A Book Club Ends (Knowledge, Reality, & Value)
On Econlog, Bryan Caplan just posted his concluding comments on my book, Knowledge, Reality, and Value:
Part 1: https://www.econlib.org/knowledge-reality-and-value-rejoinder-to-huemer/
Part 2: https://www.econlib.org/knowledge-reality-and-value-rejoinder-to-huemer-part-2/
These are my concluding responses (a bit long).
First, many thanks to Bryan Caplan for the extensive, thoughtful, and interesting discussions over the course of this Book Club. Thanks also to all the readers who contributed their thoughts. I found the discussion very high quality and serious.
Bryan’s comments below, preceded by “BC”, followed by my replies.
1) On consequentialism vs. deontology in the history of ethics:
BC: Why doesn’t the Ring of Gyges thought experiment count? At least one common interpretation of the debate is that we should practice justice as an end in itself, not because of its consequences. …
As I read it, Socrates is arguing over what is good, not over whether one should maximize the good. Note that holding that justice is intrinsically good doesn’t challenge consequentialism (though it does challenge utilitarianism), because it’s consistent with holding that you should maximize the good. To see what Socrates/Plato thought about consequentialism, you’d have to look at whether they thought it was okay to commit one injustice to prevent two similar injustices.
2) On the probability of the BIV hypothesis:
BC: Lower, definitely. But why “low”? In fiction, after all, the main use of the Brain in a Vat is to simulate reality. I still say that everything hinges on the high prior probability of the Real World story relative to all alternatives.
Let’s distinguish two skeptical hypotheses:
G-BIV (generic BIV hypothesis) posits only that our experiences are explained by scientists stimulating a brain.
S-BIV (the specific, stipulative BIV hypothesis) adds to G-BIV the assumption that the scientists stimulate the brain in exactly the same way that a typical brain in the real world would be stimulated.
If G-BIV and RWH (the real-world hypothesis) have comparable initial probabilities, then G-BIV is going to have a much lower posterior probability in the light of our evidence. Equivalently: S-BIV has a much lower prior probability. This is because in the absence of evidence, we start from the principle of indifference, and the vast majority of the ways of stimulating a BIV would not result in a perfect replication of normal, real-world experiences.
Perhaps a better way to think of it is to consider hypotheses about the capabilities and motives of the scientists. Only a tiny range of possible capabilities and motives, out of all the possible ones that a troop of BIV-owning scientists could have, result in their producing a perfect simulation of a normal, average life.
Regarding fiction: The reason why fictional stories (e.g., The Matrix) invoke perfect simulations of normal life is not that that is likely (even conditional on having a BIV or similar apparatus), but simply that that makes for a good story.
3) About recognizing shapes made by intelligent agents:
BC: Not heart shapes specifically. I’m thinking of the independent knowledge that so far, things that look like carvings have always been man-made. If that sounds circular, I say it’s just the problem of induction all over again.
“Looks like a carving” sounds to me like it presupposes a sense of what kinds of things are likely to have been created by intelligence, vs. what things are likely natural – that sense is what the Argument from Design proponents are invoking, which I took you (Bryan) to be rejecting.
Not sure what you’re suggesting about the problem of induction. Maybe that circularity is okay because a circular response to the problem of induction is okay? I would deny that circularity is okay, even with respect to the problem of induction. See my paper on induction, which gives a non-circular solution: http://www.owl232.net/papers/explanation.pdf.
BC: How about a sphere? It’s awesomely “ordered.” Perfectly symmetrical…
It is ordered, but it’s a simple order, so it’s not hard to imagine simple, unintelligent mechanisms that bring it about. Like the force of gravity, or surface tension.
Here is a better example of something that looks artificial (but isn’t): I recently learned that there is a giant hexagon on Saturn, made of atmospheric gases. The hexagonal shape makes it much weirder than if it were simply round. Scientists aren’t sure why this exists, but there are some theories.
It’s pretty weird, but it’s still a much simpler order than, say, a living organism.
4) About setting the parameters of the universe:
BC: Well, we’ve looked around the universe a lot, and never seen anyone “setting its parameters.” We haven’t even found a securely locked control panel with a “parameters of the universe” label. And the very idea sounds totally fanciful, so we should assign it an ultra-low prior probability.
About the prior probability: I think the “sounds totally fanciful” reaction is specific to atheists. The vast majority of people have thought that there were supernatural beings with all kinds of awesome powers. Most today believe in a creator of the universe. I don’t have a lot of data on their psychology, but I suspect that most find it intuitively plausible that someone created the universe.
About the evidence: If there was such a creator, I don’t think we’d particularly expect to see that person doing His work, since we wouldn’t be around until after He created the universe. I don’t think we’d expect to find a control panel either, even if we looked around a lot.
I also don’t think we’ve really looked around the universe a lot. We’ve looked around the Earth a lot (and even then, mostly just the surface), but the observable universe is about 400,000,000,000,000,000,000,000,000,000,000,000,000,000 times larger than the Earth, to say nothing of the unobservable parts of the universe, if such there be. We’ve also been around for perhaps one 46 thousandth of the age of the observable universe. I think there’s lots that we don’t know, especially about questions like the origin of the universe.
5) On the “Made by God” example:
BC: My bad, I missed the latter stipulation. But then it’s practically circular, because anything that’s a “consequence of the laws of nature” is by definition not human-made (or even alien-made).
I don’t see a circularity problem. The example is just an illustration of a point about the Fine Tuning Argument. The example shows that you could have cases in which you should conclude that (aspects of) the laws of nature were designed by an intelligent (non-human) being. It’s okay that the example stipulates that the thing we observe is something we know humans couldn’t do, because that is also true of the evidence in the actual Fine Tuning argument.
I gave this example because some of the objections to the FTA actually imply that even in the “Made by God” example, even with the stipulation that the observed evidence is a consequence of the laws of nature, you still would have no reason to believe in God, or any other form of intelligent designer. This is a reductio of those objections.
6) On the improbability of a life-supporting universe:
BC: It’s only “extremely improbable” if you think the parameters of the universe could have been “set differently.” And as I said, that seems totally fanciful to me. There’s no sign that a cosmic control panel exists.
This is one of the objections that implies that, even in the “Made by God” example, you still have no reason at all for believing in any form of intelligent design; you should just shrug your shoulders and say, “That’s just the way it is.” Any such objection must be mistaken.
In this case, the mistake is the assumption that, in order to assign (non-extreme) probabilities to propositions about the laws of nature, one has to first be convinced that the laws were set by someone (or something?), or can be changed. One need not assume either of those things merely to have a non-trivial probability distribution.
7) On what is “fantastical”:
BC: What distinguishes all these cases? (a) We only hear about the fantastical stuff from unreliable sources or in fiction; the other stuff we either see first-hand or hear about from reliable sources. (b) The fantastical stuff appeals to the human emotion of wonder; the other stuff is boring by comparison.
This would be an apt response to arguments based on testimony about miracles. The unreliability of the sources would be a reason not to accept those arguments. But I don’t think it’s a reason to deny the existence of God; rather, it means that the testimony is merely, at worst, irrelevant.
Regarding the emotion of wonder, I think that could be a reason for discounting some stories that have been handed down – they could have been transmitted because of the desire for amazing stories, even if they were not true. It might also be a reason for discounting intuitions of people who have a particular passion for wonder, because that passion could be biasing them.
However, none of this is actual evidence against the existence of an intelligent designer; it just means one should discount some putative reasons in favor. I also don’t see how it shows that the prior probability is low. As a conceptual matter, if you’re citing evidence, then you’re talking about a posterior probability, not a prior probability.
About the Fine Tuning Argument: fortunately, it doesn’t rely on unreliable testimony. The empirical premises come from modern astrophysics. The fact that some other people who are unreliable also support a somewhat similar conclusion for very different reasons doesn’t count as evidence against this argument being correct.
I would say, by the way, that whatever the origin of the universe is, I think it is bound to be something strange and amazing. The standard Big Bang theory is amazing already, as are many things in modern physics.
8) Do incentives affect your degree of free will?
BC: By the logic of your alcoholic story, it seems like the more money we offer the condo owner, the “less free” he is to turn it down. That seems bizarre to me. I say you are fully free to reject a billion-dollar offer on your condo. And the alcoholic is totally free not to drink.
The alcoholic has a psychological compulsion or other disorder, which is normally not present in a condo owner considering a purchase offer. The condo owner, in the normal case, also won’t have any of the other sorts of psychological conflicts that would reduce freedom. So for the condo-owner, it’s plausible that he’s fully free regardless of the incentives.
To make the alcoholic analogous to the condo-owner, we’d have to add some stuff to the condo story. Like he’s got some kind of psychological compulsion to accept all purchase offers. Or maybe he’s got a conflict between immediate inclinations and his rational judgment. Or something like that. In that case, I think it is plausible that he’s going to be less free for some larger offers than for a ridiculously low offer. But this example is now so similar to the alcoholic case that I don’t expect it to help resolve the issue for anyone – you’ll either intuitively agree with me about both cases, or intuitively disagree about both cases.
9) On consent and my paradox for moderate deontology:
BC: How is this different from a person who foolishly refuses to consent to a vaccination, even though he admits that the benefit of the vaccine greatly exceeds the pain of the needle? As you explain in The Problem of Political Authority, we have no right to benefit him given his explicit refusal to consent.
I’m on board with the idea that it’s wrong to harm one person without his consent in order to benefit someone else. But I balk at the idea that it’s wrong to stop harming a person without his consent, in order to benefit someone else.
In particular, if someone is being unjustly tortured (which they also didn’t consent to!), I think you can just turn off the torture device. It would be weird to think that it’s wrong to stop torturing someone unless they consent to the non-torture. Similarly, if the best you can do is to turn down the torture device, you can also do that without consent.
Given that, in my example, no consent is required to reduce both people’s torture. But consent would be required (on common deontological views) to increase one person’s torture while decreasing the other’s. So that’s how you can get a case where A and B are each impermissible, but (A&B) is permissible.
10) On an immoral pool project
BC: … the mice lose their home due to your pool construction, then slowly die of starvation and exposure while they hunt for another home. … And unless you’re doing a “No True Vegan” thing, I really doubt that even many vegans would actually consider this a morally strong reason not to build a pool.
You must not know very many vegans. I think almost no one initially sympathetic to veganism would find your example persuasive. First, I think you’re underestimating how horrible factory farms are. After leaving their current home, the mice would be in a pretty normal position for wild animals. Sometimes, indeed, wild animals starve; nevertheless, factory farm life is, in general, much worse than a normal situation in the wild. At least, that’s what all or nearly all vegans believe.
We could debate how bad life in the wild is, but that’s unproductive. This argumentative strategy in general can’t help, because either you give an example that vegans will think is less bad than factory farming, or you give an example that they see as just as bad as factory farming. In the first case, the example will be irrelevant. In the second case, pretty much every vegan will have exactly the same reaction to your example that they have to factory farming.
Here’s an analogy to explain how I and other ethical vegetarians will see your arguments. Imagine arguing with a Trump-supporter about immigration. You claim to be concerned about the welfare of potential migrants. The Trumpster can’t believe that you care about foreigners, and he tries to prove to you that you don’t. So he gives thought experiments like this: “Obviously, you’d agree that it was fine to drop atomic bombs on Hiroshima and Nagasaki in World War II. But this was more harmful to those foreigners than merely denying people entry to the U.S. today.”
Why would that not be a productive line of discussion for the immigration debate? First, it’s not obvious on its face that the atomic bombing and the immigration restrictions are analogous actions. You’d have to have a lengthy debate about that. But that debate would be a time-waster because, second: if the atomic bombing was analogous, then of course you would also be opposed to the atomic bombing. There is no way that thinking about the atomic bombing will help to resolve the immigration issue.
The best I can make of it is that you (Bryan) can’t believe that anyone really cares about other species; you think everyone is like you on the fundamental level, but maybe some just got confused when thinking about a few kinds of cases. Similarly, the hypothetical Trumpster assumes everyone is like himself. Both are mistaken. Some people are in fact different from the Trumpster and the meat-eater. Some people do in fact care.
11) Does potential/species intelligence make your pain bad?
BC: Does it even slightly reduce your confidence to learn that only 10% of respondents to this survey say that I’m definitely wrong? [In reference to BC's twitter survey. --mh]
Does it reduce your confidence to learn that only 22% of respondents think you’re definitely right?
This survey does not significantly change my opinions. I note a few points:
This isn’t a random sample; respondents are people who follow you on Twitter, which means they are disproportionately likely to agree with you.
I already knew that the overwhelming majority of people are meat-eaters. That’s most likely driving their response. Respondents may have said that potential intelligence is morally significant because they know that they’re going to have to say this in order to defend their meat eating. R’s may even have actually read your earlier post. In other words, they’re rationalizing.
The survey gives no information about why (other than the rationalization theory) 52% of R’s picked “probably”. I can’t right now think of any other reasons why someone would think that.
I’m also not sure that R’s understood the sentence. It’s pretty complex. Moreover, when asked to judge the goodness or badness of something, people generally start thinking of possible instrumental reasons why the thing would be good or bad. They almost never think that you’re asking about intrinsic value. But the relevant interpretation of the quote would be that a given painful experience is more intrinsically bad, all other things being equal, if the subject is a type of being that would normally become intelligent later. Besides the intrinsic/instrumental distinction, one would have to emphasize that the being in question is not in fact intelligent, and will never be intelligent; then ask, with that understood, does the intelligence of other members of its species make this individual’s pain worse?
Once properly understood, the proposition strikes me as absurd. To me, it’s like doing a survey about whether the shortest path between two points is a curve. If 52% of people said “Probably yes”, I would suspect that R’s didn’t understand the question or made some other error. But even if I couldn’t figure out why they said what they did, I would not give significant credence to the claim that curves are shorter than straight lines, since I can just see that that’s false.
For a closer analogy, the proposition strikes me as sort of like someone saying that painting A is more beautiful than painting B, because there are some other paintings that cost a lot of money that were painted by a member of the same race as the painter who painted A. There’s no further explanation; the person claims this is just a self-evident axiom of aesthetics. I would have ~0 credence in that.
12) On plant interests:
BC: Imagine a conversation between you and someone who believes in the rights of plants. You tell him, “Plants don’t feel pain,” and he says, “That’s an arbitrary difference. Plants are still alive. They have interests, and we shouldn’t do immense harm to their interests to slightly advance our own.” You probably consider this an obtuse position – and I agree.
There is in fact an argument in the ethics literature like that – that all living things have interests (something can be “bad for a plant”, “good for the plant”, etc.), and interests are what really matters.
I would try two lines of argument against this. One line of argument would be to compare a single cell in your body to a single-celled organism. They are intrinsically very similar. If the person agrees that the cell in your body doesn’t merit intrinsic moral consideration, then it’s plausible to generalize to single-celled organisms.
The other line of argument would start by thinking about what counts as in your interests or against your interests (for us people). One can argue plausibly that this is determined by one’s mental states (such as enjoyment/suffering, or desires), rather than by pure biological functioning. E.g., it’s good to frustrate normal biological functioning when this clearly satisfies a person’s desires, causes overall happiness, etc.
About interests: I think talk about what is “good for” plants is analogous to talk about what is good for your car (like frequent oil changes). It’s a non-moral use of “good”.
Having said that, of course if someone has sufficiently strong and different intuitions, then one can’t convince them.
13) On the Argument from Conscience:
BC: Attacks on sincerity seem more futile, but how about a direct appeal to sincerity, a la my Argument from Conscience?
(The link goes to a post about how Bill Dickens claims to be a utilitarian, and is generally highly conscientious, yet he fails to come close to maximizing utility in his actions.)
I don’t find this persuasive, even though I’m not a utilitarian. The overwhelming majority of people have strong emotional motivations apart from their ethical beliefs. There is good evidence that these emotions are the main motivators for most seemingly ethical behavior. One particularly strong source of motivation has to do with the practices and norms of one’s own society, which have a powerful effect on people’s emotions and desires. But there are probably other emotional mechanisms, possibly genetically programmed, which lead people to act in accordance with deontological rules.
With that in mind, it’s entirely plausible that Bill Dickens’ emotions and desires would mostly conform to the non-utilitarian norms of his society in most circumstances, despite his belief in utilitarianism. It is then plausible that his actions would mostly be non-utilitarian. (I say “mostly”, because I expect utilitarian beliefs to have some influence. For instance, they probably make him give more to charity than most non-utilitarians do.)
It’s not weird that Dickens’ behavior appears highly conscientious by conventional standards. That just means that he is naturally high in the sort of emotional dispositions that produce that type of behavior, esp., cooperative & respectful behavior.
What is the alternative hypothesis? That he’s lying – he doesn’t think utilitarianism is true, and he’s been playing a weird hoax for the past 25 years?
Or perhaps he thinks that he believes utilitarianism, yet he doesn’t believe it? Try to fill that in a little more. When he thinks about arguments for utilitarianism, what happens? Those arguments seem right to him, or they don’t? If they do, then it’s kind of bizarre that these arguments didn’t convince him of utilitarianism (which they support) but merely convinced him of the false proposition that he believes utilitarianism (which they don’t support). If the arguments don’t seem right to him, then it’s kind of weird that he would affirm them, and that he would think he agrees with them.
Another hypothesis might be that there are two different kinds of beliefs, or belief-like states. Maybe he consciously believes utilitarianism, but unconsciously believes deontology? (The reverse wouldn’t make sense.) But then, how would this give you an argument against utilitarianism? Suppose Dickens has an unconscious belief that conflicts with his conscious belief. What’s the argument that the unconscious belief is more likely to be true? Surely in most such cases, the conscious belief is more likely to be correct.
. . .
I’m going to conclude with this YouTube video link. This is a five-year-old child who recently learned that meat is made from animals: https://www.youtube.com/watch?v=5Npv2Mpbd3w
I include this because it is extremely difficult to doubt this child’s sincerity. She hasn’t had time to come up with the sorts of rationalizations that adults come up with, nor is she repeating propaganda from other people. It’s just the natural reaction of an innocent, compassionate person.