In the past two weeks I have been riddled by a moral dilemma, involving the balancing of my duties to two different groups of people, and though I have recently resolved the dilemma, insofar as I have made a choice of which course of action to take, I still feel guilt and regret for not electing the other option; this is a common phenomena of our moral psychology and I want to consider the implications it might have for metaethical theories. Specifically, I want to consider the expressivist’s explanation of moral dilemma, and the pluralist’s response, and why the pluralist’s explanation is more adequate than the moral monist’s and moral particularist’s.
Williams’ Challenge to Cognitivism
Using moral dilemmas, Bernard Williams advances a powerful argument for expressivism/non-cognitivism, the view that our moral judgments do not state facts, but rather, they express desires, prescriptions or emotions, such that moral judgments cannot be true or false. Williams notes that when we have a moral dilemma, in his sense, there are at least two possible courses of action and it is not the case that we can do both. This is to say the two actions conflict, such that, if each were stated as a belief about what to do, they would be implicitly or explicitly contradictory because both cannot be true — otherwise we could do both actions and there would be no dilemma. Williams astutely notes that when we have two contradictory non-moral beliefs, the strength of one or more of the beliefs is weakened, which is to say, we are less certain that both beliefs are true. By contrast, when we have two conflicting desires, such that both cannot be realized at the same time, our desires are not weakened, but we elect to pursue only one, or find a way to fulfill both. Taking this point further, Williams argues that, in moral conflicts we feel regret about not pursuing the other option, a similarity moral conflict shares with conflicting desires. Therefore, when we face moral conflict the way we go about things is more similar to desires than beliefs, such that moral judgments are not beliefs but desires, and thus, expressivism is true.
It will prove useful to make clear Williams’ argument thusly,
- We experience moral dilemmas, where there are reasons to perform more than one of the actions.
- If morality is belief-based, then when two moral beliefs conflict we are less certain that there are reasons for both actions to be performed.
- It is not the case that when two moral beliefs conflict we are less certain that there are reasons for both actions to be performed.
- Therefore, morality is not belief-based (from modus tollens).
The Cognitivist Response
When we see the argument like this it becomes clear that there are two responses open to the cognitivist. One response might be that we don’t experience real moral dilemmas of the kind in question, where there are reasons to perform more than one action — this line of response is taken by the moral monist and the moral particularist. The other possible response is to deny premise 3, thus accepting the conditional posed in premise 2. The moral pluralist takes this approach; she maintains that when two moral beliefs conflict, there is more reason to perform one action than the other even though there are reasons to perform both.
This leads into the other part of Williams’ argument, that being, that we often experience regret about what we did not do. We see clearly that the moral pluralist accommodates this phenomena in positing that there were reasons for both actions. We feel regret about the action we ended up not performing because there was some reason to perform that action. Because moral monism and moral particularism entail that there are not the sorts of moral dilemmas we experience, they also entail that there is nothing to regret about the action that was not performed. Let us look at why moral monism and moral particularism entail the denial of the sorts of moral dilemmas we experience.
Monism in morality is the view that there is just one fundamental value that is to be captured by a moral principle, and this moral principle captures everything we think is morally significant. For example, utilitarianism is a monistic moral view because there is only one value present in an action that is necessary and sufficient for that action being morally good, and that is its promoting the most well-being of all options. So for the utilitarian, there is no such thing as a moral dilemma because either one of the possible actions promotes the most well-being and thus you should do that action, or two or more options promote the same amount of well-being, and you are thus morally justified in choosing any of those options. If the utilitarian is right then there is nothing to regret about the other options you didn’t pursue; all that matters is the promotion of the greatest amount of net well-being possible, so there is no dilemma.
(The original utilitarian, Bentham)
Moral particularism is the view that a limited number of features of an action cannot determine the moral value of an action, but rather, all the properties of an action together give it its moral value. Particularism, then, is a holistic view, and not an atomistic view, because it is not concerned with certain parts, but with the whole of an action. McNaughton argues that this holistic approach is appropriate because, “we cannot judge the effect of the presence of any one feature in isolation from the effect of others. Whether or not one particular property is morally relevant, and in what way, may depend on the precise nature of the other properties of the action” (David McNaughton, “Principles or Particularism”, in Conduct and Character: Readings in Moral Theory, edited by Mark Timmons, page 341). For the particularist, there are no features of an action that are always a reason for that action to be done. The same properties that were a reason to perform an action in one setting might now be a reason against taking that action in the current setting.
What this means is that there are no moral principles that will provide a decision procedure of how to act in every situation, because moral principles cannot accommodate the specific and unique circumstances of a situation that are essential to deciding what ought to be done. From this it follows that if moral principles fail because they cannot accommodate the specific facts of the situation, and there is only one set of facts for a situation, then, according to moral particularism, there is only one correct action to perform in the situation. If there is only one correct action to perform in a situation, then there cannot be a real moral dilemma, because there is only one right solution to the situation. But if this is so, then there is nothing to regret in not choosing the other options if you choose the one right solution. Thus, for the moral particularist, there are no real moral dilemmas, and there is nothing to regret if one acts appropriately given the facts of the situation.
(Sartre was an early pioneer of particularism)
Therefore, moral monism and moral particularism entail that moral dilemmas do not have the agonizing feeling that we experience, they would be easily resolved if we followed a moral principle correctly (monism) or tended to the specific details of the situation (particularism). What is more, if we follow these prescriptions, then there is nothing to regret in not choosing one of the other options. When we deal with moral psychology the data we are working with is the first-hand qualia we experience when we engage in moral considerations. I submit that the phenomenological data shows that we do experience real and agonizing moral dilemmas, and that we do experience regret about the options we do not choose, even when we know we chose the best option. Since moral monism and moral particularism deny what our phenomenological data shows to be true, we must revise these views, or abandon them for a view that accommodates them. I have suggested already that moral pluralism accommodates this data, so it should be no surprise that I think we should abandon monism and particularism for pluralism.
Moral monism and moral particularism do not make it through the first round against Williams’ argument because they deny the moral phenomenological data that his argument makes use of. Given the option of accepting his argument or denying the phenomenological data, I accept his argument due to the truth of the phenomenological data, and its implications for moral monism and moral particularism. Where I disagree with Williams is in his postulation that only expressive can accommodate this phenomenological data; I have endeavored to show that moral pluralism explains our experience of moral dilemmas and moral regrets. Thus, Williams’ argument does not spell defeat for the cognitivist, so long as he is a pluralist.
Why Non-Cognitivism Fails
While this argument does not turn things in favor of the cognitivist, I want to suggest here that given that phenomenological data is essential to conducting research in metaethics, our general phenomenological data on moral considerations should create a general dismissal of non-cognitivism/expressivism. Non-cogntivism/expressivism holds that moral statements are not truth functional, meaning that they cannot be true or false, because they are declarations of emotions or preferences. For example, when I cry out, “Pie!”, that declaration is not truth-apt because it is not known what would have to be true of the world in order for that declaration to be true. Non-cognitivists/expressivists hold that moral statements are just like ‘Pie!’ in that they do not describe the world, but rather, exclaim something, in such a manner that the statement cannot be true or false. The view is reductionistic, in that it holds that all moral statements are really reducible to exclamations. So for the non-cognitivist/expressivist, when I say “it is morally wrong to steal” what I am really saying is “Boo stealing!”; the former statement can be reduced to the latter in a way that gets us closer to the truth of moral reality.
As Michael Huemer points out, this reductionist thesis is empirical and not a priori, meaning that it is not a conceptual truth that makes it so that moral statements are reducible to exclamations, but rather that we think that experience shows that our moral judgments can be best explained by emotive declarations (Michael Huemer, Ethical Intuitionism page 18).
(Fantastic read, one of my favorite works on metaethics)
This means that, given cognitivism and non-cognitivism are competing claims about the function of moral vocabulary, if non-cognitivism entails that moral statements reduce to confusing exclamations when they are quite intelligible taken as is, which is the view of the cognitivist, then our phenomenological data of how we go about morality is in favor of cognitivism.
The objection that best suits my goals here is the Frege-Geach problem, and it challenges the expressivist to explain how “in some sentences, evaluative terms appear without the speaker’s either endorsing them or impugning anything, yet the terms are used in their normal senses (Michael Huemer, Ethical Intuitionism page 21). The problem is that sentences like “Is stem cell research wrong?” or “if pleasure is good then pornography is good” do not make much sense for the non-cognitivist, as these sentences would reduce to “stem cell research, wrong![?]” and “pleasure yes! pornography yes!”, respectively. But these two questions make plenty of sense to the average person using moral vocabulary, and to the cognitivist, who might be a moral pluralist, which suggests that the experiential evidence of how we actually use moral vocabulary is as the cognitivist posits.
What is more, it seems that the non-cognitivist cannot accommodate those persons who would state they are experiencing a moral dilemma, or regret about the action they did not take. Consider: “Helping my dad is good, and helping my mom is good, but who would it be more good to help?” There is nothing out of the ordinary about this type of moral dilemma, and it makes plenty of sense to you and me. But the non-cognitivist is committed to saying that this statement reduces to “Hooray for helping dad! Hooray for helping mom! Who more hooray?! helping dad?! helping mom?!” This is entirely non-sensical. The non-cognitivist holds that we are questioning and yet exclaiming at the same time when we are faced with a moral dilemma. This is false; the qualia associated with a moral dilemma are not exclamations but inquisitions and puzzlement. Thus, non-cognitivism does not accommodate the phenomenological data we have for moral dilemmas.
If we are to take seriously the phenomenological data we have on moral dilemmas, as I have argued we must in metaethical inquiry, then we must put aside moral monism and moral particularism in favor of views that accommodate the fact that we experience agonizing dilemmas and profound regret despite making good choices. I have suggested that this leaves us with a face off between moral pluralism and general non-cognitivism.
Williams’ argument was that understanding moral judgments as expressions of desires better accounts for the phenomenological data we have on moral dilemma, because if moral judgments are beliefs, then in moments of conflict we would be less certain that both actions have as much reason to be done. The pluralist gladly accepts this, holding that we might have different prima facie duties at stake for different choices, but inter alia only one actual duty after sufficient deliberation. These competing duties account for why moral dilemmas are so puzzling in the moment, and why in hindsight we feel regret for an act we did not do — because there was a prima facie duty to perform that action. Thus, the pluralist can accept that in moments of conflict our certainty that both acts have as much reason to be done weakens, thereby maintaining that moral judgments are beliefs.
I then argued, against Williams, that the phenomenological data tells against non-cognitivism, as that data entails that in moments of moral dilemma I am genuinely questioning what is the correct course of action. Non-cognitivism entails that when faced with a moral dilemma I am all at once affirming and questioning my course of action. This certainly goes against the phenomenological data we have for moral dilemmas. Therefore, Williams faces a dilemma; either we ignore the phenomenological data, in which case his argument does not work against moral monism and moral particularism, or we embrace the phenomenological data and conclude that non-cognitivism/expressivism fails. Thus, Williams’ argument is interesting, as it either comes to nothing or is an argument for what it was meant to negate: cognitivism.