Let’s say a theory in normative ethics is subjectivist just in case, according to it, what one ought to do is determined ultimately by the attitudes of some subject or group of subjects, such as the agent himself, the agent’s society, some ideal observer, or God. So ordinary subjectivism, cultural relativism, ideal observer theory, and divine command theory are all forms of subjectivism. Here’s a dilemma for such views.
(Sorry, this post is on the longer side. But it is (I hope) easy reading – no fine points, nothing technical.)
For ease of presentation, let’s consider the view according to which the relevant group is the agent’s society. (The argument carries over to other forms of subjectivism.) Here’s the view, Societal Subjectivism (a.k.a., cultural relativism):
Societal Subjectivism (SS): an act (token) is morally right iff it is not prohibited by the moral code of the society of the agent of the act at the time of the act.
Though this definition will be a point of dispute in a moment, let’s say that the moral code of a society is the set of moral norms (or moral rules, or moral principles) accepted by the society. These are sentences that attribute a normative status to some act type. Some simple norms: “Lying is wrong,” “Adultery is wrong,” “Passive euthanasia is ok.” The norms people actually accept are probably more complicated, with exceptions built in, such as: “Killing a person is wrong, except in self-defense,” “If you make a promise, you ought to keep it, unless breaking it is required to save someone from extreme harm,” “If a person’s life’s is in danger and only you can help, and doing so isn’t very risky, then you ought to do so.”
To accept a norm is to be inclined to assent to it, to be inclined to try to act according to it, to tend to feel guilty if you don’t; to tend to feel contempt towards others who don’t, etc.
It’s no part of SS that societies ever actually write down their moral code or anything. The code is determined by the psychological dispositions of the members of the society. It would be very hard in practice to be certain exactly what some society’s moral code is, but this is an empirical question. (Let’s ignore the question, How many members of a society must accept a norm before we can say that the society accepts it?)
Here’s the dilemma. Either the theory restricts the notion of a moral code to include only basic moral norms, or it doesn’t. Either option leads to implausible consequences.
First, let me explain this restriction to basic moral norms. A moral norm is basic for a society if the society doesn’t derive it from other norms. So the norm “Adultery is wrong,” might be derived from the norm “If you make a promise, you ought to keep it, unless breaking it is required to save someone from extreme harm.” The latter norm might be basic. The norm “It’s wrong to throw away motor oil with the rest of the garbage” would very likely be non-basic. Some society might derive it from some basic norm concerning refraining from harming others; another society might derive it instead from a basic norm against damaging the environment that doesn’t derive from a norm against harming people.
Now back to the dilemma:
Why it’s implausible NOT to restrict the theory to basic moral norms. If we don’t restrict the theory to basic moral norms, then the theory generates implausible results concerning moral reasoning.
Suppose you’re trying to decide whether you ought to shop at Wal-Mart, or water your lawn everyday, or drink on Sunday, or … . If we don’t restrict the theory, and so include non-basic norms in the code, the code will contain a norm for each of these very specific act types. It follows that the way to know whether it’s ok to shop at Wal-Mart would NOT be to engage in ordinary moral reasoning from basic principles (i.e., basic norms) that you and your fellows accept. Instead, the way to discover whether you ought to shop at Wal-Mart would be to take a poll on the very specific issue of shopping at Wal-Mart. Then you’d have to do a different poll for lawn watering, drinking on Sunday, etc.
But this is insane. No one – not even those inclined towards some sort of subjectivism – thinks we ought to decide the issue this way. What we think is that we ought to reason from basic principles. Now, if ever it comes to questioning these basic principles, we will, on SS, have to defer to collective opinion about the principle. You will find this objectionable if (like me) you find all forms of subjectivism objectionable. But I hope you agree it is clearly less objectionable than those views according to which taking a poll on every specific moral issue is the way to go.
The implausibility of allowing non-basic norms in the moral code is especially clear when it comes to Personal, rather than Societal, Subjectivism. On such a form of Personal Subjectivism, whatever a person happens to accept at the moment about the morality of shopping at Wal-Mart will automatically be correct. Again, this flies in the face of ordinary practice. We reason from more basic principles we accept. It is far less crazy, and much more in line with common practice, to suppose that the fundamental moral norms are in the eye of the beholder than it is that the answer to every specific moral issue is in the eye of the beholder.
Why it’s implausible to restrict the theory to basic moral norms. Suppose some members of some society became convinced of Societal Subjectivism WITH the restriction to basic moral norms. That is, they accept:
SSb: an act (token) is morally right iff it is it is not prohibited by the moral code of the society of the agent of the act (where the moral code of a society is the set of basic moral norms accepted by the society).
Suppose those convinced of SSb go on a campaign to convince all their fellows of the truth of their theory. Suppose they are successful. Everyone now accepts SSb. They think all of our moral obligations derive ultimately from SSb.
But then look what happens. This is just to say that there is now really just one basic moral norm that they accept – namely SSb itself. So now their view comes to this: do what society thinks you should do; and all society thinks you should do is whatever society thinks you should do. If this is their view, then their view entails that no act is prohibited. In cases where everyone accepts the view, the view reduces to Nihilism. And all because people actually come to accept the theory!
Consider an analogy. Suppose my theory says that I ought never to violate any rule written on this sacred tablet. Suppose the sacred tablet contains just one rule: never violate any rules on this tablet. It would be impossible to do anything wrong according to this theory.
So the restricted forms of subjectivism have the following feature: if the relevant subject or group of subjects mentioned in the theory comes to accept the theory, then Nihilism follows. This seems to me to be an implausible consequence of any theory, and a consequence that even defenders of subjectivism would not welcome.
Two final notes (feel free to skip them if you’re tired of reading). First, this is not the objection that the theory is self-effacing or self-defeating. It’s not that people are more likely to violate the theory if they accept it. It’s rather that, if the relevant subject or group of subjects accepts the theory, then anything goes (for some people – who it is depends on the form of subjectivism under consideration).
Second, for versions of subjectivism that appeal to an ideal observer or to God, the problem remains. What would follow is that nothing that any ideal observer or God does could be wrong, so long as the ideal observer or God accepts the ideal observer or the divine command theory. An ideal observer, or God, could go around torturing people for fun, and there would be nothing wrong with that. But not even God can go around torturing people for fun.
Chris,
Perhaps this is too obvious a move, but why categorize SSb as a moral norm? The force of your complaint that SSb is vacuous if SSb is treated as a basic norm. But it’s not obvious that SSb is a sentence that attributes a “normative status to some act type” in the same way that “adultery is wrong” attributes a normative status to an act type. SSb doesn’t mention any substantive acts (types or tokens) at all. So why couldn’t a Subjectivist take SSb as a statement about moral norms but not a moral norm itself? After all, some statements about a discourse are also statements in the discourse (“All English sentences contain a subject and a verb”) but some aren’t (“All French sentences contain a subject and a verb”). I agree there’s something funny about SSb advocacy, because the advocates would be advocating the seemingly trivial “let’s accept what we accept!”, but the SSb-ist could advocate both SSb AND advocate some basic first-order moral norms as well. They’d just be advocating propositions at different levels of moral discourse.
“Second, for versions of subjectivism that appeal to an ideal observer or to God, the problem remains. What would follow is that nothing that any ideal observer or God does could be wrong, so long as the ideal observer or God accepts the ideal observer or the divine command theory. An ideal observer, or God, could go around torturing people for fun, and there would be nothing wrong with that.”
I don’t see it yet with the ideal observer/agent version. (I’ll leave out discussion of gods, since I don’t think such views are better than ideal observer/agent theories.) One might be able to specify the conditions that make an observer ideal such that they would not accept certain things. For example, suppose one says that a certain amount of sympathy was required, and you thought that such sympathy would make it impossible to approve of torture or to accept a norm that required torturing.
Or, if one thinks there is nothing wrong with importing normative constraints into the conception of ideality, suppose you say ideal observers must be rational or reasonable and you then think that this involves meeting certain normative constraints that rule out torture.
Relatedly, I think that this sort of move helps suggest how such a subjectivist should treat the distinction between basic and non-basic norms. I think s/he can safely ignore it. The idea might be that what is right for a person is what it would be rational for that person to do (or accept norms requiring) given the commitments s/he starts out with and full information. You don’t need to decide which norms are basic and which are not so long as you think that rational revision in light of increasing information would eventually settle what to do for the cases you think we need moral judgements about.
“If we don’t restrict the theory, and so include non-basic norms in the code, the code will contain a norm for each of these very specific act types.”
A subjectivist needn’t accept this conditional, right? She could accept that there are some non-basic norms in the code, just one, “It’s wrong to shop at Wal-Mart” and then accept only basic norms in addition to this one.
Perhaps the force of the conditional is that she needs to accept an exclusion principle, a principle which will allow “It’s wrong to shop at Wal-Mart” to be non-basic but in the code, but not “It’s wrong to buy a Whopper from Burger King.” But the subjectivist could deny that an exclusion principle is needed, all she thinks we need is sufficient overlap between intuitions of the folk in virtue of which a norm will correctly count as in the code.
I’m not convinced by either horn of the dilemma.
1. On the first horn. In general, a statement of the form “(x)(Fx <-> Gx)” does not imply that the only way to know that, say, Fa is first to find out that Ga and then infer that Fa. (For example, it’s true that, for all x, x is an equilateral triangle iff x is an equiangular triangle, but one may know that a given triangle is equilateral without first measuring its angles.) Thus, one may accept SS, that an act is morally right iff it isn’t prohibited by society’s moral code, without accepting that the only way to know that, say, shopping at Wal-Mart is morally right is first to find out that society’s moral code would not prohibit it and then infer that it’s morally right.
2. On the second horn. You say:
But isn’t the last sentence a non sequitur? Why can’t one accept SSb without thinking that all moral obligations derive ultimately from it?
“But this is insane. No one – not even those inclined towards some sort of subjectivism – thinks we ought to decide the issue this way. What we think is that we ought to reason from basic principles.”
I have a question about your basic/non-basic distinction. Consider, for example, Hare’s indirect (two-level) view, based on an ideal perscriber account of the basic norms. I would think we should apply your terminology as follows: Hare holds that (most of the time) we should reason on the basis of non-basic norms instead of basic ones.
If that gloss is right, then you need to do more work to dismiss the view as insane. If the gloss is wrong, I want to know why your distinction should not be applied in that way.
Michael,
I think SSb is a moral norm. The act type it mentions is being prohibited by the moral code of the society of its agent. This is an act type: some acts fall under it; others don’t. SSb is really a conjunctive norm. It says that all acts that have this feature are wrong, and that all acts that lack it are right.
Or do you think there is a principled way for the SSb-er to distinguish between this norm and more ordinary norms like “adultery is wrong”?
Chris,
Not being much of a friend of subjectivism, I’m definitely in arguendo mode here, but perhaps this is what I was driving at with the insertion of “substantive” earlier: SSb doesn’t attribute a moral status to an act type that would enable anyone not already familiar with some another norm, namely SSb itself, to accord their behavior with it. So even if I were ignorant of SSb, I could accord my behavior with the moral norm “adultery is wrong” without having to refer to some other moral norm, whereas according my behavior with SSb requires me to know what moral norms my society’s moral code permits. (I will of course have to know the truth of other propositions, such as those about who my wife is, in order to accord my behavior with “adultery is wrong,” but I don’t have to refer to some other norm.) Perhaps that’s a circuitous way of saying that SSb is not action-guiding, which you incorporated not into your description of norms but into your description of their acceptance:
“To accept a norm is to be inclined to assent to it, to be inclined to try to act according to it …”
So if a norm requires that it be action-guiding for those who accept it, then perhaps SSb isn’t such a norm, since acceptance of it does not by itself indicate how one ought to behave?
“So now their view comes to this: do what society thinks you should do; and all society thinks you should do is whatever society thinks you should do. If this is their view, then their view entails that no act is prohibited. In cases where everyone accepts the view, the view reduces to Nihilism.”
This is not quite right; nihilism does not follow. Suppose SSb is sole the rule telling you that moral rightness is determined by appeal to basic moral norms and SSb is itself the sole basic moral norm.
In that case the basic norm (BN) that I must follow (according to SSb) is the one that states only that I must follow it in order to act morally. So the norm says only this,
BN. Follow BN
or more simply,
BN. Follow me
So there is a basic norm that my actions must conform to (viz. BN) but none of my actions could conform to it. I do something prohibited everytime I fail to follow the basic norm ‘follow me’. But it is not possible to follow a rule that states ‘follow me’. So all of my actions are prohibited.
Mark,
About building in normative constraints (like rationality) to an ideal observer theory: I do think there is something wrong with that. Here’s why. (Another dilemma … sorry for the monotony.) Either the theory of rationality that we plug in will be “objective” or “subjective.” (A theory of rationality is objective just in case, according to it, what it is most reasonable for an ideal observer to prohibit does not depend upon the attitudes of some subject. It is subjective just in case it does.) If the theory of rationality we plug in is objective, then the resulting moral theory isn’t really a form of subjectivism. In fact, the appeal to an ideal observer seems to me to be superfluous. Instead just appeal to whatever objective values ground reasonable prohibition.
If, on the other hand, the theory of rationality we plug in is subjective, then I don’t see what can guarantee that he won’t approve of torture.
This brings us to your first remark, about building in sympathy. You say, “such sympathy would make it impossible to approve of torture.” I’m gonna be really boring an present yet another dilemma. Either you specify what sympathy is using evaluative terms or using only non-evaluative terms.
If you specify what sympathy is using evaluative terms, then the definition is going to go something like this: to be sympathetic is to be disposed to disapprove of actions that you believe are harmful to people (I realize that sounds artificial, but I think my point will apply to more natural definitions as well). Now, either you build in that, to be an ideal observer, you must have true beliefs about benefit and harm, or you don’t build this in. (Ok, this is getting silly with the dilemmas.) If the former, then the theory isn’t really a form of subjectivism. If the latter, then the problem remains, because it is not logically impossible for an ideal observer to believe falsely that suffering is good for people.
If, on the other hand, you specify what sympathy is using only non-evaluative terms, then, the problem still remains, because how could it be logically impossible for someone with whatever psychological profile you specify (non-morally) to approve of torture? I don’t see how there could be a necessary connection between whatever psychological features you specify and disapproving of torture — unless you cheat (in some way like the way described in the previous paragraph).
Or, if I am missing how there could be such a connection without cheating, perhaps you could give an illustration of how it could go.
Let me say one more thing about ideal observers that bear on what I say above. The theory is always specified counterfactually: “what an ideal observer would approve of.” This suggests that ideal observers have some connection to the actual world, as if we’re supposed to ask ourselves, “If there were an ideal observer (i.e., a person with features X, Y, and Z) here in our world, what would he approve of?” This is different from the following question: “What must an ideal observer (i.e., a person with features X, Y, and Z) approve of?” It could be that, if there were an ideal observer, he would disapprove of torture. But I don’t think it could be that an ideal observer must (of metaphysical necessity) disapprove of murder — unless you cheat in your specification of what an ideal observer is.
Why is this latter claim relevant, since the theories are always stated counterfactually? It is relevant because, given that it’s metaphysically possible for an ideal observer to approve of torture, I think it follows that the counterfactual — if there were an ideal observer, he would approve of murder — is true in some world. THAT ideal observer is the person who is allowed to torture, if he accepts the ideal observer theory.
(Man, I am really in a rut. I need a new argument form.)
Mark, you make an additional remark. I’ve been long-winded enough responding to your other remarks, so let me shut up for a minute.
Christian,
I think it would be hard for a subjectivist to deny that conditional (i.e., “If we don’t restrict the theory, and so include non-basic norms in the code, the code will contain a norm for each of these very specific act types”) as applied to our society. An unrestricted code would include all the norms we accept. And to accept a norm is just to be disposed to do certain things and have certain attitudes. Given the dispositional nature of acceptance, norm acceptance comes pretty cheap. Doesn’t it?
I’m not sure I was getting the stuff about exclusion principles. Could you elaborate?
I think this is along the lines of many of the other criticisms, but as you seem to dismiss them, I’ll see if presenting another perspective on the problem helps. It is surely not SSb alone which makes acting according to a basic norm of society right. Rather, that something has the property of being a basic norm of society plays a crucial role in making the following of it correct. So if SSb is a norm, it does not seem to follow in any sense that it is more basic than other norms; it seems you need all the basic norms for the theory to work.
I’m also in sympathy with the people who question whether SSb itself is even a norm in the first place, but if it is a norm, I see no reason to view it as more basic than the other basic norms.
I’m thinking that if accepting a norm is being disposed to assent to it, for example, being disposed to accept that shopping at wal-mart is wrong is to accept the norm that shopping at wal-mart is wrong, then people could have these dispositions both towards basic norms, like harming others for no good reason is wrong, and also non-basic norms, like shopping at wal-mart is wrong.
You seem to suggest, and I might be misinterpreting you, that if people are disposed to accept at least one non-basic norm, then they are disposed to accept a truck-load of them. But I don’t see why we should think that. Perhaps people are disposed to accept some basic norms and they reason from them together with the empirical facts, and people also reson from non-basic norms, like killing an infant for fun is wrong, and x is relevantly similar to that. So if it is just a matter of enough people in a society sharing dispositions to assent to a certain norm, then it seems plausible to me that there will be both basic and non-basic norms that a lot of people assent to, although there will be many non-basic norms that people are undecided about.
Campbell,
I definitely agree that one could accept SS without accepting that the only way to know that shopping at Wal-Mart is morally right is first to find out that society disapproves of shopping at Wal-Mart and then infer that it’s morally right. SS doesn’t demand that we reason in this way. But it allows it, and that’s what I’m objecting to. We never reason in this way that SS allows. No one decided whether it was ok to unplug Terri Schiavo by consulting the Time/CNN poll on the matter (they instead checked out whether she gave any advanced directives, whether her brain might possibly heal, etc.)
It’s actually an understatement to say that SS merely allows this way of reasoning. I think it’s fair to say that SS would regard this as the “canonical” way to reason morally. (“Canonical” in the sense that, according to consequentialism, the canonical way to reason morally is to weigh consequences; according to Rossianism, the canonical way is to weigh prima facie duties; according to virtue ethics, the canonical way is to figure ask WWJD; acccording to Kantianism, the canonical way is to see if you can universalize your maxim; etc.)
Perhaps another way to get at what I’m objecting to about SS is to consider what it implies about moral explanation. Given SS, the following is not only a perfectly acceptable moral explanation, it’s really the “canonical” way to explain why some act is right or wrong:
But no one offers or accepts such moral explanations. Instead we appeal to general principles (e.g., about when, if ever, pre-emptive war is in general justified; this in turn would likely be explained by another, more fundamental, principle). So our views about what sorts of moral explanations are appropriate favors SSb.
About the second horn, you ask, “Why can’t one accept SSb without thinking that all moral obligations derive ultimately from it?” This is because SSb isn’t just a biconditional — it’s a normative theory. It purports to identify that feature in virtue of which all right acts are right. So to accept SSb is to accept that, whenever some act is obligatory, it is obligatory because it is required by the set of basic moral norms accepted by our society.
If one accepts some version of consequentialism, one thinks all moral obligations derive ultimately from it, don’t they?
Brad,
That’s a good point. Hare (and many others) actually hold that we ought not to reason from basic principles. If we try to do this, we’ll just screw it up. Better to use mid-level “rules of thumb.”
Two things to say. First, SS still permits a sort of reasoning that, it seems to me, not even these “mid-levelers” advised. For any even very specific act type (e.g., whether it’s ok “pad” one’s resume), SS would allow us to decide it by taking a poll on that very question. But Hare would say (I would presume) that we ought to use a rule of thumb (like “don’t lie,” or “do unto others,” or whatever). SS wouldn’t require this.
Second, I’d like to re-emphasize the moral explanations argument I suggested above in my post to Campbell. Though Hare thought we should reason from non-basic rules of thumb, he held, correctly, that the ultimate explanation for why some act is wrong must appeal to the correct basic principle or principles (in his view, utilitarianism). SS implies that many such moral explanations we would find totally bizarre are in fact perfectly acceptable.
Protagoras,
I am the measure of all things, and I say SSb is more basic!
Just kidding.
I agree that, even if SSb is true, it need not be a basic norm for some society (the society could be, say, a society of utilitarians, in which case utilitarianism is their only basic norm). But my objection to SSb didn’t depend upon this. Rather, it asks us to imagine some society all of whose members accept SSb as the correct moral theory. From this I think it follows that SSb is a (in fact, the) basic norm for this society.
Chris,
I take your point that SS permits moral-reasoning-by-opinion-poll without requiring it. But I don’t see how SS implies that this is the “canonical” form of reasoning. Consider:
SS and U are consistent. One could accept both. In which case one might say that the canonical form of moral reasoning is to consider the effects of one’s actions on people’s utilities. (I would object in a similar way to your claim that SS implies that a certain form of explanation is canonical.)
You go on to say:
I think this goes to the heart of the issue between us. I was assuming that SSb was just a biconditional (well, actually a biconditional within the scope of a universal quantifier). It certainly looks for all the world like a biconditional. It would be helpful if you could explicitly include the extra bit, the bit which makes it more than a biconditional, in the definition. (By the way, it’s not clear to me why a normative theory could not be “just a biconditional”.)
Chris,
I have a follow up question about Hare which relates to your response to Mark. I think your argument about moral explanations brings out what I find counter-intuitive about the social version of subjectivism, but that ideal actor/prescriber forms of subjectivism are harder to rebuff.
Consider: Although Hare would say that the basic norms are utilitarian in content, he derives this as a conclusion from his account of universal prescriptivism, which helps make up his account of moral discourse. If I understand it aright, Darwall’s current project reminds me of this – Simplifying, I think he plans to derive a basic Kantian norm from an account of second-personal demands.
The basic formula of these constructist accounts might be put like this: observer/prescriber/agent + extensive knowledge of the facts + thin rationality + account of moral address/discourse/practice => basic moral norm
Given what you said in response to Mark, you would presumably claim that these broadly constructivist accounts are not forms of subjectivism. You wrote:
“If the theory of rationality we plug in is objective, then the resulting moral theory isn’t really a form of subjectivism. In fact, the appeal to an ideal observer seems to me to be superfluous. Instead just appeal to whatever objective values ground reasonable prohibition. If, on the other hand, the theory of rationality we plug in is subjective, then I don’t see what can guarantee that he won’t approve of torture.”
I wonder how you would respond to this: the accounts of moral address (universal prescription or second-personal address), which these theories use to get to a basic norm, are objective – in the sense you mention in the response to Mark – but that does not make the theories of morality subjective in the sense of presupposing that there are any objective “values” out in the world; their objective assumptions are about the nature or concept of morality, and they can make due without a contentious, substantive conception of rationality or reasonableness, which you rightly point out seems to depend on positing substantive values.
Chris,
Perhaps a better definition of the view you have in mind is the following.
Your claim, as I understand it, is that if the basic moral code includes SSb*, then it cannot include anything else; if society accepts SSb*, then any other moral norm it accepts is non-basic. I see the idea, but I’m still not quite convinced. Suppose society accepts both SSb* and U (as I defined it above). Does it follow that society “derives” U from SSb*? I don’t yet see why it does. For one thing, it could be that society accepts U before it accepts SSb*.
One last thing.
The following seems a natural principle for ordering moral norms in terms of how basic they are:
(Where N “settles the moral status” of an act X iff there’s some moral status M–i.e. wright, wrong, etc.–such that N implies that X is M.)
This seems to capture your example. All acts of adultery are acts of promise-breaking, but some acts of promise-breaking are not acts of adultery. So (B) says that the norm, “Promise-breaking is wrong”, is more basic than the norm, “Adultery is wrong”. However, (B) implies that SSb and U are equally basic, because both settle the moral statuses of all acts.
Campell,
That is an interesting gloss on the notion of “relative basicness” (ugg). And your extensional inclusion interpretation may capture what is at issue here. But maybe the definition should also capture the justificatory sense of basicness.
*Very* roughly:
(E) A moral norm N envelopes another N’ iff the set of acts ruled out and in by N’ is a subset of the set of acts ruled out or in by N
(B) A moral norm N is more basic than another N’ iff N envlopes N’ and we should live in accord with N’ because we should live in accord with N.
Brad,
Interesting suggestion — I suspect Chris needs something like that. But here’s a worry. (B), as you define it, seems to imply that whether N is more basic than N’ depends on whether we should live in accord with N and N’. (Here I assume that “P because Q” implies “P and Q”.) Suppose, for example, it’s not the case that we should live in accord with the norm, “Adultery is wrong”. Then your (B) implies that this norm is neither more nor less basic than any other norm. Isn’t that odd?
Well, I’ve always thought social subjectivism a bit silly, largely because it seems empirically false given the point you raise; we think polls are irrelevant. But it seems to me that individual subjectivism is extremely plausible for exactly the same reason. Surely a careful, reflective look at one’s feelings about the situation is how most of us make moral decisions when we’re being serious, and the rest of the time it’s pretty plausible that we’re making decisions based on offhand feelings in reaction to a situation.
My personal position is largely that of Hume and Schlick (and apparently Hare, based on comments in this thread); subjectivism is the only plausible source of normativity, and sufficiently reflective application of subjective norms produces, in most people, something very like utilitarianism.
Michael,
Thanks for your willingness to arguendo.
I agree that the norm “adultery is wrong” doesn’t mention other norms, whereas SSb does. But I’m not sure that’s an important difference. Intuitively, “It’s wrong to violate any of the Ten Commandments” is a norm, even though it mentions other norms, and even though acceptance of it need not by itself indicate how one ought to behave (in an action-guiding sense of ‘indicate’).
I think I disagree with something else you said (though I’m not sure if this affects your main point). I don’t think that according my behavior with SSb (or with the Ten Commandments norm above) requires me to know what moral norms my society’s moral code permits (or to know what the Ten Commandments are). I’ll illustrate why using the Ten C norm. I could accept the Ten C norm and be motivated to act according to it, even if I don’t know what they are (perhaps I used to know, and I got amnesia). Suppose I’m about to act, and I don’t want to violate any of the Ten C’s. I might ask my buddy if what I’m about to do violates any of the Ten C’s without actually asking him what the Ten C’s are. I might be prepared to refrain from acting in that way if he tells me I’d be violating one of the Ten C’s. In such a case, it seems the Ten C norm is action-guiding for me, even though I don’t know what the Ten C’s are.
Obviously, we could construct an analogous case for SSb.
Mike A.,
It still seems to me that nihilism does follow. (BTW, I should have said I’m taking nihilism to be the view that nothing anyone does or could do would be wrong. And my claim is that, given SSb and the claim that SSb is itself the sole basic norm for some society, nihilism holds for this society. And all that means is that nothing that anyone in this soceity does or could do would be wrong (so long as we hold fixed that SSb is itself the sole basic norm for the society). But I think you were on board with all this.)
Anyway, back to your point. As I suggested above, a more perspicuous way to represent SSb as a norm is to put it like this: “it’s wrong to perform any act prohibited by the basic norm(s) of your society, and it’s ok to perform any act not prohibited by the basic norm(s) of your society.” If this conjunctive norm is the sole basic norm, then it seems to me that it’s not possible for an act to be prohibited by it. Walking down the street wouldn’t be prohibited by this norm. Neither would lying, stealing, torture, etc. The norm above doesn’t prohibit these acts. And SSb says that any act that is not prohibited by the basic norm(s) is right. Since no act is prohibited by the norm above, then if this is the only basic norm for some society, and if SSb is true, then nihilism follows for the members of this society.
I think your line of thought goes wrong at the following juncture: “In that case the basic norm (BN) that I must follow (according to SSb) is the one that states only that I must follow it in order to act morally.” The norm doesn’t state that you must follow it in order to act morally. It states rather that you must not perform an act that is prohibited by it in order to act morally.
But I think your point does reveal that the fact that nihilism follows is really just an artifact of how I formulated SSb. If I had instead formulated it as follows
then I think it would follow that all actions (of those in the relevant society) would be wrong.
But this is at least as objectionable as nihilism, so I don’t think my main argument turns on this artifact of formulation.
Christian,
I don’t think I disagree with anything in your last post. I agree that, at least for societies with people like us, there will be norms we have no feelings about, even dispositionally. And I did not mean to say that if people are disposed to accept at least one non-basic norm, then they are disposed to accept a truck-load of them.
Do you think anything I agree to above makes trouble for my argument?
Campbell,
About exactly what SSb is saying. Sorry, I was being lazy. It does look for all the world like just a biconditional, but it is so much more than that. Here’s what that sentence really means:
The reason normative theories must, in my view, take this form, and can’t be mere biconditionals, is that they are meant to explain what makes an act right. That is, they must identify that feature in virtue of which right acts are right. They reveal not just an accident of rightness, but its essence, or its nature. Let me appeal to authority: Socrates sez: “Thus you appear to me, Euthyphro, when I ask you what is the essence of holiness, to offer an attribute only, and not the essence — the attribute of being loved by all the gods. But you still refuse to explain to me the nature of holiness.”
Mere (universally quantified, material) biconditionals don’t reveal natures. Suppose, as an amazing coincidence, all and only right acts are performed during an odd minute of the day (10:23, 3:37, etc.). Then this is true:
(This is a material biconditional.) This sentence would be true, but it wouldn’t be the true normative theory. It doesn’t explain why the rights acts are rights. It reveals only an “attribute” of right acts, not their essence.
Even finding something that is necessarily, and not just accidentally, connected with moral rightness may not reveal its nature. Consider:
(Notice no ‘in virtue of’ included.) Even if we grant that this sentence is true (as Socrates does) it is not a theory of rightness, since it doesn’t purport to explain what makes right acts right, or identity the feature in virtue of which they are right.
Perhaps now it is clear why I would reject your claim that SS and U are consistent (if they meant as theories rather than mere biconditionals).
About “canonical” forms of reasoning. I admit that this is dicey. As Brad pointed out, many utilitarians urge that we NOT reason in the way I said was canonical for their theory. So I think putting my point in terms of explanation, or the in virtue of relation, is better. For each theory, there IS a canonical way to explain the moral status of an action. And (to return us to the original point) SS’s canonical way is very implausible; SSb’s way is less so.
I think subjectivism implausible as a theory of morality but quite plausible in other domains.
Still I am not persuaded by the argument above. Why not just focus on the more general moral beliefs to conjoin with the subjectivist account of the moral truth maker.
Further, the thought seems to be that if the truth-maker does not determine the answer all by itself, unsupplemented by further info, then there is something vacuous about the truth-maker. But this seems wrong. Consequentialism tells us that what makes something right is that it promotes the good, but this does not tell us what is right or wrong without supplementation from a theory of the good. Yet there is no criticism of C here.
Chris,
You write:
, on the other hand, you specify what sympathy is using only non-evaluative terms, then, the problem still remains, because how could it be logically impossible for someone with whatever psychological profile you specify (non-morally) to approve of torture? I don’t see how there could be a necessary connection between whatever psychological features you specify and disapproving of torture — unless you cheat (in some way like the way described in the previous paragraph).
Or, if I am missing how there could be such a connection without cheating, perhaps you could give an illustration of how it could go.
OK, let’s see if I can answer this or say something helpful about it.
The basic idea is that something is wrong iff an observer/agent of such and such a sort would disapprove/recoil from actions of that sort in conditions of full information. For something to be wrong is for it to be the sort of thing that would be so reacted to.
I’m not actually sure that this requires it to be logically necessary for observers so specified to approve of the things that are in fact wrong, if by logically necessary you mean that the necessity has to follow by the rules of logic and the definitions of the terms in the description of the ideal observer. All that is necessary is that observers meeting those conditions would have the reaction indicative of wrongness under conditions of full information. That can be true without it being required by logic. Salt is soluble because it would dissolve if put in water (in normal conditions given the laws of physics/chemistry, etc., etc.).
So the thought behind the ideal observer/agent type theory is that given a relatively careful specification of the conditions there will be facts about what observers so-specified would react to in the required ways. If you look at Firth, he builds into his conditions an “otherwise normal” clause, which I take it to mean that observers should have the ordinary human dispositions to respond to various kinds of things. Now it might be that human psychology is not determinate enough to fix uniform responses of such observers to certain sorts of actions. But it might also be that for a wide range of cases there is uniformity of response. If so then it is true that actions of that sort are such that ideal observers/agents would respond in such and such a way to actions of that sort. And that would be sufficient for them to count as wrong according to the analysis.
Notice, that the success of the analysis depends not on whether this is guaranteed by logic and definitions, but on whether the psychological facts are such as to underwrite the claim that they would react that way.
Now suppose that even psychologically normal observers don’t respond in a uniform way. That would lead me to think that we should relativize further. An action is right for an agent if ideal agents of such and such a sort otherwise like the actual agent would be motivated to do the action if they knew all of the facts, or some such thing. Or to put it another way, an action is right if the agent would be motivated to do the action in ideal conditions of such and such a sort and with full information.
So getting back to sympathy and torture. If you think there are facts about how ordinary humans who are sympathetic would react to torture, you could think that the required subjunctive conditionals were true without it being a matter of logic.
I have a little more to say about something else, but I’ll put it in another comment.
Chris,
Fair enough on the Socrates stuff. I confess I’m sceptical about all that, but I gather this is an idiosyncrasy of mine.
Chris,
My other comment is about this:
About building in normative constraints (like rationality) to an ideal observer theory: I do think there is something wrong with that. Here’s why. (Another dilemma … sorry for the monotony.) Either the theory of rationality that we plug in will be “objective” or “subjective.” (A theory of rationality is objective just in case, according to it, what it is most reasonable for an ideal observer to prohibit does not depend upon the attitudes of some subject. It is subjective just in case it does.) If the theory of rationality we plug in is objective, then the resulting moral theory isn’t really a form of subjectivism. In fact, the appeal to an ideal observer seems to me to be superfluous. Instead just appeal to whatever objective values ground reasonable prohibition.
Suppose you thought that rationality did not put substantive constraints on ends, but only on relations between ends. This might be an objective matter. But it would not make references to an ideal observer superfluous. For what a person would aim at under conditions of full rationality and information would still be different depending on where they start from.
This idea is something like what motivates Harman’s relativism. There are rational constraints on what one should accept given what one starts out accepting. But there are no moral demands that all people should accept because what one should accept depends on what one does accept, and for any moral claim there might be someone who does not accept other things that make the acceptance of that demand rationally required.
So he suggests that a moral demand applies to someone only if they would accept it under conditions of full information and rationality. The reference to the person is not superfluous because we need it to fix the initial commitments which are then used to generate rational commitment to further demands. The demands of rationality by themselves are not sufficient to do that.
Yet the account contains a normative constraint.
You might also think that there were some substantive ends everyone should rationally have, but that these were not sufficient to fix all of the demands that applied to a person. Even according to such more substantive theories of rationality the reference to the person would not be superfluous.
So I think that it is not so easy to show that there should be no normative constraints put on ideal observers/agents for fear of no longer having a subjectivist theory. But I’m of two minds about what to do at the limit, where you think that the input conditions provided by the reference to the agent/observer play no role in determining what the output is in terms of moral requirements. I can see some reason to think that such theories are not really subjectivism anymore. But I’m not sure that is right.
Chris, thanks. You write:
“it’s wrong to perform any act prohibited by the basic norm(s) of your society, and it’s ok to perform any act not prohibited by the basic norm(s) of your society.” If this conjunctive norm is the sole basic norm, then it seems to me that it’s not possible for an act to be prohibited by it.”
I think I was less clear than I might have been. I will leave aside the second conjunct, since it does not matter to my worries. So the (trimmed) exclusive basic norm is BN.
BN. It’s wrong to perform any act prohibited by the basic norm(s) of your society.
So far, so good? Now consider the definite description in BN, ‘the basic norm of your society’. There is only one such norm, by hypothesis, and that norm is BN. So, (1) is true,
1. BN = the basic norm of your society.
Given (1), I can restate BN in this (equivalent) way.
BN’. It’s wrong to perform any act prohibited by BN’.
So what is it wrong for me to do? It is wrong to perform any act that is prohibited by BN’. What does BN’ require? BN’ requires that I do nothing prohibited by BN’. So, an example. It is wrong to lie iff. lying is prohibited by the norm that prohibits lying. And more generally,
z is wrong iff. z is prohibited by the norm that prohibits z.
But, trivially, everything has the property of being prohibited by what prohibits it. So everything is wrong.
Dear Chris,
“Do you think anything I agree to above makes trouble for my argument?”
I think so because you defend the claim that the first horn of the dilemma is implausible by saying, “It follows that the way to know whether it’s ok to shop at Wal-Mart would NOT be to engage in ordinary moral reasoning from basic principles (i.e., basic norms) that you and your fellows accept.”
But if I’m right, then doesn’t follow. Just because there are some non-basic norms in the moral code, it will not follow that we would not engage in ordinary moral reasoning. On the view I put forward, we will still engage in ordinary moral reasoning, deriving obligations from basic moral claims and the facts, for example. We will also not need to do any polling for the non-basic norms we accept. This is because we can accept that each of us shares dispositions to accept the same non-basic norms as those around us, the default position will be that we are different. If this is right, the first horn isn’t so implausible after all and we have no reason not to take it.
Mike A writes:
It is wrong to perform any act that is prohibited by BN’. What does BN’ require? BN’ requires that I do nothing prohibited by BN’. So, an example. It is wrong to lie iff. lying is prohibited by the norm that prohibits lying.
I’m missing something. BN doesn’t obviously to me seem to prohibit lying. Why does it follow from BN that lying is prohibited? What is the connection between BN and the norm that prohibits lying? Is there a hidden premise here?
Whoops, in my last post, where I have ‘BN’, I should have had ‘BN”.
Hey Mark,
What I claim is that BN entails the following,
z is wrong iff. z is prohibited by the norm that prohibits z.
My example is this,
Lying is wrong iff. lying is prohibited by the norm that prohibits lying.
And I assert that it is trivially true that everything is prohibited by what prohibits it (what else could prohibit lying, but the norm that prohibits lying?), so the right-hand side is trivially true. It follows that lying (and everything else) is prohibited. Does that sound mistaken? So I conclude that nihilism is false. It is rather the case that everything is prohibited.
Brad,
This pertains to an earlier post (as you can see, I’m having trouble keeping up, but I really do appreciate the feedback from everyone). You write:
First, did you mean to say, ” … but that does not make the theories of morality objective … “? In any event, is your question: (i) do I regard the view you describe as a subjectivist view, and (ii) if so, do I think my moral explanations objection applies to this view? I’m not totally sure what to say, since I don’t have a firm enough grasp of the kind of view you are describing, but I am inclined to think my moral explanations argument would NOT apply to this view. I’m even less sure about whether the second horn of the dilemma would apply. But I probably shouldn’t speculate, since I don’t fully understand the view.
Protagoras,
You say:
Well, I was hoping my dilemma would apply to individual, or personal, subjectivism as well, i.e., the following view:
If ‘personal moral code’ is defined so as to include non-basic norms (such as, e.g., “littering is wrong”) then PS allows implausible moral explanations (like, “your act of littering just then was wrong because you accept that littering is wrong”).
On the other hand, if ‘personal moral code’ is defined so as to include only basic norms, then if you accept PS, then that is your only basic norm (given the stuff above about how it’s a theory, and so to accept it is to accept that all right acts derive their rightness from failing to be prohibited by the agent’s personal moral code; and if there’s something norm you think all right acts derive their rightness from, then that norm is basic for you and is in fact the only basic norm for you). And if that’s your only basic norm, your view prohibits nothing.
Or that’s the argument, anyway.
Campbell,
Yeah, your SSb* is another way of capturing what I was trying to capture with the “in virtue of” talk.
However, then you write:
That’s a good question. I hadn’t thought about this until to asked this question, but it seems to me that if some society accepts SSb* at some time, then they can accept no other norms (basic or otherwise) at that time (on pain of contradiction). If they accept SSb*, they are accepting that all right acts derive their rightness from not being prohibited by the basic moral code of their society. And this is tantamount to saying that SSb* is the only basic norm (otherwise, not ALL right acts would derive their rightness from it — there could be other acts whose rightness derives from some other basic norm).
If this society also accepts U (it would have to be a non-basic norm, for the reason just given), then they will be accepting that all non-optimific acts are wrong. But they can accept, of some non-optimific act, that it is wrong only if their basic moral code contains a norm that prohibits this act. But the code contains no such norm, since it contains only SSb*. Therefore, they can’t accept U after all.
(All of this was assuming the society is not inconsistent in their accepting. I haven’t yet thought about what to say about inconsistent societies.)
Anyway, does that seem right to you?
Mike,
You write:
Hey Mark,
What I claim is that BN entails the following,
z is wrong iff. z is prohibited by the norm that prohibits z.
Yes I see that is what you think BN (or BN’) entails. But I’m missing the entailment. BN’ seems to me to say that we should not do what BN’ prohibits. But unless we have an independent argument that BN’ prohibits lying (to use your illustrative example), I don’t see how we get the claim that that entails we should not lie.
Once you get the claim that for all X, BN prohibits X, I can follow the rest of the argument. But I don’t see how you got there.
BN’. It’s wrong to perform [any act prohibited by BN’].
The question is what exactly are the acts prohibited by BN’ in the bracketed part of BN’ above. The answer is that BN’ prohibits all and only those acts that the basic norm BN’ itself prohibits. This is in (1).
1. The acts prohibited by BN’ just are the acts the basic norm (BN’) prohibits.
2. :. It is wrong to perform any act prohibited by the basic norm that prohibits it. From BN’, 2
As I read (2), it is a biconditional such a (2′), but my argument would work as well if the principle in (2′) were true right to left.
2′. z is wrong iff. z is prohibited by the norm that prohibits z
Mike, where does the clause “the norm that prohibits it” come from?
Every view, I suppose, entails that it is wrong to perform any act that a basic norm prohibits. And it seems trivially true that
(z) z is wrong if z is prohibited by a basic norm that prohibits it.
if only because the last clause is pleonastic. But why ‘the’? Is there supposed to be existential presupposition or other existential commitment in your version?
Jamie,
I think I follow you, but I’m not certain. You write,
“where does the clause “the norm that prohibits it” come from?”
BN’ states this,
BN’. It’s wrong to perform [any act prohibited by BN’].
What does BN’ prohibit? BN’ prohibits all and only those acts that “the basic norm itself prohibits”. This is where the phrase comes from. BN’ prohibits all acts that have the property of being prohibited by the basic norm (viz., BN’, since BN’ is the only basic norm). It’s trivial. So we could put it this way,
z is wrong iff. z has the property of being prohibited by the basic norm that prohibits z.
Is that any clearer?
Mike,
I think Jamie and I are having the same problem.
The claim –
‘ z is wrong iff. z has the property of being prohibited by the basic norm that prohibits z. ‘
– seems to be being used in such a way that it involves an existential commitment to a basic norm prohibiting z. If it did not involve such a commitment you would not be able to conclude from it that z is wrong. (Notice that Jamie’s version (z) which substitutes ‘a’ for ‘the’ does not entail that z is wrong.)
But if it does involve such a commitment, then it isn’t entailed by BN’. For suppose there was only one basic norm and it did not prohibit z. Then it would still be true that z would be wrong if it was prohibited by that norm. But it would not be prohibited by that norm and hence not wrong (on the assumption that wrongness requires being prohibited by a basic norm).
Mark, you say this,
“If it did not involve such a commitment you would not be able to conclude from it that z is wrong. (Notice that Jamie’s version (z) which substitutes ‘a’ for ‘the’ does not entail that z is wrong.)”
But I’m not assuming that there is a norm that prohibits z, or I’m certainly not trying to do that. Mty conclusin is supposed to follow trivially from the fact that everything is prohibited by a norm that prohibits it. Or, to put it negatively, nothing is prohibited by a norm that does not prohibit it. So I’m happy to replace ‘the’ with ‘a’. Strictly, the trivial claim is this,
1. (Vz)(z is prohibited by a norm that prohibits z).
(1) just says that nothing is prohibited by a norm that does not prohibit it. Or put positively,
2. [](Vz)(Vn)(norm n prohibits z only if z is prohibited by n)
Since it is true of every action that it is prohibited by a norm that prohibits it, every action is wrong. That’s the idea.
Mike,
If (1) in your immediately previous response is trivial, is it trivial enough to be true in situations where there are no norms?
Mark,
I’m pretty sure it is true under your assumption.
1. (Vz)(z is prohibited by a norm that prohibits z).
But, as I mentioned above, (1) just says that nothing is prohibited by a norm that does not prohibit it. I’ll state that explicitly in (1′).
1′. ~(Ez)(En)(norm n prohibits z & z is not prohibited by n).
(1′) is true in worlds where there are no norms, and therefore so is (1). Incidentally, (1′) is just the negative version of (2) in the previous note above.
I think that means it will also be true in situations where there are just two norms, one prohibiting lying, and the other prohibiting cursing. But in that situation it is not the case that everything is wrong, because any acts which are not either lyings or cursings will not be wrong (again on the assumption that the way acts get to be wrong is by being forbidden by a norm). And I thought the upshot was supposed to be that everything is wrong.
“I think that means it will also be true in situations where there are just two norms, one prohibiting lying, and the other prohibiting cursing.”
Mark, I am completely lost. I have no idea why you say this. (1′) is trivially true.
1′. ~(Ez)(En)(norm n prohibits z & z is not prohibited by n).
In (1′) z quantifies over actions and n quantifies over norms. Suppose there are no norms. In that case (a) is false,
(a) (Ez)(En)(norm n prohibits z & z is not prohibited by n).
But obviously if (a) is false, then (1′) is true. Now suppose the first conjunct of (a) is true. In that case, the second conjunct must be false (otherwise we have a contradiction). But then again (a) is false and (1′) is true.
So (1′) is trivially true: every action has the property of not being prohibited by a norm that does not prohibit it. But then every action is wrong.
David,
Sorry it’s taken me a while to respond to you. Sorry for something else, too: I’m not sure I’m following you. You say:
I’m guessing you’re offering a third option (to my SS and SSb) that avoids the defects I allege they have. Could you spell out this theory in more detail for me?
I still don’t see how the argument is supposed to work, Mike.
So (1′) is trivially true: every action has the property of not being prohibited by a norm that does not prohibit it. But then every action is wrong.
Seems like you’re missing a premise. It must be this one, or something equivalent to it:
Every action that is not prohibited by a norm that doesn’t prohibit it, is wrong.
This premise is pretty obviously false, so I guess your point must be that it is implied by the version of subjectivism in question. But I don’t think it is.
Mark,
This is about an earlier post of yours about the ideal observer theory (not the current thread with Mike and Jamie).
I’m totally on board with everything you say, including that the theory appeals to what norms an ideal observer would accept. I agree that this view is a genuine version of subjectivism — it doesn’t “cheat.”
BUT, I still think the second horn of my dilemma would work against this view. That horn would ask us to imagine that the ideal observer herself accepts the ideal observer theory (the “basic norm” version of the theory). Call this theory ‘ISb’ for ‘Ideal Subjectivism (with the restriction to basic norms)’. If there were an ideal observer who accepted this theory, then her only basic norm would be this very theory (as we’ve said, that’s entailed by accepting it). And if this is her only basic norm, then, supposing this theory is true, no acts would be prohibited for her.
This is compatible with the claim that if there were an ideal observer, she would accept basic norms other than ISb, including one that forbids torture. For, if there were an ideal observer, odds are she wouldn’t accept ISb. In my argument, we ask, essentially, if there were an ideal observer who also accepted ISb, would she accept a basic norm that forbids torture? Given that theories just are basic norms and that only one theory can be true, the answer must be No.
Mike,
Mark, I am completely lost. I have no idea why you say this. (1′) is trivially true.
Basically the main reason is the same thing Jamie is pointing at – that I cannot see how we get from 1′ to everything is wrong. But more specifically, I was giving a model for a situation in which nihilism was false (there are some true moral claims – the ones about lying and cursing), 1′ was true, and not everything one could do was wrong.
I’m not sure if there is more I can say to help explain my thinking.
Chris,
I’m going to reread your comments upthread and think a bit before I answer your latest comments, but I will do so later today or tomorrow.
Christian,
You say:
I agree that we wouldn’t. (In fact, maybe there actually is more or less such a thing as the set of our basic and non-basic moral norms. Still, we reason in the ordinary way — i.e., not from the non-basics.) Indeed, part of my argument was that we wouldn’t and we don’t reason in this radically non-basic way, and, more importantly, we don’t recognize it as a legitmate way to reason morally. But if SS were true, it would be totally legitimate to do so.
But maybe I’m misunderstanding you. Are you denying that if SS were true, it would be totally legitimate to do so.
(For the record, for reasons that have been discussed, it would be better to cast the points above in terms of moral explanation rather than moral reasoning.)
Chris,
I think I have at least one response. You say:
If there were an ideal observer who accepted this theory, then her only basic norm would be this very theory (as we’ve said, that’s entailed by accepting it). And if this is her only basic norm, then, supposing this theory is true, no acts would be prohibited for her.
It looks to me like this norm by itself could not be the only basic one accepted by an ideal observer, at least on plausible versions of the theory.
Remember that one can accept a norm in the relevant sense if one is disposed to act in accordance with it. Quoting from above, “To accept a norm is to be inclined to assent to it, to be inclined to try to act according to it, to tend to feel guilty if you don’t; to tend to feel contempt towards others who don’t, etc.” Since we know that ISb by itself is insufficient to generate any moral commitments, and we have characterized the psychology of the relevant ideal observers as one in which they tend to feel sympathy (or whatever we have built in) we might not implausibly think that sympathetic people will be disposed to choose actions that don’t harm others, feel guilt when they do, feel contempt at those who needlessly harm others etc. So they will count as accepting further norm beyond ISb in your sense. And since this is not derived from ISb it too is basic.
So it looks like she might well count as accepting a norm that forbids torture, either by itself or in conjunction with ISb.
We can make all of this even more complicated by remembering that many versions of the ideal observer theory involved not the judgements of ideal observers, but certain affective responses. And it isn’t obvious that accepting ISb needs to displace any affective responses that an ideal observer might feel toward act types that she would otherwise regard as wrong. And if they did change those responses, an adherent of the theory might just stipulate that idea observers could not accept any judgements or moral judgements that influenced what affective responses such observers would have.
(I’m not saying that I myself like these versions, but the old-fashioned IOT theorists like Firth were leery of any hint of circularity, so they were careful to indicate that the sort of approval and disapproval they were requiring was not based in any moral judgement. Since I’m not so worried about cirularity I’m not sure they needed to do that.)
It’s late, and I’m not sure this is very clear, but I think it is one response an ideal observer or agent theorist might make to your comment.
Jamie,
Here is how the argument is supposed to go.
1. It’s wrong to perform any act prohibited by BN’. (this is the norm BN’)
Now the question is *what exactly are the acts prohibited by BN’ in (1) above.* The answer is that BN’ prohibits all and only those acts that the basic norm BN’ itself prohibits. This is in (2).
2. The acts prohibited by BN’ just are the acts the basic norm BN’ prohibits.
Is the problem with this first inference from (1) and what BN’ prohibits to (2)?
3. :. It is wrong to perform any act prohibited by the basic norm that prohibits it. From 1, 2
Is the problem with the inference from (1) and (2) to (3)?
As I read (3), it is a biconditional stating what is in (4),
4. z is wrong iff. z is prohibited by the norm that prohibits z.
The discussion has concerned (as far as I can see) what the right hand side of (4) states. I’ve been saying that the right-hand side holds trivially for every action. The right-hand side says this,
5. z is prohibited by the norm that prohibits z
Stated formally and carefully, (5) says this,
6. ~(Ez)(En)(norm n prohibits z & z is not prohibited by n)
(6) is equivalent to the positive version in (7),
7. (Vz)(Vn)(norm n prohibits z only if z is prohibited by n)
And (6) and (7) are tivially true. So the righthand side of (4) is trivially true. But then every action is wrong.
Summing up—–
The more explicit statement of BN’ is expressed in (3). But (3) is just the biconditional in (4). The right-hand side of the biconditional in (4) is trivially true. I put the righthand side of (4) in (5). That (5) is trivially true is easy to see in its formal representation in (6). Every action satisfies (6) and therefore every action is wrong.
I think that’s as explicit as I can make the argument (I mean, in a post). Which inference are you (most) worried about?
Chris,
“Indeed, part of my argument was that we wouldn’t and we don’t reason in this radically non-basic way, and, more importantly, we don’t recognize it as a legitmate way to reason morally.”
We are likely talking past one another. I was trying to say the following claim:
“If we don’t restrict the theory to basic moral norms, then the theory generates implausible results concerning moral reasoning.”
is not true. So, I suggest there are no implausible results that are entailed by, or plausibly implied by, accepting that we reason from non-basic norms. I’m sure we do it all the time. You seem to think there are implausible results and I guess I just don’t see what these results are supposed to be.
For example, you and I agree that reasoning from one non-basic norm doesn’t imply that there will be many non-basic norms that we reason from. And I think that if we reason from one non-basic norm, then doesn’t follow that we will need to poll the public to find out what non-basic norms they accept. So, this isn’t a consequence either. But, then, what are the implausible results?
If we move from reasoning to explanation, perhaps things are different. But then I don’t understand the criticism anymore. That is, if we argue that subjectivism does not explain what it is in vritue of which something is non-basic and wrong, then I agree it doesn’t, subjectivism, relativism, etc., are all rather implausible. But this is a common criticism, not what you had in mind I think.
Mike, the ‘the’ still makes it very hard for me to tell what the statements that contain it are supposed to mean. I can’t tell whether it carries any existential commitment.
Can you state everything without any ‘the’s? Just existential quantifiers when there is existential commitment to norms of one kind or another, and universal quantifiers when not?
Jamie,
You’re putting me to work here. Ok, “de cappotta” with no definite articles.
1. It’s wrong to perform any act prohibited by BN’. From BN’
What exactly are the acts prohibited by BN’ in (1) above? The answer is that BN’ prohibits all and only those acts that the basic norm BN’ itself prohibits (and therefore that “a” basic norm prohibits). This is (2).
2. The acts prohibited by BN’ just are the acts a basic norm (viz., BN’) prohibits.
3. :. It is wrong to perform any act prohibited by a basic norm that prohibits it. From 1, 2
But (3) is just the biconditional in (4),
4. z is wrong iff. z is prohibited by a norm that prohibits z.
The right-hand side of (4) says this,
5. z is prohibited by a norm that prohibits z
Stated formally and carefully, (5) says this,
6. ~(Ez)(En)(norm n prohibits z & z is not prohibited by n)
(6) is equivalent to (7),
7. (Vz)(Vn)(norm n prohibits z only if z is prohibited by n)
And (6) and (7) are trivially true. So the righthand side of (4) is trivially true. But then every action is wrong.
Hm.
There seems to be a scope ambiguity in (4). It could be
(EV) (En)(Vz)(It’s wrong to perform z iff z is prohibited by n)
(Let ‘n’ range over basic norms, ‘z’ over acts.)
That is entailed by (BN’), although it doesn’t entail it. To get something equivalent to (BN’) I guess we’d need
(En)(Vz)[(It’s wrong to perform z iff z is prohibited by n) & n = BN’]
but if you are only trying to show that something follows from (BN’), it will be good enough if you show that it follows from (EV).
But, you say that the right hand side of (4) is (5). That certainly means that my (EV) isn’t what you meant by (4). I don’t see how the right hand side of (4) could be something with two quantifiers in it, actually, since the quantifier for the acts must have scope over the entire conditional.
“That is entailed by (BN’), although it doesn’t entail it. To get something equivalent to (BN’) I guess we’d need
(En)(Vz)[(It’s wrong to perform z iff z is prohibited by n) & n = BN’]”
But the right conjunct is stipulated as true in the hypothesis. We are assuming here that there is only one norm and that that norm is BN’. My claims are about situations of this kind. If you’d like, then, you can add as a premise in the argument the assumption (1a) and the rest goes as before.
1a. (Vn)(n = BN’)
But my point was about the scope of the quantifiers.