Abstract
Hume seems to claim that there does not exist a valid argument that has all non-ethical sentences as premises and an ethical sentence as its conclusion. Starting with Prior, a number of counterexamples to this claim have been proposed. Unfortunately, all of these proposals are controversial. Even the most plausible have a premise that seems like it might be an ethical sentence or a conclusion that seems like it might be non-ethical. Since it is difficult to tell whether any of these counterexamples are genuine, we need a taxonomy that sorts out ethical sentences from non-ethical ones. We need to know the difference between an ‘Is’ and an ‘Ought’. In the first part of the paper, I establish the need for a taxonomy. I consider some of the most influential ‘Is’–‘Ought’ derivations. These include proposals by Prior and Searle. I argue that each proposal has a premise whose status as ethical or non-ethical is difficult to determine. In the second part of the paper, I consider taxonomies proposed by Karmo and Maitzen. I argue against both taxonomies. I end with the claim that we need a taxonomy of ethical sentences and that none of the current proposals are adequate.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
However, Searle says, “[t]his list will no doubt be somewhat indefinite because the boundaries of the concept of a promise, like the boundaries of most concepts in a natural language, are a bit loose” (1964, p. 45).
As Searle (1964, p.46) points out, “it is possible for all sorts of things to happen which will release one from obligations one has undertaken.” So C* has other members as well.
As Searle points out, there are other conditions that rule out cases in which “something extraneous to the relation of “obligation” to “ought”... interfere[s].”
I must admit that Searle does not explicitly say this. So I am being a little loose with the text here. But I have a justification for this attribution. As we will see later, it is needed to explain why Searle endorses the formulation of ‘other things are equal’ in the first passage and resists the formulation of ‘other things are equal’ in the second passage. Without stipulating some kind of connection like this, Searle’s preference cannot be explained.
So the rough idea is: other things are equal (or other things are equal*) if and only if it is not the case that (i) there is some actual person (or possible person) who judges that ‘Jones ought to pay’ is false and (ii) ‘Jones ought to pay’ is false. I think this is the core idea. But it does not capture all the details.
My interpretation explains why Searle (1964, p. 47) thinks that reading “other things are equal” as other things are equal “justif[ies] [his] claim that the satisfaction of [the ceteris paribus] condition does not necessarily involve anything evaluative.” Recall that the big difference between this way of reading the phrase and reading it as other things are equal* has to do with whether or not you allow possible agents in as candidates for satisfying this requirement. An essential feature of the argument that (5) is ethical was my appeal to the existence of a possible ideal agent. Now, it is not too controversial to suppose that there is a possible agent that is ideal. Using this supposition we could show that (5) looks like it might be ethical. It is much more controversial, however, to say that some actual agent is ideal. For this reason, a similar argument cannot be used to show that (5), on Searle’s preferred reading, is ethical. So my interpretation is able to explain why Searle thinks that reading (5) as saying that other things are equal makes (5) a non-ethical sentence.
The logical-autonomy-of-ethics debate is the debate between those who believe you can derive an ‘ought’ from an ‘is’ and those who believe that you cannot.
Here is Karmo’s proof: “if sentences S1, ..., Sn (where n > 0) entail sentence S(n + 1), then for any possible world W in which S(n + 1) is ethical, if all of S1, ..., Sn are true in w, then at least one of S1, ..., Sn is ethical in w. (Proof: since all of S1, ..., Sn are true in w—that is, are true in w with respect to the correct ethical standard E—S(n + 1) is true in w with respect to E. If S(n + 1) is to be ethical in w, then there must be some alternative ethical standard E’ with respect to which S(n + 1) is false in w. Since replacing E with E’ changes the truth value of S(n + 1) in w, and S(n + 1) is entailed by S1, ..., Sn, replacing E with E’ changes the truth value in w of at least one of S1, ..., Sn; so at least one of S1, ..., Sn is ethical in w.)” (p 256–7).
Recall Prior’s example. If you agree that the sentence ‘Either tea-drinking is common in England or all New-Zealanders ought to be shot’ is either ethical in all contexts or non-ethical in all contexts, then either way you get a valid ‘Is’–‘Ought’ derivation.
You may be worried that if tea drinking is common in England, then some moral property or other is bound to be instantiated. That is probably right. But that does not mean that some given moral property is instantiated. I think, for Maitzen, the idea is something like this: Pick any moral property you want, say wrongness. There is a possible world where nobody does anything wrong. The English all treat each other very well and so wrongness is not instantiated at that world. So a given moral property is not instantiated.
Although there maybe be an ethical standard according to which all New Zealanders ought to be shot, such a horrible standard is, of course, not the correct one.
Well I’m exaggerating a little bit. Maitzen’s taxonomy is only a partial one. His taxonomy tells you when a sentence is definitely not ethical. But it does not tell you when a sentence definitely is ethical.
For another interesting objection to Karmo’s taxonomy see Nelson (2007). He argues against the claim that Karmo’s taxonomy has the result that there is no sound derivation of an ‘ought’ from an ‘is’. Consider this argument:
-
(1)
Stalin authorized the Katyn Massacre.
-
(2)
Therefore, it is not the case that both ‘If Stalin authorized the Katyn Massacre, then Stalin was evil’ and ‘It is not the case that Stalin was evil’.
Nelson argues that Karmo’s taxonomy classifies (1) as non-ethical and (2) as ethical. This is a topic for another paper and there is not space to go into it here. But it seems to me that Karmo’s taxonomy classifies (2) as non-ethical at every world at which (1) is true. So Karmo’s claim that his taxonomy gives the result that there are no sound ‘Is’–‘Ought’ derivations goes unscathed.
-
(1)
To see this, consider two ethical standards, say E and E*. With respect to E, the sentences ‘No ethical sentence is true’ and ‘Torturing babies is wrong’ are both false. With respect to E*, the sentence ‘No ethical sentence is true’ is false and the sentence ‘Torturing babies is wrong’ is true. Now pick any possible world you want, say W. At W, the sentence ‘Either no ethical sentence is true or torturing babies is wrong’ is false with respect to E and true with respect to E*. So the sentence in question is ethical at W. But W can be any possible world. So Maitzen is right to claim that, for Karmo, the sentence ‘Either no ethical sentence is true or torturing babies is wrong’ is ethical at all possible worlds.
Maitzen (1998, p. 357) anticipates this argument and offers a response to it:
Surely, moreover, there is such a thing as the ‘nihilistic ethical standard,’ the standard that assigns ‘false’ to every member of the set U of uncontroversially ethical sentences.... One might reply that some pairs of ethical sentences... are logical contradictories.... [O]f course [‘It is wrong to purchase coffee that is not fairly traded’] and [‘It is not the case that it is wrong to purchase coffee that is not fairly traded’] are contradictories, but [‘It is not the case that it is wrong to purchase coffee that is not fairly traded’] is not then an ethical sentence, for reasons that will emerge shortly.
Maitzen’s formulation of the argument depends on the premise that ∽FT is uncontroversially ethical with respect to Karmo’s taxonomy. Since Maitzen’ doesn’t classify ∽FT as ethical, I concede that it is not. But my version of the argument depends only on the premise that, as I show in the main text, ∽FT is ethical with respect to Karmo’s taxonomy even though it is not uncontroversially ethical.
The problem I have been discussing in this section of the paper is related to what Pigden (2007) calls the Doppelganger Problem. The problem is this: Ethical Nihilism is the thesis that every ethical sentence is false. Suppose that S is an ethical sentence. If S is ethical, then ∽S is ethical too. But either S or ∽S has got to be true. So Ethical Nihilism is false. I propose that we understand Ethical Nihilism in terms of moral properties rather than in terms of which sentences are true and which sentences are false. Pigden deals with this problem differently. He continues to construe Ethical Nihilism in terms of sentences. Pigden’s strategy is to reformulate Ethical Nihilism as a claim about a restricted set of ethical sentences, what he calls “non-negative atomic moral judgments.” These sentences are such that S’s belonging to that restricted set does not entail ∽S’s also belonging to that set. Moreover, if S is a non-negative atomic moral judgment, then ∽S is not. Formulating Ethical Nihilism as the claim that all non-negative atomic moral judgments are false, then, does not get you into a contradiction.
A reviewer gives a response on behalf of Maitzen: Consider the universal being-an-act-of-torturing-babies. This universal exists at W. Consider the property is-a-wrong-making-feature. This is a moral property. What is more, this property is instantiated by being-an-act-of-torturing-babies at W. So it seems like for any world at which (3) is true, the moral property is-a-wrong-making-feature is instantiated. Therefore, Maitzen can classify (3) as an ethical sentence.
I think that this is a really interesting and promising way for Maitzen to respond. I do not have a full counter-response worked out. But let me give you a sketch of a counter-response. I suspect that just as it is difficult to tell, in some cases, whether a sentence is ethical or non-ethical, it will be difficult, in some cases, to tell whether a property is moral or non-moral. I suspect that different plausible answers to the question “What is a moral property?” will give you different answers to the question “Are some properties of universals such as is-a-wrong-making-feature moral properties?” In other words, I think that is—a wrong-making-feature’s status as moral or non-moral is controversial. So we won’t know whether my objection to Maitzen’s taxonomy is successful until we know what counts as a moral property.
Now, Maitzen takes moral properties to be undefined and primitive. He says that “in order to avoid an infinite regress, we may have to take something, perhaps the notion of moral property, as primitive, as undefined, but then any (finite) conceptual analysis must take something as primitive” (Closing the ‘Is’–‘Ought’ Gap, p 362). I have no complaints about taking some concepts to be primitive in an analysis. But I do not think that taking moral properties to be primitive is an option for Maitzen. Here is why: To my mind, the motivation for introducing taxonomies of ethical sentences in the first place is the need to evaluate alleged ‘Is’–‘Ought’ derivations. There are sentences whose status as ethical or nonethical is difficult to determine. We need to know how to classify them. But Maitzen’s proposal defines ethical sentences in terms of something that is equally problematic—moral properties. There are cases, such as the one the reviewer raises, of properties whose status as moral or non-moral is difficult to asses. Until we know what a moral property is, we cannot know whether Maitzen’s taxonomy is adequate and, therefore, we cannot know which way to classify the controversial sentences appearing in Prior and Searle’s ‘Is’–‘Ought’ derivations. In other words, the same problem remerges. So if Maitzen takes moral properties to be primitive, then the motivation for introducing his taxonomy—the ability to evaluate Prior and Searle’s proposals—is undermined. So Maitzen owes us an account of what a moral property is.
In light of the reviewer’s point, I would restate my objection as a challenge. If you can tell me what a moral property is, then I’ll tell you whether Maitzen’s taxonomy is correct. If you cannot tell me what a moral property is, then we had better look for another taxonomy that does not take moral properties to be primitive because otherwise the same problems about determining the status of ethical sentences remerge with determining the status of moral properties. Again, what I have here is just a sketch. I think the reviewer’s comment opens up an interesting line of response for Maitzen. There is more to be said here.
Here is the passage where Maitzen (1998, p. 361) motivates his taxonomy:
Our partitioning of ethical and non-ethical sentences should, among other things, capture the distinction between sentences that are merely about morality (and thus aren’t ethical sentences) and sentences that are genuinely of, or ‘belong to,’ morality (and thus are ethical sentences). [Consider the sentence] Morality often goes by the label ‘ethics.’ While, intuitively, [Torturing babies is wrong] is a sentence of morality [Morality often goes by the label ‘ethics’] is a sentence merely about morality. One way of capturing this distinction is to require that any genuinely ethical sentence entail that a given moral property (such as moral rightness or wrongness) is possessed by at least one object.
Karmo’s taxonomy can capture this distinction too. The truth of ‘Morality often goes by the label ‘ethics’’ does not vary with respect to different ethical standards. So Karmo does not classify this sentence as ethical. It seems to me that this is not enough of a motivation for a taxonomy that makes deriving an ‘Ought’ from an ‘Is’ so easy.
Huemer (p. 82), for example, says that EN “seems to be an evaluative statement.”
Thanks to Pete Graham for pressing me on this.
Prior (1976, p. 89) says that such sentences “belong to the logic of ethics, or ‘deontic logic’ as it is sometimes called, but not to ethics itself.” However, he adds, “It will be appreciated that in piling on these conditions I am making my own job of deducing ethical conclusions a harder one.” Huemer (2007, p. 81) agrees that an interesting ‘Is’–‘Ought’ derivation will not have as its conclusion a necessary truth but still classifies such sentences as ethical. He says:
[T]he conclusion should not be a mere tautology such as ‘Right acts ought to be performed’ or ‘Bad things are bad’ (these statements trivially follow from any set of premises). I assume that an evaluative statement is non-trivial if there is at least one value system that rejects it.”
Geach (1977, p. 476) employs an allegedly necessary truth as a premise in his proposal. He classifies it as an ethical sentence but argues that although the sentence “is indeed an ought thesis,... it is vacuous, it imposes no logical restriction on our views of what in particular we ought or ought not to do.” Of this Huemer (2007, p. 76–7) says:
If [the sentence] were non-evaluative, the counter-example would seem to be in good order, but in fact [it] is evaluative. Nevertheless, Geach observes that [it] is an analytic truth.... So at least Geach seems to have shown how to go from non-evaluative statements and/or tautologies to a non-trivial evaluative conclusion.
So both Geach and Huemer think there are necessarily true ethical sentences.
Searle anticipates the claim that one of the premises in his proposal is an ethical sentence. He (1964, p. 46) says “[My opponent will claim] the derivation rests on the principle that one ought to keep one’s promises and that is a moral principle, hence evaluative. I don’t know whether “one ought to keep one’s promises” is a “moral” principle, but whether or not it is, it is also tautological.” The thing to take from this is that Searle thinks there can be tautological ethical sentences. Here is the upshot of this brief survey: There is a prominent tradition in the literature according to which there exist necessary and impossible ethical sentences.
See footnote 21.
A second problem with appealing to principle (H) is that it is false. (H) is a conditional: If (i) and (ii) are satisfied, then there are no sound ‘Is’–‘Ought’ derivations with respect to really ethical sentences. We can test (H), therefore, by assuming that (H)’s consequent is false and asking whether it follows that (H)’s antecedent is false as well. So assume there are no sound ‘Is’–‘Ought’ derivations. Does it follow that (i) and (ii) are not satisfied? No. I can show this by showing there is some way for (H)’s consequent to be false even though (i) and (ii) are both true. So, by hypothesis, I get to stipulate that there is a sound ‘Is’–‘Ought’ derivation. If I can show there is some scenario according to which this hypothesis does not entail that (i) and (ii) are satisfied, then (H) is false.
So here is my strategy. I am going to state a taxonomy. I am going to come up with a scenario in which (H)’s consequent is false but (H)’s antecedent is true. This will show that principle (H) is false. Here is the taxonomy I propose:
A sentence, S, is ethical** if and only if either (a) S is ethical or (b) S is a premise in a sound argument that has a conclusion that is ethical.
The things Huemer says to defend Karmo’s taxonomy can be used to defend the crazy taxonomy I have proposed. You cannot complain about the fact that the classificatory conditions of really ethical sentences differ from the classificatory conditions of ethical** sentences because my taxonomy is only stipulative. I am just picking out a certain set of sentences and I’m calling them ‘ethical**’. I never claimed that these were one and the same as real ethical sentences. If this kind of reasoning gets Karmo off the hook then it gets me off the hook too.
Now, by hypothesis, there is an argument that is such that its conclusion is ethical and its premises are non-ethical. In order to show that (H) is false, I need to describe a scenario in which there is a sound ‘Is’–‘Ought’ derivation with respect to real ethical sentences but there are no sound ‘Is’–‘Ought’ derivations with respect to ethical** sentences and all ethical sentences are ethical** sentences. In other words I need to show that there is a scenario consistent with our hypothesis in which (i) and (ii) obtain. Here is one such scenario: Suppose there exists a sound argument, A, that is such that A has as its conclusion an ethical sentence and all of A’s premises are ethical**. Now, in order to be counted as ethical** a sentence must either (a) be ethical or (b) be a premise in a sound argument with an ethical conclusion. Suppose that all of A’s premises fail to satisfy (a). So they are not ethical. They only count as ethical** because they satisfy (b). In other words, they are ethical** because they are premises in a sound argument with an ethical conclusion. But the premises are not really ethical. So I have described a situation in which the consequent of (H) is false that is consistent with my crazy taxonomy. Nothing inherent to the taxonomy I have proposed requires that there are sound ‘Is’–‘Ought’ derivations with respect to ethical sentences.
Now I need to show that even though (H)’s consequent is false on this scenario, its antecedent is true. You can show that there are not any sound derivations of ethical** sentences from non-ethical** ones. Here is how: Suppose there is a sound argument, A, that is such that all of A’s premises, P1 and P2 and..., are non-ethical** and A’s conclusion, C, is ethical**. By the definition of ethical**, either C is ethical or C is a premise in a sound argument that has an ethical sentence as its conclusion. Suppose that C is ethical. By the definition of ethical**, P1 and P2 and... are all ethical**. Suppose instead that C is a premise in a sound argument that has an ethical sentence as its conclusion. Then there is a sound argument that has P1 and P2 and... as premises and an ethical sentence as its conclusion. Namely, the argument that derives C from P1 and P2 and... and then goes on to derive an ethical sentence. So, by the definition of ethical**, P1 and P2 and... are all ethical**. Therefore, there is no sound argument, A, that is such that all of A’s premises, P1 and P2 and..., are non-ethical** and A’s conclusion, C, is ethical**. So condition (i) of (H) is satisfied.
You can also show that any sentence that is ethical is also ethical**. After all, any ethical sentence that you pick will satisfy condition (a) of my crazy taxonomy. What is more, satisfying either (a) or (b) is sufficient for a sentence to be classified as ethical**. So all ethical sentences are also ethical**. Condition (ii) of (H) is satisfied. The antecedent of (H) is true. So we have a taxonomy with respect to which (H)’s consequent, there are no sound ‘Is’–‘Ought’ derivations is false but (H)’s antecedent, (i) and (ii) are satisfied, is true. Therefore principle (H) is false.
A reviewer points out that the sort of ‘Is’–‘Ought’ gap that Huemer is interested in is different than the traditional one—the one that I am interested in. Huemer’s project is an epestimelogical one. He aims to show that there is no way to get from foundational empirical knowledge to knowledge of moral claims. We need intuition to account for moral knowledge. If this is right, then perhaps my argument does not bear on Huemer’s project. In that case, I would say the following: It is still interesting to see if Huemer’s point in defense of Karmo can be successfully made in the context of the debate that I am interested in—the traditional ‘Is’–‘Ought’ gap.
I am indebted to Fred Feldman for uncommonly generous comments and guidance on numerous drafts of this paper. I also thank Pete Graham for helpful criticism. In addition I thank Gary Matthews, Lynne Baker, Phil Bricker, and a referee for this journal for comments that improved this paper.
References
Geach, P. T. (1977). Again the logic of “Ought”. Philosophy, 52, 473–476.
Huemer, M. (2007). Ethical intuitionism. Palgrave Macmillan.
Humberstone, I. L. (1982). First steps in philosophical taxonomy. Canadian Journal of Philosophy, 12, 476–478.
Humberstone, I. L. (1996). A study in philosophical taxonomy. Philosophical Studies, 83, 121–169.
Hume, D. [1739] (1992). Treatise of human nature. Reprint, Buffalo, N.Y.: Prometheus.
Karmo, T. (1988). Some valid (but no sound) arguments trivially span the “Is”–“Ought” Gap’. Mind, 97, 252–257.
Maitzen, S. (1998). ‘Closing the “Is”–“Ought” Gap’. Canadian Journal of Philosophy, 28, 349–366.
Maitzen, S. (2006). The impossibility of local skepticism. Philosophia, 34, 453–464.
Nelson, M. T. (2007). More bad news for the logical autonomy of ethics. Canadian Journal of Philosophy, 37, 203–216.
Pigden, C. (1989). Logic and the autonomy of ethics. Australasian Journal of Philosophy, 67, 127–151.
Pigden, C. (2007). Nihilism, nietzsche and the doppelganger problem. Ethical Theory and Moral Practice, 10, 441–456.
Prior, A. N. (1976). The autonomy of ethics. In P. T. Geach, & A. J. P. Kenny (Eds.) Papers in Logic and Ethics (pp. 88–96). London: Duckworth.
Searle, J. R. (1964). How to derive “Ought” from “Is”. Philosophical Review, 73, 43–58.
Schurz, G. (1994). Hume’s is–ought thesis in logics with alethic–deontic bridge principles. Logique et Analyse, 37, 265–293.
Schurz, G. (1997). The Is–Ought problem: An investigation in philosophical logic. Springer.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Hill, S. ‘Is’–‘Ought’ Derivations and Ethical Taxonomies. Philosophia 36, 545–566 (2008). https://doi.org/10.1007/s11406-008-9131-y
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11406-008-9131-y