Undoubtedly, philosophers do make moral judgments about particular cases. For example, they make judgments about actual historical cases – as G.E.M. Anscombe famously judged that it was wrong of President Truman to order the dropping of two atomic bombs on Hiroshima and Nagasaki in 1945.
However, the ethical intuitions that moral philosophers rely on in testing moral theories – intuitions about trolley problems and the like – are not intuitions about particular cases. They are intuitions about types of cases. Admittedly, they are relatively specific types – but they are types nonetheless. As I shall explain, this has some importance for both moral epistemology and the methodology of moral theory.
The “particular cases” that philosophers present us with in attempting to elicit our intuitions are obviously not actual cases – no effort is undertaken to ensure that any such case has ever actually occurred (typically, no historical sources or newspaper reports are ever cited). Indeed, most probably, no case exactly like the one that the philosopher is presenting us with has ever actually occurred.
This does not matter. All that is required is that it is possible – that there should be some non-actual possible world where such a case occurs. However, it seems that it is never feasible for us to pick out a unique non-actual non-possible world. All that we can do is to quantify over such non-actual worlds – that is, to talk about a range of such possible particular cases, or in other words about a general type of case.
For example, suppose that a philosopher presents us with a trolley problem. If this were a particular case, there would be answers to the questions: Who is the particular agent in this case? (Is it you, or me, or President Trump, or who?) At what time did the trolley start running down the track? (Was it on a Tuesday, or a Thursday?) Where did this happen? (In China, or in Italy? In another solar system?)
However, it is clear that these questions are not meant to have any answers at all. In reflecting on the case, we are thinking about a general type of case – a type of case that would be exemplified equally well if a trolley were to start running down a track in China on a Tuesday, or if a numerically distinct but qualitatively similar trolley were to start running down a track in Italy on a Thursday.
It follows that the content of our intuitive judgment on this type of case is already a general moral principle – perhaps a highly specific moral principle about a highly specific type of case, but a general principle nonetheless.
Indeed, arguably, the type of case that we are focusing on is typically not just determined purely by the explicit description that the philosopher has presented us with – since we typically also need to rely on a mass of background beliefs and expectations about how the world normally works to fill in the details of the case.
In effect, then, the way in which we are thinking of the relevant type of case is as the type exemplified by normal cases of that kind (where ‘that kind’ picks out the kind exemplified by all and the only the cases that satisfy the explicit description that the philosopher has presented us with).
Thus, the content of our intuitive judgment is something like the following general principle: “In normal cases of that kind, it’s wrong for the agent of the case to push the bystander into the path of the trolley.”
I believe that several important consequences follow from this point.
- Methodological arguments like those of Peter Singer that rely on the idea of a sharp distinction between “case intuitions” and “principle intuitions” are unsound. The difference is not a difference in kind, but only a difference in degree – the so-called “case intuitions” are relatively specific and concern a narrow range of cases, while the “principle intuitions” are more general and concern a wider range of cases – but in principle there is a continuous spectrum of intuitions, leading from those that focus on highly specific types of case and those that focus on highly general types of case.
- A perceptual model for moral intuitions seems highly questionable – because perception by its nature is an awareness of a genuinely particular state of affairs, and not of some general truth about a general type of case. In crucial ways, ethical intuition must work differently from sensory perception.
- More generally, any moral philosophy that tries to build on such intuitions, and views them as akin to the everyday moral thinking of ordinary people, is committed to rejecting particularism (of the sort that has been defended by Jonathan Dancy). Since the contents of these intuitions are general principles, friends of these intuitions must reject the particularist assumptions that we should not expect to be able to articulate true non-trivial general principles, and that these principles play no important role in moral thinking.
“A perceptual model for moral intuitions seems highly questionable – because perception by its nature is an awareness of a genuinely particular state of affairs, and not of some general truth about a general type of case. In crucial ways, ethical intuition must work differently from sensory perception.”
It doesn’t seem to me that this follows directly from your main point. After all, our visual imaginings are often not about actual cases in your sense; yet, it is very plausible that the epistemic power of such imagining is derived in some way from their connection to perception. What seems important in the visual imagining case is that one imagines *some* things that are sensorily determinate in a certain way. The defender of a perceptual account of intuition should hold an analogous position. Of course it is not specified if the trolley is in Italy, but one does need to imagine certain determinate features in order to induce intuition.
Might it not be that these judgments are of particulars, but of the particulars across all the possible worlds we are picking out via our description of a case? I agree we can also think of the judgment as (or as corresponding to) a principle. I like to think of them as “That’s It!” principles: “X, Y, Z, and that’s it for morally relevant features!”
I had a similar worry to Errol’s above. It seems to me that a perceptual model for moral intuitions (in a broadly Aristotelian vein like McDowell’s or Platonic like Murdoch’s; they come to the same thing WRT this issue I think) will hold that in the first instance it is particular cases we are aware of, and we move from those to more general principles by abstraction or by reflection on what is at work in particular cases. Maybe this isn’t the sort of perceptual model you were thinking of? The kind I have in mind tends to go closely with particularism, or something like it. Even still, it is consistent with thinking that intuitions about trolly problems are probative insofar as they help elucidate heuristics that are more or less reliable, or defeasible principles, so I also am not sure I buy the third point either.
The general thought here is totally persuasive to me! I just think it cuts in favor of a particularistically-flavored, perceptual model of moral knowledge like the kind McDowell and Murdoch endorse.
Thanks for the interesting post, Ralph!
One lesson I’m tempted to draw from the history of substantive normative theorizing (“first-order” ethical theorizing) is that the true, _complete_ normative theory that tells us what we ought to do in any case is either a highly complex version of one of the major traditions (e.g. consequentialism, contractualism, etc.) with which ethicists have spent a lot of time wrestling or some radically different theory that isn’t even on our collective radar yet. If the true normative theory takes the former shape, then we might worry, assuming our intuitions about cases can even have all of it as its content, that we can’t articulate it. If it takes the latter shape, it also seems to me, again assuming our intuitions can even have all of the true normative theory as its content, that we can’t articulate it. Either way, ethicists can’t articulate the true, _complete_ normative theory – we can’t capture it in the form of some principle that is stated explicitly.
I suppose that makes me a kind of particularist. But I’m not sure, and I’d be interested to hear what you think, that it’s a version of particularism that is incompatible with your suggestion that the content of our intuitive judgments about cases takes roughly the form “In normal cases of that kind, it’s wrong for the agent…”.
Thanks, Errol, Daniel, Jack, and Nick!
I’ll start by responding to the issue that was raised by Errol and Jack. When I said “a perceptual model for moral intuitions seems highly questionable”, all that I meant was that a perceptual model of *these* ethical intuitions (i.e. intuitions about purely hypothetical cases) would be highly questionable.
Admittedly, perhaps that point is too obvious to be interesting — and so it was quite reasonable of you to take me to be making a more general claim, attacking the very possibility of moral perception. As it happens, I have complicated views about moral perception (I think it’s possible, but not epistemically fundamental), but I can’t really get into that issue here.
However, I am sceptical of the particular perceptual model that you describe, on which the epistemic power of ethical intuitions is derivative from that of moral perception of actual particular cases, in just the same way as the epistemic power of visual imagining is derivative from that of visual perception. I don’t think that we can learn many important new empirical truths about the world from mere imaginative visualizing — whereas we can learn lots of new ethical truths by canvasing ethical intuitions…
This is partly because practical reasoning itself crucially involves forming ethical judgments about hypothetical cases — viz. about the acts that one decides not to perform. One needs to be able to judge that if φ-ed, one’s φ-ing would wrong, even if one never φ-s, and so never perceives the wrongness of one’s φ-ing.
Incidentally, I am also pretty sure that McDowell and Murdoch misread Aristotle’s talk of “perception” in the discussions of practical reasoning in e.g. Nicomachean Ethics Book 6. I think that the kind of “perception” that he is talking in those passages about consists in (a) your noticing certain opportunities (i.e. that certain practical options are available), and (b) the right questions about these opportunities’ occurring to you (e.g. questions of the form ‘If I φ-ed, would my φ-ing be F?’). I don’t believe that we have to interpret Aristotle as believing in literal moral perception!
Daniel, thanks for your comment! Yes, in general, I am very sympathetic to your view. But I don’t think that it is possible for us to pick out a unique particular non-actual possible case in the way that you describe.
Suppose that I were try to specify a case by picking out a bunch of particulars: “Imagine that mayor Eric Garcetti sees trolley #5472 running down track #5 at Union Station at 12 noon on Sunday 17th February 2019 …” Even so, there could be lots of different ways in which a case of this sort could happen (different velocities for the trolley, and so on), and these would be slightly different possible cases. It also seems of course that naming the agent, the time, the place, and so on, in this way would be a mistake, because these are essentially irrelevant details.
So, I am inclined to think that the only way in which we can pick out a genuinely particular case is by picking out an actual historical case (like Truman’s dropping the bombs on Hiroshima and Nagasaki).
Thanks, Nick!
I agree with almost everything that you say here. I.e., I agree with you that (1) the true complete moral theory will have a non-trivial degree of complexity, (2) in at least certain respects, it will be probably differ from all the moral theories that dominate the contemporary debate, and (3) we will never know exactly what this true complete moral theory is in all details.
However, I don’t see why this makes me a “particularist” (at least not in the same sort of way as Jonathan Dancy). Just because we can’t know *all* moral principles, it doesn’t follow that we can’t know *some* moral principles. And I think that we do know quite a few moral principles.
It’s true that we can’t reduce our ethical competence to knowledge of such principles, since we can come to know new moral principles without simply inferring from them the principles that we already know. But you don’t have to be a particularist to think that our ethical competence consists of a lot more than just knowledge of principles.
Although I am interested in modeling the epistemology of thinking about cases on the epistemology of sensory imagining, my initial point was much more modest. It doesn’t seem to me that a perceptualist account of intuition is troubled by the fact that we don’t imagine unique worlds–i.e., fully determinate worlds. When it comes to my initial point, the visual imagination was just supposed to support this; I wasn’t trying to argue that intuitions about cases are the output of off-line perceptual processes (although I am sympathetic to that). You can think that the epistemic power of some states that take as their objects less than fully determinate worlds is explained on a perceptualist model. That is, at a minimum, what ordinary visual imagining shows.
(Of course, you *inferred* from the claim that the worlds we think about are less than fully determinate that our intuitions are just about principles. And *that* immediately clashes with perceptual accounts of intuitions about cases. One way of thinking about my point is that it questions this inference. You might have intuitions about determinate parts of less than fully determinate worlds, and they might be explained on a perceptualist model.)
Errol, I am having a really hard time interpreting what you’re saying here….
If I understand you correctly, you are conceding that real perception is of genuinely particular states of affairs (i.e., of states of affairs that cannot be exemplified more than once). You are also conceding that we can’t pick out any unique non-actual possible world. If that is true, then surely it is equally plausible that we can’t pick out any unique non-actual possible individual. (E.g. I don’t think it is possible for us to pick out a particular non-actual possible person. We can think of a possibility in which there is some person or other x who meets condition C, but then it’s clear that there is *also* a possibility of there being some *other* person y distinct from x who meets condition C…)
So, I take it that you’re also conceding that the content of these ethical intuitions is, as I say, a general principle, such as “In every normal case of that kind, it is wrong for the agent of the case to push the bystander into the path of the trolley”.
But you then imply that “the epistemic power” of “states that take” such general principles “as their objects” can plausibly be “explained on a perceptualist model.” Surely, this is incompatible with conceding that all real perception is of genuinely particular states of affairs?
Sorry for not being clear. The issues are complicated and it’s always difficult to try to have these conversations via blog.
One more try. Hopefully we can talk in a more efficient way IRL some time.
Why think *individuals* are what matter? In the case of visual imaginings, we can easily learn about qualitative properties by visualizing them. You can visualize a coke red armchair and learn things (e.g., that nothing could be colored *like that* and also be forest green). You don’t visualize a particular individual chair, sure. But that doesn’t mean that the sensory nature of your state is irrelevant to the story of how you learn. I think it would be bordering on absurd to say that what is really going on is that you are having some non-perceptual intuition about general principles like ‘For every normal chair like this, it can’t be coke red and forest green.’ So, in general, it doesn’t look great to infer the claim about principles from the claim of not picking out particular individuals. That’s another thing visual imaginings show.
Now, I didn’t say how to make the ethical case analogous to this. My starting suggestion is to focus on the qualitative nature of what we can imagine and how that connects to the ethical features of things. If our intuitions are reactions to and in some sense explained by our capacities to be sensitive to those connections, then it seems to me that one could defend a perceptualist view while acknowledging both of the claims that look to be your main premises (first, that we don’t imagine fully determinate worlds and, second, that we don’t imagine individuals).
So, to answer your questions:
“So, I take it that you’re also conceding that the content of these ethical intuitions is, as I say, a general principle, such as “In every normal case of that kind, it is wrong for the agent of the case to push the bystander into the path of the trolley””
No, I’m not conceding that. It’s a false dilemma to say that we are either related to individuals or we are related to principles.
“But you then imply that “the epistemic power” of “states that take” such general principles “as their objects” can plausibly be “explained on a perceptualist model.” Surely, this is incompatible with conceding that all real perception is of genuinely particular states of affairs?”
No, it’s not incompatible with conceding that all real perception is of genuinely particular state of affairs. At least, it’s not incompatible with conceding the true version of this, where ‘perception’ is used as a success term. The key is understanding the nature of the content we entertain when we visually imagine. It is true that it is different than the content of our perceptions (again, when ‘perception’ is used as a success term). But that doesn’t mean it is not sensory. Nor does it mean that its epistemic power is completely divorced from perception.
I realize that that probably looks mysterious. That’s fair. My views on these things are tied up in much more general views about sensory content and how that relates to imagining, perception, hallucination, and many more things. Being able to clearly articulate in blog comments is hard!
Errol,
Surely my intuition about colour exclusion *is* essentially general! Its content surely is something like “Nothing can be both red all over and green all over” (i.e. an essentially quantified proposition). Even if I get to that intuition by imagining “an arbitrary object”, it’s a strange view (defended to my knowledge only by Kit Fine) that such “arbitrary objects” really exist, and are thought about by us.
Most philosophers would say that, really, in your example, your visualizing a chair is just a way for you to conceive of an essentially general proposition, such as the proposition that there is some chair or other that that is both red all over and green all over — and this conceiving then enables me to intuit the impossibility of this proposition.
Propositions, as I see it, are either particular — they concern individual people, places, and times, or the like — or they are general — they quantify over such individuals, people, places and times. Your mysterious “third way” is positively baffling to me.
It also seems to me most unwise to hitch your moral epistemology to a Kit-Fine-like view of the feasibility of thinking of “arbitrary cases” that are somehow neither particular (concerned with individuals) nor general…
Wait, are you denying you can visualize a coke red chair that isn’t a particular chair? Or are you saying that the coke red chair you visualize is really just a proposition? I certainly hope most philosophers don’t endorse either of those claims.
There’s a pretty big literature about visual imagining, and many people endorse the basic picture I was sketching. See, e.g., the work of Amy Kind, Peter Kung, Dominic Gregory, and Magdalena Balcerak Jackson. Tim Williamson doesn’t explicitly endorse the stuff about content, but his view about offline processing is often taken as a version of this view.
In a different corner of the philosophy of mind, Mark Johnston defends a view about the object of hallucination that is just like the view I was sketching. Susanna Schellenberg defends a related view. Elijah Chudnoff and John Bengson have discussed this type of view in the context of the literature on intuitions.
In “The Obscure Object of Hallucination”, doesn’t Mark Johnston defend the view that the content of hallucination is general — the complex property that your hallucinatory experience represents your environment as instantiating (whereas with genuine perception, the content is the particular instantiation of that complex property)?
I’m afraid don’t know the work on visual imagining of Amy Kind, Peter Kung, Dominic Gregory, and Magdalena Belcerak Jackson. But the only sense that I can make of the view that you’re ascribing to them is as a version of Kit Fine’s view.
If the content of my visual imagining really were something like ‘That chair is red’, we could ask, “Which chair? Where was it made?” etc.”If the answer is, “There’s no answer to the question of where it was made”, we’re committed to a bizarre view like “There is a chair such that there is no answer to the question of where it was made”.
This, I take it, is Kit Fine’s view. Is this really what Amy Kind, Peter Kung, Dominic Gregory, and Magdalena Belcerak Jackson all think?!
A much better view, surely, is a view like Mark Johnston’s, that the content of the imagining is something like a complex general property — the property of being a chair with such-and-such features. But that is exactly what I was saying about ethical intuitions.
Yes, on Johnston’s view, we don’t hallucinate individuals. But we are aware of things that have a qualitative nature. We are aware of sensible profiles. These are complexes of universals, but they are essentially qualitative and we are acquainted with their qualitative nature when we hallucinate them. On Johnston’s sort of view, we can have perceptual experiences (not in the success sense) that have as their objects chairs when no such individual chair exists (that’s because chair-ness has a qualitative nature, at least on Johnston’s view). (Also, fwiw, he doesn’t think representation has anything to do with it. We are directly acquainted with sensible profiles when we hallucinate and directly acquainted with individuals when we see.)
Of course, the people in the visual imagining literature don’t necessarily buy into all of this. But they do buy into the thought that we can visualize qualitative properties and combinations of them that add up to visualizations of possible scenes (which sometimes include coke red chairs even though they don’t pick out any individual coke red chairs). The qualitative nature of these states and their source in perceptual capacities is an important part of the epistemology of the imagination these people defend. My point is that the defender of a perceptualist account of intuition should look to these sorts of states to avoid your objections.
Errol,
My point is exactly that these “sensible profiles” and “complexes of universals” are general properties or universals or types. (Of course, these general properties / universals / types also have “qualities” or “qualitative natures”: properties have properties and stand in relations, after all…) So the “possible scenes” that we’re imagining are also, strictly speaking, general properties or types or universals. But any view that accepts that the objects of these states are general properties or universals or types *agrees* with what I was saying. I was not in any way objecting to this view.
I am also not objecting to the view that ethical intuitions have a phenomenological or qualitative character. (Indeed, given my own views about their connection to our emotional dispositions, I actually think that they do in a way have such a phenomenological character.) I was objecting to the idea that these states are “perceptual” in the strict sense on which perception is necessarily of particular states of affairs.
As a technical matter, I think it’s incredibly misleading to say that in hallucinating or visual imagining you are “aware of chairs” — since in these cases, there need be no chairs such that you are aware of them. After all, you might visualize or hallucinate a unicorn, and this visualization or hallucination is only in a very weird sense an awareness of unicorns! When we say ‘You are visualizing a unicorn’, the phrase ‘a unicorn has to be taken as occurring in an intensional context. The real content of your visualization is something like a general property of a situation, such as the property of being a situation involving unicorns.
Furthermore, if we agree that the content of ethical intuitions is — like the content of visual imagining — essentially a general property or universal or type, then I just don’t see any room to deny that if the content of the ethical intuition is a proposition at all, it must be a general proposition or principle like “In every normal case of that kind, it is wrong for the agent of the case to push the bystander into the path of the trolley”.
There are no non-general propositions that could with any plausibility be taken to be the content of the intuition. If we try to take the content to be “It is wrong for that agent to push that bystander into the path of that trolley”, we can’t escape the weird questions like “Which agent? Who is the bystander? Where is the trolley’s serial number, and where was it made?”!
Ralph, in general I find your view plausible, but I fear you have exaggerated its novelty somewhat. RM Hare made essentially the same point in his essay on “Universalizability” (found in his _Essays on Ethical Theory_), noting in particular that the idea, sometimes suggested by people who tend towards a particularist way of thinking (casuists, virtue theorists, etc.), that the rich description of particular cases in novels makes them exemplary tools for moral education by inciting such intuitions, ironically defeats its purpose, since works of fiction are by definition about non-particular persons/situations, and hence can our moral judgments about them can only make sense if there are some kind of universal moral principles.
I’m also confused about how you think you are disagreeing with Singer (note: a student of Hare). You don’t cite the source for his distinction, but a quick google search suggests that this is raised in his remarks in _Singer and His Critics_. Where, not surprisingly, he thoroughly rejects the idea of reaching sound moral conclusions via (1) intuitions about “concrete cases,” e.g., “case intuitions,” rather than (2) intuitions about general principles (which he also thinks need to be tested by reasoning, of course). It is not clear to me that this is exactly what you mean by (3) “intuitions about particular cases,” since you seem to suggest that this (non-existent thing) would be some direct perception of a moral quality in a particular instance, and you claim that Singer’s (1) is simply the intuition of more specific principles than in (2). But if 1 is not 3, then your rejection of 3 (which Singer would also surely reject as a basis for sound moral reasoning) seems orthogonal to his distinction between 1 & 2. Since some people surely do think that the more particular your case-description, the more likely your intuitions about it are true/foundational/useful/etc., then it seems useful for Singer to reject this claim, and I see nothing in your remarks to suggest that you would disagree with him on this point, either, though it is possible that you do not. Perhaps (indeed: very likely) I am misunderstanding what you think Singer is saying about (1); but if so, I hope you can clarify this and its relevance to your point.
Thanks, Scott!
I actually never claimed that my point was “novel” (with hastily written blog posts, it happens from time to time that one makes a point without knowing that it has already been made by another philosopher…). But thank you so much for reminding me about R.M. Hare’s point from “Universalizability”.
Hare’s point is certainly similar to mine: he rightly emphasizes that there are no particular situations that fictions (unlike e.g. historical novels) are about — and I guess that we can take the hypothetical cases that many moral philosophers like to describe as, in effect, short little fictions. I think, however, that Hare draws a slightly different conclusion from mine. I am conceding that we can actually learn moral truths from such fictions, and then arguing that these moral truths must be general principles (rather than any particular moral truth about an individual case). I could be misremembering him, but Hare seems to be arguing that we can’t really learn anything from considering such fictions, because we must already have some moral principles to guide us in responding to them. So, there is still something modestly novel about my point, I hope!
On Singer, you’re quite right: I am thinking of his “Response” in the volume “Singer and his Critics”. In this response, Singer argues forcefully against “the view that we should take our intuitive responses to particular cases as the test of a sound theory” (p. 316). My question is, What does he even mean by talking about “our intuitive responses to particular cases”?
There are real “responses to particular cases” — like G.E.M. Anscombe’s response to Truman’s dropping the atomic bombs on Japan. But it looks as if what Singer is referring to is actually not these responses to actual historical cases, but rather our intuitive response to considering somewhat specific types of cases. As I try to show, however, our intuitive responses to specific types of cases actually consist in our acceptance of (somewhat specific and limited) general principles.
So, the only difference between our “intuitive responses to specific types of cases” (which Singer criticizes) and our reliance on intuitively plausible general principles (which he accepts as methodologically sound) is the degree of generality — the latter are more general than the former. But it is far from clear that the arguments that he raises show why the more specific principles that we find intuitively plausible should be any less reliable than the less specific or more general ones.