in the case of major surgery, those wounds are predictably very serious. None of that would happen without she surgery. But the surgery is still overall good for the patient, and rational, at least in most cases, even if it is in some way contrary to what is good for the patient, and in spite of the fact that it is metaphysically possible that the patient would make a full recovery without the surgery.

It might be suggested that in those cases, the surgery is in no way against what is good for the patient, because otherwise the patient probably or very probably would have suffered a worse fate. However, there are metaphysically possible cases in which also something worse happens to the person if the person does not use a faculty in a manner contrary their proper function. So, if, for the purposes of Premise 1, surgery in those cases is considered not against what is good for the patient – not even in a minor respect -, on account of the fact that it probably or very probably it prevents something worse – when probabilities are assessed rationally from the epistemic perspective of the person choosing whether or not to have surgery - , then some of the examples on the previous subsection are also not against what is good for the person who actsnot even in a minor respect -, falsifying Premise 1.

Moreover, we do not need to consider only cases in which major surgery is performed for the benefit of the patient. We can consider organ donation again. At the very least some of those cases – actual cases, not only metaphysically possible ones are cases in which the donor’s behavior is not irrational in the least, let alone immoral. In fact, in some cases, the behavior is even morally praiseworthy. Yet, it is clearly in some way contrary to what is good for her to donate the organ, since she is losing a healthy organ – not to mention the wound she suffers in the extraction procedure -, and her health predictably deteriorates as a result, at least one of her faculties is damaged, etc.

In short, there are plenty of acts that are in some way contrary to what is good for the person A engaging in the act, but in which A acts rationally and in a morally neutral or even morally praiseworthy fashion, even if A has as much knowledge of the situation as humans normally and/or usually have. For that reason, even if it were the case that all acts in which a human agent A uses a faculty F in a manner contrary to its natural purpose E, are cases in which the act is at least in some way contrary to what is good for A, that would still provide no good reason to even suspect that all of those acts are irrational and/or immoral.

But perhaps, Feser did not mean that, and what he meant when he stated the requirements for a genuine counterexample is that in order to be a genuine counterexample to Premise 1, a metaphysically possible scenario must show an action that involves the active frustration of the natural end of a faculty and yet which is not overall contrary to what is good for the agent who acts in that manner?

If this is what Feser meant, then the case is similar to one already considered, and several of the examples in the previous subsection are decisive counterexamples, since I constructed some scenarios in which actions involve the active, even deliberate frustration of the natural end of a faculty, and are not overall contrary to what is good for the person who acts. A potential objection to this would be to deny this, and insist that the actions in all of my examples are overall against what is good for the person who acts, even if overall good simpliciter, or overall good for others, etc. However, that objection has at least two key problems:

I. If scenarios such as Scenario S2, Scenario S3, Scenario S4 are examples of actions that are overall against what is good for the person who acts, then – since the actions are clearly rational -, there are metaphysically possible cases in which it is rational to act in a way that is overall against what is good for the person who acts, and even when erroneous or unusually limited information is not the problem. But if so, then Premise 1interpreted in this manner – would still fail to play any role regardless of whether it is true, since there would be plenty of cases involving rational – and even informed - actions that are against what is overall good for the person who acts.

II. In the distortion goggles examples, as I pointed out, there is nothing negative for the person – not even a little bit. But even assuming otherwise, at the very least it should be even more obvious that the actions of a person who participates in those experiments are not always overall against what is good for her: for example, maybe she wants to contribute to science a little bit, and there is clearly nothing in the experiments that would outweigh that and make them overall negative for her. The same can be said about for provoking vomit when there is no health need for it, or similar examples.

The upshot: If one interprets it is metaphysically impossible for it to be good for Ain Premise 1 to mean that it is metaphysically impossible for it not to be overall contrary what is good for A, then the premise is false, as several examples show. In fact, it clearly conflicts with common sense. On the other hand, if the premise is interpreted to mean that it is metaphysically impossible for it not to be at least in some respect – even if minor - against what is good for A to behave in that manner, then the premise is still probably false, and additionally, even if it were true, it would be pointless to have a premise like that in the context of an argument against the rationality of the behaviors that concern us.

4. Aristotelian-Thomistic metaphysics.

In his defense of the perverted faculty argument, Feser claims that Premise 1 follows from Aristotelian-Thomistic metaphysics of the good, and generally bases his argumentation on Aristotelian-Thomism. While Feser acknowledges that the metaphysical picture in question is controversial and does not defend it in the paper, he makes sufficient claims to raise a number of objections, which I will do in this section.

That said, the following objections are not required for the case against the perverted faculty argument to succeed. In fact, that case has already been made in the previous section, and it stands on its own. However, given how strongly connected the Aristotelian-Thomistic metaphysical view and Feser’s defense of the premises of the perverted faculty argument seem to be, it seems adequate in this context to raise objections to that theory as well.

4.1. Kinds, natures, species, and the ‘old’ natural law theory.

Feser’s perverted faculty argument is based on what he describes as the “old” natural law theory. As Feser explains, this theory of ethics is grounded on Aristotelian-Thomistic metaphysics. Two key features of this metaphysical view are described by Feser as follows:

Feser: In particular, natural law theory as Aquinas and the Neo-Scholastics understand it presupposes an essentialism according to which natural substances possess essences that are objectively real (rather than inventions of the human mind or mere artifacts of language) and immanent to the things themselves (rather than existing in a Platonic third realm); and a teleologism according to which the activities and processes characteristic of a natural substance are “directed toward” certain ends or outcomes, and inherently so, by virtue of the nature of the thing itself (rather than having a “directedness” that is purely extrinsic or entirely imposed from outside, the way artifacts do).

I think there is such thing as objective proper function and similar notions, and also that our claims about the world around us – including but not limited to what Aristotelian-Thomism would classified as “natural substances” - are generally objective, but not in the senses that would follow from Aristotelian-Thomism; this should become clear as I address some of Feser’s examples. So, let us first consider triangles:


A triangle drawn hastily on the cracked plastic seat of a moving bus might fail to be completely closed or to have perfectly straight sides, and thus its angles will add up to something other than 180 degrees. Indeed, even a triangle drawn slowly and carefully on paper with an art pen and a ruler will contain subtle flaws. Still, the latter will far more closely approximate the essence of triangularity than the former will. It will be a better triangle than the former. Indeed, we would quite naturally describe the latter as a good triangle and the former as a bad one.

That is all correct, as long as the word ‘triangle’ is being used in a colloquial sense, rather than a strict mathematical definition of a Euclidean triangle: in such strict sense, we have never seen a triangle, and nothing we can draw actually meets the criterion, but in colloquial speech, the word ‘triangle’ is not used in that manner, and as Feser correctly points out, there are good and bad triangles, triangles that are better than others, etc., and all of them are triangles.

But now, suppose we make a computer program that draws triangles and, in many different ways, very gradually transforms them – in a gazillion steps, changing one pixel at a time – into circles. So, we start with a computer generated triangle, and that is an extremely good triangle. In fact, it is better than the triangle drawn carefully on paper in Feser’s example[D]. And if only one pixel is changed, the result will still be an extremely good triangle. However, some of the gazillion triangles that the computer generates before we get a circle on our screen will be mediocre ones. Some will be pretty bad triangles. And some of the transition figures will neither be triangles nor circles.

Yet, there seems to be no good reason to think there is a specific ‘first non-triangle’ in the series of figures that starts with a very good triangle and ends with a very good circle, or a specific ‘first bad triangle’. In fact, it seems improbable that there are such firsts. Similarly, there seems to be no good reason to think there is a specific ‘first circle’ or ‘first good circle’. Rather, it seems - at least - probable there is no objective fact of the matter as to whether some of the figures that are triangles are bad ones, or whether some of the figures are bad triangles or non-triangles, or whether some are circles, or good circles, etc., in the usual sense of the expression ‘no objective fact of the matter’. This does not appear to be an epistemic issue: It is not that we are not in a epistemic position to tell which figure is the first non-triangle, etc. Rather, the problem seems to be a problem of language: our colloquial term ‘triangle’ is very probably not precise enough for there to be an objective fact of the matter as to whether some figures are triangles, or bad triangles, or good circles, and so on.

Now, if there were a ‘first non-triangle’, let F(1) be the first figure on screen, and F(k0) the last triangle. Then, F(k0+1) is just F(k0) with a single pixel changed, and very far more similar to F(k0) than F(k0) is to F(1). In fact, F(k0+1) is so similar to F(k0) that they are indistinguishable to the naked human eye. On the other hand, both F(k0) and F(k0+1) are vastly different from F(1). They are vastly different in the way they look to us humans, and in the number of pixels they have in common.

It seems at least very improbable that there is a metaphysical chasm between the almost identical F(k0) and F(k0+1), whereas there is metaphysical similarity between F(1) and F(k0) – which allegedly belong to the same kind, “triangle” and share the same essence. Yet, the Aristotelian-Thomistic picture supported by Feser seems to imply that F(k0) and F(1) belong in the same ontological category ‘triangle’, whereas F(k0+1) does not. Similar considerations apply to the question as to whether there is a first ‘bad triangle’, or ‘not good triangle’, etc.

Now suppose that some intelligent, scientifically advanced aliens evolved on a different planet – let us call them ‘species #2’. We should not expect that, in their language, they would have a colloquial word that has the same meaning as the colloquial English word ‘triangle’, though maybe we should expect a similar one. Assuming they do have a language similar in structure to ours, since they and humans probably had to resolve similar problems along their evolutionary past, it seems pretty plausible that they will have words in their language or languages – let’s say a single language to simplify – very similar in meaning to the colloquial words ‘triangle’, or ‘circle’, but we should not expect that the fuzzy transition between ‘2-triangle’ (in their language) and ‘2-non-triangle’ were almost indistinguishable from the fuzzy transition between ‘triangle’ and ‘non-triangle’. In fact, it might well be that some of the figures generated by the computer would be triangles, but not 2-triangles. Members of species#2 would be able to tell the difference between a 2-triangle and a figure that is not a 2-triangle – at least, when the figures do not fall in the fuzzy transition zone -, but that would not allow them to tell the difference between triangles and non-triangles, which may well have quite a different fuzzy transition zone – i. e., the fuzziness may well be in different places. If they wanted to learn more about the difference between triangles and non-triangles, they would need to study humans, and more precisely human language, human minds, etc.

Similarly, #2-aliens would know how to tell the difference between 2-good 2-triangles and 2-bad 2-triangles. However, in order to tell good triangles from triangles that are not good, they would have to study humans – and not just triangles.

Why should we think such differences are metaphysically possible, and even nomologically possible?

Because it is metaphysically and even nomologically possible that both their sensory apparatus and the way their minds categorize objects in several respects could be are different from ours.

But moreover, there would be no good reason to assume that even if metaphysically possible, something like the scenario described above would be improbable. In fact, it’s rather probable if there are intelligent, rational, talking aliens, since in spite of the likelihood of convergent evolution and even the probability that most if not all such aliens will have a word like ‘triangle’, there seems to be no way for evolution to make the similarities in both the sensory apparatus and the associated mental capacities between different intelligent species so vast as to prevent the transition between triangles and non-triangles and between their alien analogues to happen in different places.

In fact, figures that look very similar to us might look quite different to them, due to their different visual system. They might things we do not. And whether something is a triangle or not – or a good or a bad triangle - does depend on the way it looks to a human with a normal sensory apparatus – not to an alien. Surely, we would not say that the good, carefully drawn triangle Feser was talking about is actually a bad triangle just because some aliens with a much sharper visual system qualify it negatively in their language and with respect to their concepts; i. e., that might be a #2-bad #2-triangle, but still a good triangle.

Additionally, – and I think even more importantly -, due to differences in their minds and categorizations, even if, say, a figure F1’ on the screen looks to them as F1 looks to us, and F2’ looks to them as F2 looks to us, it may well be that F1 and F2 belong in the same category ‘triangle’ in English – or in the same category in other human languages – but while F1’ belongs in their category ‘2-triangle’, F2’ does not.

The upshot: there is objectivity in the sense that in nearly all practical cases, there is an objective fact of the matter as to whether something is a triangle, or as to whether a triangle is good or not. But it seems probable that in some metaphysically and even nomologically possible cases, there is no objective fact of the matter, and the transition is fuzzy. Moreover, the classification of objects in triangles and non-triangles, or in good and bad triangles, is, in a sense, depending on human minds. Of course, that does not make it subjective in the sense of the colloquial sense of expressions like ‘it’s a subjective matter’. As I mentioned, there usually is an objective fact of the matter as to whether something is a triangle, or a good triangle. And to be absolutely clear – and to dispel a potential misunderstanding - it is of course not the case that whether a figure is a good triangle – or a triangle at all – depends on whether the human assessing whether it is so believes it, of course. That claim would be mistaken in more than one way. However, the whole classification of figures in triangles and non-triangles, and of triangles in good and bad ones depends on how the words are used by a human linguistic community, which in turn depends on human minds. Whether they also depend on human culture – and so, in addition to species-relative, they would be culture-relative – depends also on human minds: some concepts seem to be human universals in the sense that all human communities have them – e. g., “morally wrong” -, and others are not. And in the case of universal concepts, usually languages have a word for them. Maybe the concept of a triangle is a human universal, or maybe not. I do not know. But it’s extremely improbable that the colloquial concept of a triangle is an all-metaphysically-possible-rational-beings universal.

So, our regular assessments about good and bad triangles, non-triangles, etc., are generally correct, but the essentialism in the Aristotelian-Thomistic metaphysical picture of the world is not. In a weaker sense, there is an essence of triangles, but given by our human word ‘triangle’ - and thus, by our human minds -, and furthermore, it is fuzzy.

But let us now turn to living organisms.


Any particular living thing can only be described as an instance of a species, and a

species itself can only be described in terms of Aristotelian categoricals stating at least its general characteristics. If a particular S happens not to be F - if, for example, a particular cat is missing a leg—that does not show that S’s are not F after all, but rather that this particular S is a defective instance of an S.

The claim about the cat is true, but the conjunction of the assertions that an individual living organism can be described only as a member of as a species and that species can only be described in terms of Aristotelian particulars is very probably false. In fact, both conjuncts are probably false.

Regarding the first conjunct, why only a member of species?

Biologists also distinguish between genera, and between subspecies. And botanists also distinguish between varieties. There are plenty of classifications in biology. If there is only one on Aristotelian-Thomism (namely, species), does it match one of the scientific ones? Is it something else entirely?

Maybe this Aristotelian-Thomistic concept of species is not related to the scientific one?

In any event, let us leave the first conjunct aside, since there might be some terminological misunderstandings, and let us focus on the second conjunct - namely, that species “can only be described in terms of Aristotelian categoricals stating at least its general characteristics”. Let us consider a concrete case: Cecil the Lion. Cecil was definitely a lion. His father was also a lion. And so was his grandfather, and so on. However, Cecil’s most recent common male ancestor with Dolly the Sheep was surely not a lion. Let L(1) be Cecil, L(2) be Cecil’s father, and generally, let L(k+1) be the father of L(k) for all k ∈{1,…,n-1}, and let L(n) be the most recent male common ancestor of Dolly and Cecil. If every individual is a member of a species that can only be described in terms of Aristotelian categoricals, it seems this would cause strict separations between species, so that every individual belongs to one of them. If this is so, then for every k ∈{1,…,n}, there is a (perhaps extinct) species S(k) - each described in terms of Aristotelian categoricals - such that L(k) is a member of the species S(k). Clearly, S(n)≠S(1). Therefore, on the Aristotelian-Thomistic picture, it seems there is a minimum k ∈{1,…,n-1} for which S(k+1)≠S(1). Let k0 be that minimum. It follows that L(k0+1) was a member of the species S(k0+1) which is not the same species as S(1) – i. e., he was not a lion, if lions make up a species in this Aristotelian sense -, but fathered a lion L(k0). Yet, L(k0) was almost certainly much more similar - genetically, morphologically, physiologically, psychologically, behaviorally, etc. - to L(k0+1) than he was to Cecil – or, for that matter, to any living lion. Was L(k0) a very defective lion? [E]

The point is that there is no good reason to think there was such an individual. In fact, if human scientists (coming from a parallel universe, or whatever) had found L(k0) and L(k0+1), they would almost certainly have classified them as members of the same species, if they classified biological organisms as we do. Would they have been in error? Are our present-day classifications generally unreliable too?

It seems probable that species are not categorically different in an Aristotelian sense. In fact, that theory would seem to require that there would be a ‘first lion’, and so – as we know from observations so far - there would be almost no difference in terms of genetics, psychology, behavior – of course, informed by psychology -, shape, etc. between the first lion and his father, yet metaphysically, there would be a leap between them, whereas there would be individuals sharing the same nature and being far more different in any of the previously mentioned senses. That metaphysical picture is very improbable. Much more probably, what we classify as a species depends on what population of organisms we are observing, and this applies to science as well as to our colloquial terms. For example, since we observe present-day lions, we have a term ‘lion’ that refers to some present-day organisms, and organisms that are ‘similar enough’, in the sense that they share with some present-day organisms a number of properties, and of course the transition as we go back in time between lions and non-lions is fuzzy. However, if, instead of the lions we have observed, we had encountered individuals just like their ancestors 1200000 years ago, we would have instead a term – say ‘lion*- that would pick a similar but different set of properties, and the fuzzy transition would lie elsewhere, even if the referent of the term would have some overlap with the referent of our term ‘lion’ in the actual world.

Moreover, if aliens were to visit Earth and study the planet’s biosphere, the term in their language that referred to lions may well not have the same referent as our term ‘lion’, even if a very significant degree of overlap in their referents – in the actual world[F] – is probable.

Does any of the above imply that there is no such thing as objective judgments about species, either in the sense or senses of ‘species’ in which biologists use the term, or in a colloquial sense if there is one and they differ?

No, not at all. There is an objective fact of the matter as to whether Cecil was a lion. He was. But for that matter, if we decided to classify individuals not in species, but in any way that we fancy (e. g., for a classifications of animals, the first category is animals that have a Wikipedia article about them, the second are those animals not in the first category and accidentally killed by vehicles on some road, the third are those not in either of those two categories and eaten by people, etc., and we make other categories like that such that the last one is the category of those animals not in any of the previous ones), there would be an objective fact of the matter as to whether or not an individual belongs to one of those categories or another – well, at least in nearly all actual cases, if we are precise enough when we come up with the rest of the categories in our example. So, objectivity in our categorization is metaphysically cheap so to speak. Note that this is similar with the case of artifacts: purely for example, in at least nearly all cases, there is an objective fact of the matter as to whether, say, a vehicle is a car, or a truck, etc.

Does this mean that our usual categorizations of biological organisms – whether colloquial or scientific -, while objective in the sense described above, are just something we make up, without any restrictions imposed by the things we study, just like the example I gave above?

As before, the answer is negative. Our colloquial terms track features that some organisms have in common, and those are features generally salient to us because of the minds we have. Our scientific terms are usually either a refinement of colloquial terms, or are introduced as a means of studying features of organisms we find around us and that are also interesting objects of study – pretty much to any human who studies those organisms, or close to that. So, the categories are constrained both by what we find in the biosphere around us, and by human psychology. While they are similar to a categorization like “animals with a Wikipedia article”, etc., in terms of objectivity, they are dissimilar in other respects.

In short, I am not suggesting any sort of radical skeptical view here. Our usual terms do refer. Our scientific and even colloquial claims about different organisms are usually objective, and often true – even if less often in the colloquial case. They are also useful for studying the world around us, etc. What seems very improbable is the Aristotelian-Thomistic picture, particularly in light of gradual evolution.

But is evolution as gradual as I have described above? Or are there always distinct ‘jumps’, which might warrant the claim that there are sharp metaphysical boundaries?

It is an empirical matter, but the evidence clearly and very strongly supports gradualism in the case of complex species – at least. On that note, let us consider the evolution of polar bears from brown bears as an example:

Polar bears have a number of adaptations to the cold that are not present in brown bears, and it is surely not the case that a female brown bear just gave birth to full-fledged polar bear cubs. Indeed, we have observed many generations of brown bears. Hybrids aside – more on those below -, in every case that has been observed, from two brown bears we got cubs that were brown bears - unsurprisingly. Some of the cubs might have a mutation that results in slight differences from their parents, and some of the mutations sometimes might increase their chances of reproducing, in some particular environment. But the cubs will still grow up to look, behave, etc., almost like their parents. In order to go from brown bears to polar bears, many steps were required, and in every one of them, the offspring was surely much more similar to the the parents than to their descendants after many generations.

Are polar bears somehow an exception in which evolution happened very slowly?
Actually, they are an example of very
rapid evolution for such a complex species. Yet, what is very rapid in evolutionary terms, is still so slow that the description of the gradual process that I gave above holds. The following quote is from a study on polar bear evolution. [4]

Assuming an average generation time of 11.35 years (Cronin et al., 2009; De Barba et al., 2010), the distinct adaptations of polar bears may have evolved in less than 20,500 generations; this is truly exceptional for a large mammal. In this limited amount of time, polar bears became uniquely adapted to the extremities of life out on the Arctic sea ice, enabling them to inhabit some of the world’s harshest climates and most inhospitable conditions.

So, it took at least about 20000 generations to go from brown bears to polar bears, and that is exceptionally fast for large mammals.

Granted, the Aristotelian-Thomist might say that while there are like 20000 generations from brown bears to polar bears, there are in fact many intermediate species, at least in the Aristotelian-Thomistic sense of species that does not need to match usage by biologists. However, this would miss the main point of the gradualism objection, which is the fact that the members of one generation were extremely similar from those in the next, and actually more similar to them than the latter were to those several generations down the line. Positing many Aristotelian-Thomistic species that do not match our concepts would not seem to work, since there would still be big metaphysical jumps where there is very little genetic, morphological, psychological, etc., difference, and no metaphysical jumps where there is much bigger difference.

The problem of gradualism holds for non-human mammals – and birds, and reptiles, and many other species at least -, but it becomes particularly salient in the case of humans, where the metaphysical leap is far bigger on Aristotelian-Thomism than in the other casesa matter I will address later in greater detail.

Perhaps, the Aristotelian-Thomist might suggest considering each new trait resulting from a mutation or mutations a new species, but that would not work for at least the following reasons:

First, there are plenty of adaptations (in different organisms, but especially complex ones, like mammals) that occur gradually, and are the result of many mutations over many generations, rather than a single mutation. Considering every one of the tiny steps the emergence of a new species would not seem reasonable, in light of our usual concepts: it would imply that - for example - a brown bear that, due to a mutation, tends to have a very slightly greater amount of fat than her mother, is a new species. But given the differences within species, it would turn out that most of the species we regularly talk about are actually many different species – in fact, single-individual species, or close to that -, hybridizing pretty much every time they breed. Again, that does not seem to be in line with the way we normally talk.

Second, new traits often do not result in new species, under any reasonable understanding of the concept of ‘species’ that matches or approaches the way we regularly talk. For example, some human groups have adaptations for living in high altitudes, and others for digesting milk even when they are adults, but that does not make them members of different species. More and bigger adaptations would eventually result in speciation, but then, that would take much longer...and still, there would no specific, precise amount of change that would bring about the change from humans to some other species. Rather, the transition would be – as usual – fuzzy.

Now, polar bears and brown bears can - and sometimes do - mate and have fertile offspring. So, it might be suggested that they belong in the same species. However, this sort of objection is not an option in the Aristotelian-Thomistic sense of species, for the following reason:

Lions of any generation – for example - had the capacity to mate with lions of the immediately previous generation, and leave fertile offspring. But we can go back to before there were any lions, one generation at a time. So, assuming that the capacity to mate and leave fertile offspring implies belonging in the same Aristotelian-Thomistic species, then by transitivity of being the same Aristotelian species, we can go back from lions to, say, the latest common ancestor between lions and mosquitoes, and then forward to mosquitoes, concluding that mosquitoes and lions belong to the same Aristotelian-Thomistic speciesan absurd result.

In short, that sort of objection does not work for the Aristotelian-Thomist. Moreover, hybridization on its own might be a problem for Aristotelian-Thomism: the offspring of a female polar bear and a male brown bear is not a brown bear or a polar bear. So, what species is it?

Still, let us leave bears aside, and focus now on humans. Feser says that humans are ‘rational animals’, and considers that all other species are not, and indicates a radical distinction between humans and non-humans. For example:

Feser: Hence while perception is a good for both non-human animals and human beings, that perception in our case participates in our rationality makes of it a different and indeed higher sort of good than that of which non-human animals are capable. Other goods we share in common with animals similarly participate in our rationality and are radically transformed as a result.

Now, on Aristotelian-Thomistic metaphysics, the goods depend on our nature, and particularly the nature of our faculties. So, according to this, the nature of our faculties would be radically transformed as well. Indeed, this is what Feser says about our sexual faculties: while he claims that they keep the reproductive purpose of other animals, he ascribes to them a radical transformation. I will address some of his claims about our sexual faculties in the next section, but here, I would like to address the alleged radical transformation, and argue that it seems extremely improbable, given the gradual picture of evolution. Of course, the difference can be huge after a very long time, but the problem is that when there is only a very small difference even psychologically between generations, a radical metaphysical difference seems very improbable.

For example, let us say that Lucy and her daughter Joan – just to pick two names; there is no further meaning to them – are two of our ancestors, at some point after the split between the lineage leading to chimpanzees and bonobos, and the lineage leading to modern humans. Given the almost zero genetic distance and the fact that they grew up in essentially the same social group, the brains and minds of Joan and Lucy were very probably extremely similar. In fact, Joan would have been raised among her mother’s social group, learning from them – by looking at their behavior and intuitively picking things up or by being taught - things like what is safe to eat and what isn’t, which animals are dangerous, how to communicate with other members of the group in whatever language or proto-language they had, and even the rules of behavior, tool-use and tool-making. The point is that Joan and Lucy would be extremely similar not only physiologically, morphologically, and genetically, but also behaviorally and psychologically. The idea of a huge leap that would render the functions of Joan’s faculties radically different from those of her mother Lucy’s – or generally, that the nature of the goods for Joan would be radically different from the nature of the goods for Lucy – is extremely improbable.

4.2. Goodness.

In this section, I will continue to address Feser’s claims about the teleology of organisms, in particular focusing on the matter of goodness.


In living things the sort of norm in question is, as Foot also notes, inextricably tied to the notion of teleology. There are certain ends that any organism must realize in order to flourish as the kind of organism it is, ends concerning activities like development, self-maintenance, reproduction, the rearing of young, and so forth; and these ends entail a standard of goodness.

Does the nature of an organism entail a standard of goodness?

Before we address the matter, let us consider a color analogy. Surely, we make objective statements of color very often. For example, if there is a court case about whether a car driver ran a red traffic light, generally there is an objective a fact of the matter as to whether the light was red when the car began to cross the street (or some part of the car, if more precision is required). Moreover, our statements ascribing color to objects are usually true.

Now, many other animals have visual systems that react to parts of the electromagnetic spectrum considerably differently from ours. For example, some animals can perceive wavelengths that we do not perceive, and are ultraviolet or infrared. Some of those animals, in turn, cannot perceive some of the wavelengths we perceive. So, we may consider a hypothetical scenario in which intelligent agents also have a different visual system.

S5: On a distant planet, an intelligent species evolved – say, species#3. They have language, science, technology, and a visual system somewhat similar to ours: they perceive color (or color-like, or whatever the correct terminology is) differences between things in their environment – much like we do -, but their perceptions are associated with considerably different ranges of wavelengths. In particular, some objects that look the same to us color-wise do not look the same to them, and vice versa.

That scenario is clearly metaphysically possible. In fact, I would say it is also nomologically possible, though this is not required for the argument. But in species#3-language (but let’s say it’s a single one to simplify), there would be no synonyms for the color words we have in English, or the color words in any human language. They would have words for #3-colors, but not for colors. Just as our color vision tracks some properties in the world around us – reflective properties and/or dispositions and/or some other properties, whatever the correct account of color is -, their #3-color vision would track other properties, and their language would reflect that. Sometimes, two objects would be of the same #3-color but not the same color, or vice versa.

Another way to see this is as follows: suppose species#3 civilization was already very advanced long ago, and some of those aliens visited the Earth 180 million years ago, in the Jurassic. It is extremely improbable that they would have learned the colors of any objects on Earth. They would have been able to determine the #3-color properties of many objects. They could have studied the visual systems of many animals on Earth. But it is very improbable that there would have been anything close enough to our own visual system to get the aliens to detect colors. In fact, these aliens would not have even considered categories such as ‘red’ orgreen’, but analogues. However, even if – though extremely improbable – there was on Earth something with a visual system similar enough ours 180 million years ago, we just need to set an earlier date – say, 500 million years ago, during the Cambrian, and there is no way there was anything close enough to us. Crucially, without studying humans, or some animals close enough to humans, there is no way these very advanced aliens could have determined the color of things, or even think of color.

It might be objected that #3-color would still be color, so the aliens would learn some of the colors of objects – just not the colors we talk about, but other colors we do not know about. However, if color words are like that and #3-colors are colors, the point would be that without studying humans – or some animals close enough to humans -, the aliens would not have been able to learn that some of the things they found on Earth were green, or red, etc.

The upshot is that while it seems very probable that color is in a real sense in the objects to which we ascribe colors perhaps some reflective properties, or some dispositions to have certain effects, or whatever it is -, and our senses are tracking some real stuff in the world around us, the standard of color is not in the objects: that one is in our language, in the meaning of our words, and we use those words because of the visual system and minds we have, so in the end the standard is in us humans though there is room for different classifications of colors even in human languages, for the most part our visual systems are roughly the same color-wise.

Let us now turn to goodness. At this point, it is useful to conceptually distinguish at least three different senses in which we can evaluate goodness.

G1: What is good.

G2: What is good for some agent A.

G3: What would make an agent A not defective and/or a good example of some species S(A) of which A is a member.

It is a property of individual organisms to be defective or not, healthy or not, etc. However, and similarly to the color case, it seems probable that the standards are not in individual organisms or in their nature, but in our minds, and in that manner, in our language. Now, if intelligent aliens with language evolved on another planet, I think they would probably have words with referents very similar to our words ‘illness’ and ‘health’, ‘properly functioning’, or ‘defect’, in the sense that there would actually be a vast overlap between the referents of some of our words and the alien close counterparts. In fact, I think their assessments – and their language – would likely be much more similar to ours than in the color case. Yet, we should not expect the same meaning, or referent. One reason for that is that the orders would probably be somewhat different: To give a simplified example, it might happen be that individuals A and B – say, two lions – both have 20 defects and 20 #3-defects (in the language of species#3 in the previous example), and even that each of the 20 defects is a #3-defect and vice versa, but overall, A is somewhat more defective than B – he has somewhat worse defects -, whereas B is somewhat more #3-defective than A – he has somewhat #3-worse #3-defects. There appears to be no way evolution would result in uniformity across the universe if there are many planets like ours, even if a lot of similarity is to be expected.

So, even in the case of G3 above, it is probably nomologically possible – and hence, metaphysically possible – that there would be aliens with a different language, talking about different stuff - even if very similar to human language -, so their (analogous) standard would be different.

The upshot is that probably, the standard of goodness in the sense of G3 is not in the nature of things, but in our human minds, even if the illness, or defect, or badness, is in the individual organism, event, etc.

The difference probably would increase (I would say significantly) when it comes to judgments about something analogous to what is good or bad in the sense of G1 above. But in any case, the point is that there could be differences too. For example, let us consider the following scenario:

S6: On a distant planet, there are intelligent aliens with language, science, technology, etc. They evolved from something like squid. They have evaluative language, and they make judgments about what is alien-squid-good or bad, alien-squid-better, etc.

On a second planet, there are intelligent aliens with language, science, technology, etc, also with language like that, but they evolved from something like elephants.

Now, in a situation like that, it would seem improbable that alien-squid-bad things – events, states of affairs, objects, etc. - are the same as bad things or the same as alien-elephant-bad thing, even if there is wide overlap in the referents in the actual world[F]. For example, if scientists from both species come to Earth today and they see a pride of very hungry lions trying to bring down an adult elephant, it may well be that the elephant-like aliens truthfully reckon it would be alien-elephant-better if she escapes and the lions starve to death, whereas the squid-like aliens truthfully reckon it is alien-squid better – or alien-squid-neutral - if they bring her down and ear her – while she’s still alive for a considerably long time, since lions do not have the means to swiftly kill an adult elephant.

Why should we think that that is likely?

The meaning of words is determined by usage, and their usage would depend on their minds. Given the big differences between their minds, and in particular between some of the things – events, states of affairs, etc. - that individuals of the different species generally value positively or negatively, one should expect different meanings and referents in their words as well.

So, if those squid-like or elephant-like aliens had come to Earth in the Jurassic rather than today (for example), then – and just as in the color case -, they would not have been able to figure whether any of the events or states of affairs, animals, events, etc., that they encountered on Earth were good or bad events, states of affairs, things, etc., even if they generally would have been able to ascertain what was alien-squid(or elephant)-good or alien-squid(or elephant)-bad, and there would have been a big overlap between those and what is good or bad. In fact, in order to learn about good or bad things – in the sense of G1 above -, they would have had to study humans - or at least, some non-human animals that are relevantly close enough to humans; maybe bonobos or chimpanzees - but nothing living on Earth in the Jurassic would have given them any information about good or bad events, states of affairs, etc.

Similarly, there probably would be some differences between goodness in the sense of G2 and an alien analogue – though I think probably the difference would be much smaller than in the G1 case, but that is a side issue.

None of this is is not a problem for our ordinary assessments about whether something is a defect, or an illness, or whether something that happened is a good or a bad thing, event, state of affairs, and/or bad or good for an individual organism etc., just as alien colors are not a problem for our ordinary assessments about the colors of things. In particular, there usually is an objective fact of the matter regarding questions about G1, G2 or G3 above, just as there is in the case of color questions.

However, the Aristotelian-Thomistic metaphysical picture seems to be wrong on this as well: the standards are in us, even if the goodness, badness, defect, etc., are in the individuals, things, events, etc., we talk about.

4.3. Rationality and morality.

In his defense of the perverted faculty argument, Feser makes a number of claims about reason and morality. In particular, he seems to base morality on rationality – following Aristotelian-Thomism. That seems to be misguided. Humans have a sense of right and wrong, which is the result of evolution and came gradually from other primates that had a somewhat different moral or proto-moral or moral-like sense – whatever the correct semantics of the terms are. But if aliens evolved on different planets, they might – and they metaphysically possibly could, and even probably would - not have morality, even if they were rational, and they could talk, build spaceships, etc., and even if they had a close analogue to morality. This is without even considering AI with superhuman intelligence and capacity to reason, which might be radically different and have no analogue to morality, at least not an analogue to the idea of moral obligations.

But Feser says:


Not only the content of our moral obligations but their obligatory force are thus determined by natural teleology.

This metaphysical picture seems to imply that any rational agents – like any advanced aliens – would have moral obligations – rather than, say, alien-squid-moral obligations -, and that it would be irrational on their part to behave immorally. Neither implication appears plausible, but leaving that aside, on the Aristotelian-Thomistic view, the aliens’ moral obligations would be determined by what is good for them. That might lead to all sorts of horrific moral obligations. For example, it might be good for a squid-like alien stranded on Earth in 50 CE to hunt and eat humans if they’re available, and it might even be bad for such an alien not to hunt and eat humans. But it would be a bad thing if it did hunt the humans, and surely it would not be morally obligatory for the alien to do so just because it would be bad for it not to do it.

Granted, the squid-like aliens are a hypothetical species only. But apart from the fact that they may very well be plenty of advanced aliens somewhere out there that, even if not squid-like, are different from us in the senses that are relevant in this context, these sorts of aliens seem to be metaphysically possible, and Feser’s claims in this context are about metaphysical possibility and necessity.

Still, let us grant for the sake of the argument that the objections I just raised in this subsection do not succeed, and let us take a look at the way Feser’s explanation of the grounds of moral obligation. With a number of qualifications, Feser explains the grounds of moral obligations, according to the Aristotelian-Thomistic view he adheres to.

Feser: The hypothetical imperative (1) If I want what is good for me then I ought to pursue what realizes my natural ends and avoid what frustrates them is something whose truth follows from the metaphysical analysis sketched above. By itself, it does not give us a categorical imperative because the consequent will have force only for someone who accepts the antecedent. But that (2) I do want what is good for me is something true of all of us by virtue of our nature as human beings, and is in any case self-evident, being just a variation on Aquinas’s fundamental principle of natural law. These premises yield the conclusion (3) I ought to pursue what realizes my natural ends and avoid what frustrates them. It does have categorical force because (2) has categorical force, and (2) has categorical force because it cannot be otherwise given our nature. Not only the content of our moral obligations but their obligatory force are thus determined by natural teleology.

It should be clear by now that (1) is false due to several counterexamples, like those in Scenarios S2, S3, S4, the case of distortion goggles, and others. Examples like that also show that there are metaphysically and nomologically possible situations – and of course, actual ones, but we do not need that in this context – in which if a human being has no psychological defect and wants to do what is good for him, then a rational and morally permissible – or even morally praiseworthy - course of action would be for him to use a faculty in a way contrary to its function or natural end. Moreover, sometimes it would be irrational on his part to fail to do so.

So, (1) is false. Let us now turn to (2). Do I want what is good for me? In a qualified manner, which does not seem to be enough for the premise. To see why, let us first consider an analogy with justice. Do I want that that justice be done?

In a sense, yes, but that is conditional to ‘justice’ being similar enough to what I generally reckon justice is. For example, there are Christians who believe that all human beings deserve infinite torment in Hell – even if they believe Christians will be spared. Of course, I find all of that ridiculously improbable, but assuming for the sake of the argument that that would be a just punishment for all human beings, I do not want justice to be done.

Moreover, according to some Christians, human beings who fail to believe that Jesus is their lord and savior until they die (under conditions that vary with the Christian; e. g., that the Gospel was preached to them, etc.), will be punished with eternal torment, and that is just, deserved punishment. Now, it is pretty clear to me that such punishment would be undeserved to an absurd degree. However, assuming hypothetically and for the sake of the argument that those Christians are correct about the justice of the punishment and about who will be punished, it remains the case that I do not want people to be punished for that reason. Of course, under that assumption, that would happen regardless of what I want, but I do not want it. In fact, if the referent of “just” happened to be like that sort of religion claims, I do not in general want justice to be done – rather, that would very much depend on the specific case at hand.

I think my position on this is not unusual. When people say they want justice, they say it with implicit qualifications. Their qualifications may well be very different from mine – in the cases of some people, they surely are -, but I think most people who would agree to ‘I want justice to be done’ are not implying for example - that if justice involves eternal torment for every human being who has ever lived, they want that to happen.

Going back to the case at hand, do I want whatever is good for me?

In a sense yes, but this is also a qualified ‘yes’, similar to the sense in which I want justice done. In particular, I am assuming that what is good for me – even if I do not always know it – is similar enough to what I generally reckon is good for me. In general, I would make the following assertion instead of (2):

(2)’’(From my perspective): I do not always know what is good for me, but whatever it is, I generally want what is good for me, as long as that ‘whatever it is’ is qualified as follows: I believe my ordinary assessments about what is good for a person – and, in particular, for me – are not radically wrong, as they would be if, say, some metaphysical and/or metaethical theories and/or religions were true. So, if the ‘whatever it is’ happens to be radically different from my ordinary assessments of goodness, then in order to tell whether I want what is good for me I would need to address the matter on a case-by case basis.

In particular, under the hypothesis that it is never good for any human being to use a faculty in a manner contrary to its function or natural end, then it is not the case that I always want what is good for me, no matter the circumstances. For example, if – given the situation - the achievable alternative to using a faculty in such a manner by, say, smoking “to excess”, is to be tortured to death but – by assumption – that is the best for me, then I no longer want what is best for me: I would much rather smoke “to excess”, and avoid the torture to death, even if the course of action I choose is bad for me for some obscure metaphysical reason that has no practical consequences.

But moreover, there does not need to be a threat. Granting that it is bad for me (due to the Aristotelian-Thomistic metaphysics of the good), I would still agree to wear distortion goggles for a few minutes, given adequate, non-coercive incentives. For instance, I would agree if I’m at a science fare, a friend is doing the experiments and asks me to participate, and I have the time. I would not have any inclination at all to reject something simply because it is bad for me because of some obscure metaphysical fact (we assume that it is a fact, that is), but has no actual impact on my life that I value negatively, by my own evaluative function. True, under those metaphysical assumptions, my evaluative function would seem to have a defect, but – as in the case of justice or what is good for me - my positively valuing having no defects is also subject to some implicit assumptions about what is or might be a defect.

5. Our sexual faculties.

In his defense of the perverted faculty argument, Feser also makes some claims about the end of the sexual faculties – both in humans and in other animals:

Feser: That sex considered from a purely biological point of view exists for the sake of procreation is uncontroversial. This is true even though people have sexual relations for various reasons other than procreation, since we are talking about nature’s ends here, not ours. In particular, it is true even though sex is pleasurable and human beings and animals are typically drawn to sex precisely because of this pleasure. For giving pleasure is not the end of sex, not that for the sake of which sex exists in animals. Rather, sexual pleasure has as its own natural end the getting of animals to engage in sexual relations, so that they will procreate.

It is at least controversial that there is a function or ‘natural end’ that is not biological in the case of humans, though it might be a matter of terminology. But let us leave that aside. What is not controversial is that one of the functions of the sexual organs – and our sexual faculty or faculties - is reproduction. Yet, that does not imply or suggest that that is the only function of the sexual organs or the sexual faculty, and that there are no other functions. There might be more functions even in some (or many) non-human animals, even if reproduction is the main one. This is so not only for humans, but for non-human animals as well.

For example, same-sex sex in bonobos may well have non-reproductive functions too, like a function of reducing conflict, or reinforcing alliances between females, or both. Indeed, what the evidence so far indicates that same-sex relations – involving bisexual individuals – are common among bonobos, both in captivity and in the wild[5] [6]. Now, it might be that their same-sex behavior is a defect in their sexual function, but it might not be so, and by looking at how common that behavior is, there seems to be a good chance that it is not a defect – sometimes, defects might be prevalent in a species over non-defective traits, but that is not the most common situation.

Similarly, masturbation may also have a function in some species – say, to reduce stress, or keep the sexual organs in good health in some cases, increasing fertility.

In humans too, it might be that sexual organs and the sexual faculty or faculties have one or more non-reproductive functions in humans that come from our non-human ancestors, perhaps involving masturbation or same-sex relations in some circumstances. Note that this might be so even if one granted for the sake of the argument that there was such a radical metaphysical chasm between humans and non-humans: even a radical transformation in the functions of the sexual organs and/or faculties would not need to make the all of the already existing functions other than reproduction go away, if there were any such functions.

Now, I am not claiming that there are such functions. I am saying that there might be – though I think whether there are such functions is not morally important in this context.

Let us address another point:

Feser: And we’re built in such a way that sexual arousal is hard to resist and occurs very frequently, and such that it is very difficult to avoid pregnancies resulting from indulgence of that arousal. The obvious conclusion is that the natural end of sex is (in part) not just procreation, but procreation in large numbers. Mother Nature clearly wants us to have babies, and lots of them.

That tells us that an aspect of our nature inclines us to have lots of sex. Similarly, we can take a look at our gustatory and olfactory senses, and conclude that some aspect of our nature inclines us to not only eat sugary meals, but to eat a lot of them: after all, sugary meals are often hard to resist, taste great, and so on. We have that inclination because in the ancestral environment, seizing the opportunity to eat that sort of thing would have been conducive overall to reproductive success, even if it would generally be bad for us in an environment in which there is a lot of sugar available.

So, given the available evidence, the proper assessment is that some aspect of our nature – rather than “Mother Nature” in an unqualified manner - makes us inclined to eat a lot of sugary food. But in some – many - circumstances, it would be a bad idea to “let nature take its course” when it comes to sugar. Better options are refraining from eating a lot of sweet things, or eating artificial sweeteners – which allow us to get some of the pleasure of eating sugary foods, without some of its negative effects.

Back to the question of having lots of babies, a very probable assessment from the available evidence is that in the ancestral environment in which humans evolved, having many children – at least, many by today’s averages in developed countries - was a trait selected for. Many of those children would have died of starvation, disease, etc., but having many would have increased the odds that some might reproduce.

But while an aspect of our nature makes us inclined toward behaviors which, lacking contraception, would result in a numerous offspring, that does not tell us that it is good for us under all conditions, or even under most present-day conditions, to have many children, or even to have any children at all – just as our inclination for sugary foods does not tell us that such foods are good for us under all conditions, etc.

In fact, it should be clear that it would be bad for at least some – many - of us to have many babies – babies who, predictably, if they did not die very early would grow into malnourished children, also with little education and little parental care due to lack of available time, as both parents would have to spend most of their time just getting enough food on the table to keep all of those children alive, even assuming they are both present. Sure, abstinence from sex is an option, at least under reasonable conditions of freedom. But for many people, having a sex life results in a happier life, and happiness is an important human good.

In general, our natural inclinations do not always map to moral obligations, or even to what is good for us – which is of course a matter different from that of moral obligations, but even if we leave that aside.

Of course, knowing human psychology can be useful as a tool to better know and understand what is good for us, and even our moral obligations. However, not all normal human psychological inclinations are part of our moral capacities, or lead to morally better behavior, and not all normal human psychological inclinations are something it is a good idea to follow in present-day circumstances, at least for many humans who have no defects in the relevant capacities.

To finish this section, I would like to leave aside for a moment the question of the ends of our faculties, and focus on a claim about the consequences:

Feser: So, nature’s taking its course thus seems to leave mothers and offspring pretty helpless, or at any rate it would do so if there weren’t someone ordained by nature to provide for them. But of course there is such a person, namely the father of the children.

Actually, if “nature takes its course”, mothers would in plenty of cases be pretty helpless even with the assistance of the father or fathers, given that their resources would also be too limited, and so the fathers too would be pretty helpless. So was the case in the past for many people, and so is the case for many even today.

If people generally have many children – as some of their natural inclinations would lead them to - , plenty of children would still – and do still – grow in situations of dire poverty, hunger, lack of education, etc., even in those cases in which the father of all or some of the children is around. In fact, as a predictable consequence of the lack of use of contraceptives – because they’re not available or by choice - and a lot of sex – letting “Nature” take its course -, millions of children are born who will routinely fail to achieve the higher goods Feser talks about, and will suffer from hunger, all sorts of infectious disease, poor health in general and in particular poor mental development, and so on. With more contraception, of course there would be fewer children, with a good chance of living on average much happier, healthier lives, with more education and far better chances of flourishing. Many people want that – for example - for the children they already have - rather than having more children and making life worse or even much worse for the ones they have –, and also want to have a good sex life with their partner or spouse – rather than abstinence and less happiness.

In short, it seems pretty clear to me that contraception can be and is a good thing for many people, and this is so not only in metaphysically possible but unrealistic scenarios, but in millions of ordinary cases.

6. Conclusion.

In his defense of the perverted faculty argument, Feser makes claims that imply a radical departure from common sense judgments of morality and rationality. Additionally, he bases some of his key points on a metaphysical theory that, given the available evidence and reason, is very probably false, even on some of its most crucial features.


[A] I would call it a function, not a ‘natural end’, and of course I disagree with Aristotelian-Thomistic metaphysics. But I don’t want to focus on the terminology at this point.

[B] There is no stipulation that the AI is conscious, or in any way has a mind, or that it does not. Thus, neither the impossibility of conscious superintelligent (i. e. with superhuman intelligence) AI nor the impossibility of unconscious superintelligent AI would, on its own, block this scenario.

Moreover, if one does not believe that all AI with superhuman intelligence is metaphysically possible (but why not?), one may instead assume it’s advanced aliens from another planet who invaded the Earth and control the robots. If one believes that that is also metaphysically impossible (but why?), one may instead stipulate the ruler is an evil, sadistic, psychopathic human being, who has robots and accomplices to enforce his evil rules. One way or another, scenarios like these in the relevant sense – i. e., regarding the rationality of smoking “to excess” - are surely metaphysically possible.

[C] If one does not believe that those methods are metaphysically possible (but why not?), one may instead stipulate that the AI uses artificial insemination and forced impregnation. It takes eggs from human females – by force -, and sperm cells from human males who either cooperate – some or all under threat -, or do not cooperate but the sperm is extracted by force, to make more human embryos, which it forcibly implants into women, etc. There are plenty of metaphysically possible variants one can choose from.

[D] It might be suggested that, for some reason, the images on the screen do not qualify as triangles and somehow paper is needed. But that would go against our common usages of the words. Of course, triangles on a computer screen are still triangles, and can be good or bad triangles, etc. But in any case, one can stipulate that the triangles are printed – which is metaphysically possible for sure.

[E] Someone might suggest that while there is a species S(1) of which Cecil is a member, it does not match the referent of the word ‘lion’. However, this would not block my objection: Since L(n) is not a member of S(1), the argument goes through just as well, and we can just say ‘S(1)’ instead of ‘lion’ throughout.

[F] Talk of possible worlds is not essential here, or anywhere else in this essay. We can just talk about what actually happens, is the case, etc., what possibly happens, is the case, etc., and make essentially the same arguments.


[1] Feser, Edward, Neo Scolastic Essays, South Bend, Indiana. St. Augustine's Press. [2015] One can find a link at:


[3] See, for example,,

[4] Liu, S., Lorenzen, E. D., Fumagalli, M., Li, B., Harris, K., Xiong, Z., Zhou, L., Korneliussen, T. S., Somel, M., Babbitt, C., Wray, G., Li, J., He, W., Wang, Z., Fu, W., Xiang, X., Morgan, C. C., Doherty, A., O'Connell, M. J., McInerney, J. O., Born, E. W., Dalén, L., Dietz, R., Orlando, L., Sonne, C., Zhang, G., Nielsen, R., Willerslev, E., … Wang, J. (2014), Population genomics reveal recent speciation and rapid evolutionary adaptation in polar bears, Cell, 157(4), 785-94.

[5] Brand, Colin & J. White, Frances & Thompson Handler, Nancy & Hickmott, Alexana & Boose, Klaree. (2018). Adaptive functions of GG rubbing among female bonobos (Pan paniscus) at Lomako, DRC.

[6] De Waal, F. (2006). Bonobo, sex and society. Scientific American. 16. 14-21. 10.1038/scientificamerican0606-14sp.