= 27d567bc2d48d9f40e9ccc20ea370b15)
Bugmaster comments on White Lies - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (893)
I understand your point, but I still do not understand the motivation behind it. Are you following some sort of a consequentialist morality, or a deontological one that states "overt lies are bad, lies of omission are fine", or something else ?
As I see it, if a friend asks you "do you like this painting ?" and you reply with "the background color is nice", the top most likely outcomes are:
Similarly, if your friend asks you to review his paintings and you refuse, you'd better have a good reason for refusal (i.e., the truth or some white lie); otherwise, anyone of average intelligence will interpret your response as saying "I hate your paintings but I won't tell you about it".
None of what I wrote above matters if you only care about following prescribed rules, as opposed to caring about the effects your actions have on people. Perhaps this is the case ? If so, what are the rules, and how did you come by them ?
I'm Less Wrong's token deontologist. I thought most people around here knew that. I wrote this article about it and my personal brand of deontology is detailed in this comment.
Sorry, I did not, in fact, know that; and most people here are consequentialists, so I assumed you were one as well. I'd skimmed your post on deontology that you linked to earlier, but I did not understand that it was meant to represent your actual position (as opposed to merely being educational).
As I said above, if your moral system simply has a rule that states "lying is bad except by omission", or something similar, then none of my points are valid, so you are right and I was wrong, my apologies.
That said, personally, I don't think that deontology makes any sense except possibly as a set of heuristics for some other moral system. That's a different line of debate however, and I won't push it on you (unless you are actually interested in pursuing it).
I'm willing to answer questions about it if you're curious, but since I dropped out of grad school I haven't devoted much time to refining either my ethical theory or my ability to explain it so the old article will probably be just about as good. I get enough debating in just from hanging out with consequentialists all the time :P
To expand on what blacktrance said:
As I understand it, deontological systems are, at the core, based on lists of immutable rules.
Where do the rules come from ? For example, one rule that comes up pretty often is something like, "people have inalienable rights, especially the right to A, B and C". How do you know that people have rights; what makes those rights inalienable; and what makes you so sure that A, B and C are on the list, whereas X, Y and Z are not ?
And consequentialist systems are, at the core, based on an immutable utility function.
Where does this function come from?
Well, no. Utilitarian systems are based on a utility function (although I'm not aware of any requirement that it be immutable... actually, what do you mean by "immutable", exactly?). Consequentialist systems don't have to be utilitarian.
Even so, the origin of a utility function is not that mysterious. If your preferences adhere to the von Neumann-Morgenstern axioms, then you can construct a utility function (up to positive affine transformation, as I understand it) from your preferences. In general, the idea is that we have some existing values or preferences, and we somehow assign utility values to things ("things": events? world states? outcomes? something) by deriving them from our existing preferences/values. It's not a trivial process, by any means, but ultimately the source here is the contents of our own brains.
The problem is that most (all?) people's preferences don't.
That's a valid question, and, admittedly, there's no good answer that I'm aware of. One might say that, ultimately, the function can be derived from some basic principle like "seek pleasure, avoid pain", but there's no objective reason why anyone should follow that principle, as opposed to, say, "seek paperclips, avoid non-paperclips".
I will grant you that both consequentialism and deontology are based on some a priori assumptions; however, I would argue that the fact that consequentialism is based on fewer such assumptions, as well as its flexibility in the face of new evidence, make consequentialism a more efficient moral system -- given that we humans are agents who are reasoning under uncertainty using a comparatively limited amount of data.
I would argue that this "fact" is not in fact true, or at least not obvious. It's not even clear to me what the content of that claim is supposed to be. If you mean that it takes fewer bits to encode a utility function then a collection of maxims, then this will obviously depend on which utility function or set of maxims is used, also as Eliezer points out here this is a really really bad way to compare moral systems.
Huh? If you're claim is that consequentialism is more flexible in the face of new evidence then deontology, you're going to have to provide some justification for it (as well as specifying precisely what you mean). As I see it, both are inflexible in the sense that ideal agents of either type are incapable of changing their utility function or set of maxims in the face of any evidence, and flexible in the sense that they can use evidence to determine how to pursue their maxims or maximize their utility function, and also in the sense that actual humans will in fact update their maxims or utility function in the face of evidence.
Not necessarily. You are correct in saying that any given arbitrary utility function can be a lot more complex than any given arbitrary set of rules; so strictly speaking I was wrong. However, in practice, we are not dealing with arbitrary functions or rules; we are dealing with limited subsets of functions/rules which are capable of sustaining a human society similar to ours in at least some way. Of course, other functions and other rules can exist, but IMO a moral system that effectively commands its followers to e.g. kill themselves ASAP is not very interesting.
Given this restriction, I believe that consequentialist moral systems which satisfy it will require fewer arbitrary assumptions, in part due to the following:
Changing the maxims is exactly the problem. Given that deontological maxims are essentially arbitrary; and given that the space of all possible human behaviors is quite large; it is already pretty difficult to construct a set of maxims that will account for all relevant behaviors that are currently possible. Of course, you could always create a maxim that amounts to saying, "maximize this specific utility function", but then you're just reducing deontology to consequentialism.
In addition, though, as humans acquire more knowledge of and more power over their environment, the set of possible behaviors keeps changing (usually, by increasing in size). This presents a problem for the deontologist, who has to invent new maxims just to keep up (as well as convincing others to use the new maxims which, as you recall, are entirely arbitrary), as well as to possibly revise existing maxims (ditto). The consequentialist, on the other hand, can apply his existing utility function to the new behaviors, or plug the new data into it, in order to come up with a reasonable re-evaluation of the morality (or lack thereof) of each behavior.
To use a trivial example, at some point in human history, it became possible to digitally copy musical performances without paying any money to the original authors. The deontologists are debating to this very day whether such actions count as "theft" or not, "theft" being a prohibited behavior under one specific maxim. Unfortunately, this new behavior doesn't quite fit the parameters of the original maxim (which was invented before information technology became widespread), hence the debates. But if we dispense with the labels, and attempt to evaluate whether digital music copying ultimately causes more harm than good (or vice versa), then we can at least make some progress.
Upvoted for this, and the excellent (if trivial) digital copying example.
I will add that progress in such cases may also sometimes be made by attempting to discern just what are the origins of our moral intuitions about the wrongness of theft, seeing if those intuitions may be decomposed, and whether they may be reconstructed to yield some concepts that are appropriate to the digital realm. (I've got an essay where I attempt to do just that for software piracy, which I may post online at some point...)
The general principle here is that since the basis of our consequentialism systems is the contents of our brains, we can refer to the source material for guidance (or attempt to, anyway). With deontology, since it doesn't reduce to anything, that move is not open to us. (I think. I remain unclear about where the rules in a deontological system come from.)
I think that rights drop naturally out of personhood. Being a person is to be the kind of thing that has rights (and the obligation to respect same). The rights are slightly alienable via forfeiture or waiver, though.
I don't quite understand what you mean. Even if we can agree on what "personhood" means (and I've argued extensively with people on the topic, so it's possible that we won't agree), what does it mean for a right to "drop out naturally" out personhood ? I don't understand this process at all, nor do I understand the epistemology -- how do you determine exactly which rights "drop out naturally", and which ones do not ?
To use a trivial example, most deontologists would probably agree that something like "the right to not be arbitrarily killed by another person" should be on the list of rights that each person has. Most deontologists would probably also agree that something like "the right to possess three violet-blue glass marbles, each exactly 1cm in diameter" should not be on the list. But why ?
I think Alicorn's answer concerned the ontological status of rights, not the epistemology thereof.
Understood, but I would like to understand both...
Likewise. For what it's worth, though, I don't actually think there is a good answer to the epistemological questions you asked; that's one of the reasons I favor consequentialism rather than deontology. Of course, I imagine Alicorn's views on the matter differ, so I, too, would like to see her answer (or that of any other deontologist who cares to respond).
In another branch of this thread I've just walked through an assessment of whether a provided example contained a rights violation. Does that help?
EDIT: Sorry, turns out you already answered my question. Here are some replacement questions.
You've said that you will do nothing, rather than violate a right in order to prevent other rights being violated. Yet you also say that people attempting to violate rights waive their rights not to be stopped. Is this rule designed for the purpose of allowing you to violate people's rights in order to protect others? That seems unfair to people in situations where there's no clearly identifiable moustache-twirling villain.
You have also said that people can waive any of their rights - for example, people waive their right not to have sex in order to have sex, and people waive their right not to be murdered in order to commit suicide. Doesn't this deny the existence of rape within marriage? Isn't it, in fact, the exact argument that was used to oppose laws prohibiting rape within marriage? This seems worrying. (Obviously, there are other, similar cases that can be constructed, but this one is a major problem.)
Finally, you mention that some actions which do not violate rights are nonetheless "being a dick", and you will act to prevent and punish these acts in order to discourage them. Doesn't this imply that there are additional aspects to morality not contained by "rights"? Do you act as a Standard-LessWrong-Consequentialist-Utilitarian™ with regards to Not Being A Dick?
I wish everyone in this thread would be more careful about using the word "right". If you are trying to violate somebody's rights, you don't have "a right not to be stopped". You have your perfectly normal complement of rights, and some of them are getting in the way of protecting someone else's rights, so, since you're the active party, your (contextually relevant) rights are suspended. They remain in effect out of that context (if you are coming at me with a knife I may violently prevent you from being a threat to me; I may not then take your wallet and run off cackling; I may not, ten years later, visit you in prison and inform you that your mother is dead when she is not; etc.).
That's a good question, but the answer is no. A marriage does not constitute a promise to be permanently sexually available. You could opt to issue standing permission, and I gather this was customary and expected in historical marriages, but you can revoke it at any time; your rights are yours and you may assert them at will. I don't object to people granting each other standing permission to do things and sticking with it if that's how they prefer to conduct themselves, but morally speaking the option to refuse remains open.
No. There's morality, and then there's all the many things that are not morality. Consequentialists (theoretically, anyway) assign value to everything and add it all up according to the same arithmetic - with whatever epicycles they need not to rob banks and kidnap medical test subjects - but that's not what I'm doing. Morality limits behavior in certain basic ways. You can be a huge dick and technically not do anything morally wrong. (And people can get back at you all kinds of ways, and not technically do anything morally wrong! It's not a fun way to live and I don't really recommend it.)
No. Actually, you could probably call me sort of virtuist with respect to dickishness. I am sort of Standard-LessWrong-Consequentialist-Utilitarian™ with respect to prudence, which is a whole 'nother thing.
Well, sure. I did read your explanation(s). I was assuming the worst-case scenario for the hypothetical, where you have to violate someone's rights in order to protect others. For example, the classic lying-to-the-nazis-about-the-jews scenario.
Not anymore, no. Because we changed the rules. Because of all the rapes.
I ... see, that seems consistent. I assume you can waive the right to abolish an agreement at will, too? That's the foundation of contract law, but I don't want to assume.
Indeed, and that's what I'm asking about. What is this "don't be a dick" function, and what place does it hold?
Huh. Shame.
... what's "prudence" in your nomenclature? I haven't seen the term as you use it.
If you are trying to violate someone's rights, then your contextually relevant rights are forfeited. For example, the Nazi has forfeited the right not to be lied to.
I'm not sure I understand.
Well, some people are motivated to avoid being dicks, and might value information about how to do it. It's not very ontologically special.
To me, it looks like consequentialists do prudence exclusively, and name it "morality", instead of actually doing any morality. Prudence is arranging for things to be how you would like them to be.
Yes, I know. Hence my question:
You see?
Can I waive my right to un-waive a right? For example, if I waive my right not to work for someone, can I also waive my right to change my mind? As in, a standard contract?
I would assume so, but hey, it can't hurt to ask.
Are you, personally, motivated to discourage dickish behaviour? You are, right? You mentioned you would sue someone for being a dick to you, if it were illegal, even if it was perfectly moral for them to do so.
... as long as "how you would like them to be" doesn't violate any rights. I think I see what you mean by that.
... ah, and "how you would like things to be" includes "no dickishness", right?
I do not see; can you start that line of inquiry over completely for me?
I haven't actually thought about this before, but my instinct is no. Although if you arrange for it to be really hard to communicate a change of mind, someone who went by their last communication with you might not be doing anything wrong, just making a sincere mistake.
I try to minimize it from and around myself. I am not on a Global Dickishness Reduction Campaign of any kind. Maybe I should have said not that I'm not a consequentialist about it, but rather that I'm neither agent- nor recipient-neutral about it? How I would like things to be certainly refers to dickishness.
Utilitarians do this. Consequentialists don't necessarily.
Not only utilitarians, but von Neumann-Morgenstern rational agents.
Well, sure. That wasn't meant to be an exhaustive list; I only meant to highlight that consequentialism does not necessarily do this.
But yes, you're quite right.
"No. There's morality, and then there's all the many things that are not morality."
Is this only a linguistic argument about what to call morality? With ,e.g. , virtue ethics claiming that all areas of life are part of morality, since ethics is about human excellence, and your claim that ethics only has to do with obligations and rights? Is there a reason you prefer to limit the domain of morality? Is there a concept you think gets lost when all of life is included in ethics (in virtue ethics or utilitarianism)?
Also, could you clarify the idea of obligations, are then any obligations which don't emanate from the rights of another person? Are there any obligations which emerge inherently from a person's humanity and are therefore not waivable?
You could re-name everything, but if you renamed my deontological rules "fleeb", I would go on considering fleeb to be ontologically distinct in important ways from things that are not fleeb. I'm pretty sure it's not just linguistic.
Because there's already a perfectly good vocabulary for the ontologically distinct non-fleeb things that people are motivated to act towards - "prudence", "axiology".
Unassailable priority. People start looking at very large numbers and nodding to themselves and deciding that these very large numbers mean that if they take a thought experiment as a given they have to commit atrocities.
Yes; I have a secondary rule which for lack of better terminology I call "the principle of needless destruction". It states that you shouldn't go around wrecking stuff for no reason or insufficient reason, with the exact thresholds as yet undefined.
"Humanity" is the wrong word; I apply my ethics across the board to all persons regardless of species. I'm not sure I understand the question even if I substitute "personhood".
Lets take truth telling as an example. What is the difference between saying that there is an obligation to tell the truth, or honesty being a virtue or that telling the truth is a terminal value which we must maximize in a consequentialist type equation. Won't the different frameworks be mutually supportive since obligation will create a terminal value, virtue ethics will show how to incorporate that into your personality and consequentialism will say that we must be prudent in attaining it? Similarly prudence is a virtue which we must be consequentialist to attain and which is useful in living up to our deontological obligations. and justice is a virtue which emanates from the obligation not to steal and not to harm other people and therefore we must consider the consequences of our actions so that we don't end up in a situation where we will act unjust.
I think I am misunderstanding something in your position, since it seems to me that you don't seem to disagree with consequentialism in that we need to calculate, but rather in what the terminal values are (with utilitarianism saying utility is the only terminal value and you saying hat there are numerous (such as not lying , not stealing not being destructive etc.))
By obligations which emerge from a person's personhood which are not waivable, I mean that they emerge from the self and not in relation to another's rights and therefore can not be waived. To take an example (which I know you do not consider an obligation, but will serve to illustrate the class since many people have this belief) A person has an obligation to live out their life as a result of their personhood and therefore is not allowed to commit suicide since that would be unjust to the self (or nature or god or whatever)
The first thing says you must not lie. The second thing says you must not lie because it signifies or causes defects in your character. The third thing says you must not lie unless there is a compensatory amount of something else encouraging you to lie. The systems really don't fuse this prettily unless you badly misunderstand at least two of them, I'm afraid. (They can cooperate at different levels and human agents can switch around between implementing each of them, but on a theoretical level I don't think this works.)
Absolutely not. Did you read Deontology for Consequentialists?
I still don't know what you mean by "emerge from the self", but if I understand the class of thing you're pointing out with the suicide example, I don't think I have any of those.
You wrote that you believe that persons have rights. How do you determine what rights they have?
IANAlicorn, but, since I have the same belief, I'll give it a shot. My imperfect introspection tells me that, since the world where people don't have rights would quickly become unfair and full of suffering (and this has been repeatedly experimentally tested), I want to live in a world where I, my family or someone I can identify with would have less of a chance of being treated unfairly and made to suffer needlessly. Pretending that people have "unalienable rights" goes a long way toward that goal, so I want to believe it and I want everyone else to believe it, too. To dig deeper, I am forced to examine the sources for my desire for fairness and the origins of my empathy (imperfect though it is), and the available literature points to the mix of genetics and upbringing.
That sounds like rule utilitarianism, or a rule utilitarianism-like consequentialism, not like a deontological justification for human rights.
I suppose you are right. However, if you skip the introspection part, "people have rights" makes sense in most cases without having to worry about utilities. It's the edge cases, like the trolley problem, which require deeper analysis.
I agree, but that's all basically consequentialist.
A decent justification, but not very deontological. What I was curious about is how Alicorn determines what rights exist purely deontologically, without reference to consequences.
Since I'm no longer maybe going to write a thesis on it, mostly I don't work on this a lot. Not lying, not stealing, and not attacking people does pretty good for everyday. There's sort of an informal checklist when I think something might be a right - the rights have to be reasonably consistent with each other, they're overwhelmingly negative rights and not positive ones, simpler ones are better, etc. This would be easier with a sample maybe-a-right but I haven't examined any of those recently.
If I may offer one —
Suppose that I am photographed on the street outside a place that has a bad reputation (with some people). The photographer might publish the photo, which could lead viewers to believe bad things of me.
One acquaintance of mine, M, claims that I have a right to forbid the photographer from publishing this photo; I have the right to control publicity about me or the use of my image, even though the picture was taken in public.
Another acquaintance, B, claims that the photographer has a freedom-of-speech right to publish it, so long as they do not explicitly say anything false about me. B believes that it would be nice of the photographer to ask my permission, but that I do not have a right to this niceness.
Still another acquaintance, R, says that it depends on who I am: if I am the mayor of Toronto, I have no right to control photos of me, since my actions are of public interest; but if I am a privately employed engineer of no public reputation, then I do have that right.
Okay, I'll walk through my process of apprehending and making a call on this situation. It looks like a good example, thanks for coming up with it.
The conflict here is between you, and the photographer - other persons in the story have opinions but aren't directly involved. The steps are that there was an opportunity to ask you if it was okay to photograph you (which the photographer passed over), the decision to photograph you, the opportunity to ask you if it's okay to publish it (which the photographer also passes over), and the decision currently at hand of whether to publish the photo. If there's a rights violation potential or actual, it's probably in one of those places. The statement of the problem doesn't suggest that you've committed any rights violations by happening to be in this location.
The fact that two chances to ask for consent have been passed up is suspicious. It's not a guarantee that something has gone wrong - the photographer is allowed to look at you without securing permission, for instance - but it's a red flag. In the normal course of things, people waive some of their rights when asked explicitly or by other mechanisms all the time. People waive their right to refuse sex, for instance, on a per-occasion basis in order to have non-rape sex.
You don't actually do anything in this story, except exist in a place that by stipulation you are quite entitled to be, so only the photographer might be committing a wrong here.
So the obvious candidate possibilities are that you have a right not to be photographed, or that you have the right not to have your likeness publicized, or that you have no such rights and the photographer may do as they like with the photograph provided no other wrongs (such as libel, a form of lying) are committed in so doing.
But earlier I said the photographer is allowed to look at you without permission. You're in a public place, where anyone, including photographers as an unspecial class of anyone, may walk by. The upshot of a photograph is that others, too, may look at you. Any of them could have walked by in real time and seen you without wronging you. There's no obvious mechanism by which a gap in time should change things. If one of the passersby had a photographic memory, that wouldn't change the fact that they could look at you. Similarly, the fact that people who live far away from this location might not have had a chance to show up in person and espy your presence, or the information that anything going on might be of interest, doesn't seem like it has anything to do with anything.
So it seems to me that the photographer is probably, at worst, being a dick. You do not have a right to prohibit someone from photographing you and publishing a photo of you unless something else is going on. (I feel like I should mention now that I hate this result. I don't photograph well and definitely don't like the idea of my likeness being used in any which way without my agreement. In fact, if the law allowed, I might even pursue someone who did this to me via legal routes, which - I hope obviously - are separate from ethical condemnation. In any event I'm not rigging this to suit myself.)
But supposing it were the other way around, your acquaintance R might still have a point. Assuming it's known that elected politicians are treated differently with respect to the use of their photographs, pursuing a career as a politician might constitute waiving one's right not to be photographed (if we had that right, which I concluded above we probably don't). In this counterfactual, this would apply to the mayor of Toronto, but not people who became public figures without doing anything wrong (that would be a potential rights forfeiture, particularly if they are a threat to the photograph-viewing population) or choosing to become public figures on purpose.
Ok, this is quite a detailed response and I appreciate the thought that went into writing it. However, from my perspective, it raises more questions than it answers. For example, you say things like this:
But why not ? I'm not asking this just to be a contrarian. For example, many people believe that all of us have a fundamental right to privacy; this would imply that you do, in fact, have the right to "prohibit someone from photographing you and publishing a photo of you". Presumably you disagree, but on what basis ?
Furthermore, you say that you
I don't see how that works. If you believe that a person has a right to photograph you and publish the photo without your permission; and yet you are launching a legal challenge against him for doing so; then are you not engaging in an immoral attack ? Sure, it's not a physical attack, but it's an attack nonetheless, and previously you stated (assuming I understood you correctly) that attacks on people who have violated no rights are immoral.
Can you tell me where I lost you in the description of why I do disagree?
No. This isn't how the right is framed. They don't have a right to do it; no one must behave in a way to protect their ability to do so; I don't have to stand still so they can get a clear shot, I don't have to go out in public, if I happen to own a news outlet I don't have to give them a platform, if I acquire a device that makes me look like a smudge of static to nearby cameras I can carry it around without this being morally problematic. (Perhaps unless I'm under some sort of agreement to be visible to security cameras, or something like that.) I just don't have the right to not be photographed. Remember that rights are overwhelmingly negative. The fact that someone commits no wrong by doing X does not imply that others commit a wrong by making X inconvenient or impossible.
(You're also being kind of overbroad in understanding my use of the word "attack", which was intended broadly, but not so broadly as to include "seeking legal recompense in response to an upsetting, by stipulation illegal, behavior which does not happen to constitute a violation of any intrinsic moral rights".)
These are good questions. It seems like deontologists have difficulty reconciling seeming conflicting rights.
In my main reply to the original post, I discuss some of the conflicts between truthfulness and privacy. If people have a right to not be lied to, and people also have privacy rights, then these rights could clash in some situations.
Why these rights, and not others? For example, why a right to not be murdered, instead of a right to murder one person per year? A once-a-year right to murder can be formulated as a negative right, i.e. non-interference as you murder one person.
(I agree with the listed criteria for rights, BTW.)
Quantity limitations on rights are inelegant (what does one Earth year have to do with personhood-in-general?), so there's that. Even if you frame it as "the right to go uninterfered-with during the course of one murder per year", that has a heck of a lot of fiddly bits.
It also doesn't interact with itself very well. Suppose you are trying to murder me, and I'm the first person you've tried to murder all year, and I haven't murdered anyone all year either, so I try to murder you back - am I interfering with you? It sure looks like it, but I'm just exercising my own right... "Right not to be murdered" doesn't do that sort of self-defeating.
I have other reasons to prefer the "right not to be murdered" version but they are failing to come verbally clear. Something about self-containedness that I'm having trouble explicating.
Consequentialist reasoning which seems to align fairly well with Alicorn's conclusions (at least the one about it being in some situations correct to hide the truth by being selective even when this in some sense deceives the listener, and at the same time being less correct to directly lie) are touched on here if that's useful to you.
Essentially: You don't know for sure if a person wants general encouragement/niceties or a genuine critique. One way to deal with this is to say something nice+encouraging+true which leaves room for you to switch to "okay but here is what you could do better" mode without contradicting your previous nicety if and only if they communicate clearly they want your full opinion after hearing your careful wording.