Walter Sinnott-Armstrong has recently criticized moral intuitionism by bringing to light some compelling empirical evidence indicating that we are unreliable at forming moral judgments non-inferentially.Footnote 1 The evidence shows that our non-inferentially arrived-at moral convictions are subject to framing effects; that is, they vary depending on how the situation judged is described.Footnote 2 Thomas Nadelhoffer and Adam Feltz, following in Sinnott-Armstrong’s footsteps, have appealed to research indicating that such judgments are also subject to an actor-observer bias; that is, they vary depending on whether the situation judged includes the judger as an actor in, or an observer of, the situation.Footnote 3 The accuracy of this empirical evidence will not be challenged in what follows. What will be called into question is its purported relevance for moral intuitionism. To that end we will consider a version of moral intuitionism developed by Iris Murdoch that not only accommodates but essentially relies on the existence of various kinds of distorting factors including framing effects and actor-observer biases.
Moral intuitionism is threatened by the aforementioned empirical evidence, according to Sinnott-Armstrong, because moral intuitions are supposed to be non-inferentially justified. How can we simply accept intuitions as true without the backing of valid inferences, the thought runs, if we know we are not reliable at forming them? That the moral intuitionist treats moral intuitions as non-inferentially justified is uncontroversial. Less obvious are the reasons for doing so. Sinnott-Armstrong attributes to the intuitionist an epistemological reason for countenancing non-inferentially justified moral intuitions: “To avoid skepticism, moral intuitionists claim that some moral intuitions are justified non-inferentially.”Footnote 4 The intuitionist is characterized here as responding to familiar epistemological concerns about infinite justificatory regression. If an immediate moral conviction must be justified by appeal to other premises, where at least one of those premises is itself a moral conviction, then presumably the justifying moral conviction itself must be justified by further premises, at least one of them moral. And so on ad infinitum. To stop the regress, the intuitionist posits moral convictions that are justified without appeal to further argument.
But avoiding skepticism need not be the point of countenancing non-inferentially justified moral intuitions. The intuitionist’s motivation here might be construed as metaphysical rather than epistemological. More precisely, her aim, like Plato’s, might be to articulate a conception of moral reality as objective and will engaging. Understanding moral truth, on such a conception of moral reality, is a matter of revelation; a witnessing of what is really there to be seen, where seeing is not just believing (in the motivationally-neutral Humean sense) but rather is a character-altering event. If this is the moral intuitionist’s aim, skeptical concerns are at best secondary and so her appeal to moral intuitions is not primarily epistemologically motivated. Once we appreciate this reason for appealing to moral intuitions, it will be less clear that the empirical evidence cited by Sinnott-Armstrong (and supporters including Nadelhoffer and Feltz) threatens moral intuitionism. Let us begin by examining more carefully why he thinks it does.
In his paper “Framing Moral Intuitions,” Sinnott-Armstrong offers the following Master Argument:
(1) If our moral intuitions are formed in circumstances where they are unreliable, and if we ought to know this, then our moral intuitions are not justified without inferential confirmation.
(2) If moral intuitions are subject to framing effects, then they are not reliable in those circumstances.
(3) Moral intuitions are subject to framing effects in many circumstances.
(4) We ought to know this – that is, (3).
(5) Therefore, our moral intuitions in those circumstances are not justified without inferential confirmation.Footnote 5
The Master Argument serves as a template where various types of distorting factors can be plugged into premises (2) and (3). Sinnott-Armstrong plugs in framing effects while Nadelhoffer and Feltz plug in the actor-observer bias. It will be shown that the intuitionist need not accept any of the premises in this argument, given the metaphysical aim attributed to her. And as for the proposed plug-ins, we shall see that framing effects and the actor-observer bias feature as essential components in Murdoch’s moral intuitionism.
Let us begin with the first premise: If our moral intuitions are formed in circumstances where they are unreliable, and if we ought to know this, then our moral intuitions are not justified without inferential confirmation. The intuitionist has a fair question here right off the bat: Who is being referred to in the expression “our moral intuitions”? The intuitionist is surely not committed to the claim that everyone has veridical moral intuitions.Footnote 6 Perfect moral vision is an ideal, not the statistical norm. And for the virtuous—that is, for those who have attained such vision—moral intuitions are always reliable. It is these people who are skilled at discerning exactly what is called for morally in various situations. The moral intuitions of virtuous people set the standard for correct moral vision. No inferential justification is required for their moral intuitions because it is precisely their moral intuitions which reveal moral truth.
In order to get a grip on this notion of revelation by moral intuition, let us consider an example of Murdoch’s.Footnote 7 The example serves to demystify the idea of improving one’s moral vision. But it also serves here to show that the ubiquity of circumstances that tend to skew our moral perception does not tell against moral intuitionism. On the contrary, these circumstances constitute the framework for moral growth. The characters in the example are a mother, M, and daughter-in-law, D. At first, M’s view of D is unsympathetic. D fails to live up to M’s expectations. Murdoch describes M’s initial vision of D as follows:
M finds D quite a good-hearted girl, but while not exactly common yet certainly unpolished and lacking in dignity and refinement. D is inclined to be pert and familiar, insufficiently ceremonious, brusque, sometimes positively rude, always tiresomely juvenile. M does not like D’s accent or the way D dresses. M feels that her son has married beneath him.Footnote 8
But M, being the kind of person who is “capable of self-criticism, capable of careful and just attention to an object which confronts her,” comes to see things differently over time:Footnote 9 “D is discovered to be not vulgar but refreshingly simple, not undignified but spontaneous, not noisy but gay, not tiresomely juvenile but delightfully youthful, and so on.”Footnote 10 This reorientation in vision is a result of an inner monologue aimed at a less selfish conception of things: “I am old-fashioned and conventional. I may be prejudiced and narrow-minded. I may be snobbish. I am certainly jealous. Let me look again.”Footnote 11
Murdoch’s example shows that our tendency to form skewed moral judgments is not at odds with intuitionistic moral realism. Moral vision is sharpened, and moral reality thus brought into view, only by freeing ourselves from selfish biases. M discovers that the features she once thought she saw in D were never really there. Her discovery is essentially a matter of realizing that those features appeared to be there only as a result of her prejudice. Since moral revelation just is a ridding oneself of such bias, our conception of moral reality, for Murdoch, is inextricably bound up with the kinds of distorting factors that might be plugged into premises (2) and (3) of Sinnott-Armstrong’s Master Argument.
Indeed, the case of M and D is a good example of working through framing effects. Sinnott-Armstrong understands framing effects as follows: “A person’s belief is subject to a word framing effect when whether the person holds the belief depends on which words are used to describe what the belief is about.”Footnote 12 This is certainly true of M. Whether M believes that D is an acceptable mate for her son depends on the words she is disposed to use in characterizing D. But rather than telling against moral intuitionism, this phenomenon seems indispensable for a proper understanding of M’s moral achievement. M’s moral progress consists precisely in her re-conceptualizing D, resulting in a more accurate, which is to say less selfish, framing of the situation.
Importantly, the concepts M uses in describing D (“unpolished,” “pert,” “brusque,” etc., at first, and then “refreshingly simple,” “spontaneous,” “delightfully youthful,” etc., afterwards), are thick concepts. These concepts have descriptive and evaluative components inextricably bound up with one another and hence using them engages her will. Her attitude, and so her disposition, towards D changes as her descriptions change. Her ultimate aim, morally speaking, is to have her will engaged by the right concepts; namely, the concepts she would use in the ideal circumstance of having liberated herself completely from selfish fantasy. Her aim is to frame the situation correctly. If and when she does so, it will be moral reality itself, and not the mere appearances constituted by her selfish concerns, that will move her. And if this explanation of M’s moral progress is on the right track, it is not unreasonable to characterize moral reality as objective and will engaging, which is precisely the metaphysical aim attributed to the moral intuitionist at the outset.
To be sure, this is a unique form of moral realism. Murdoch understands realism as a moral achievement. Seeing things as they really are is not a matter of having a neutral conception of things. It is, rather, a matter of having a just and loving conception of things. Such a goal is an ideal and, consequently, the activity of aiming at it is infinitely perfectible. On this understanding of realism, our concern to see moral truth would not even make sense without a background of attempts at overcoming the various selfish distorting factors that all too often influence our perception. Murdoch says of M:
What M is ex hypothesi attempting to do is not just to see D accurately but to see her justly or lovingly . . . M’s activity is essentially something progressive, something infinitely perfectible. So far from claiming for it a sort of infallibility, this [activity of M’s] has built in the notion of a necessary fallibility. M is engaged in an endless task. As soon as we begin to use words such as “love” and “justice” in characterizing M, we introduce into our whole conceptual picture of her situation the idea of progress; that is, the idea of perfection.Footnote 13
Of necessity, for finite beings like us, an infinitely perfectible task will always be riddled with imperfection. But despite its in principle unattainability, the ideal is nonetheless worthy of pursuit for those concerned with moral growth.
If moral judgments are non-inferentially justified only when issued from a selfless perspective, and if attaining such a perspective is an infinitely perfectible task, then it should not be surprising that our actual moral judgments lack such justification as a result of being subject to what Nadelhoffer and Feltz call the actor-observer bias. Nadelhoffer and Feltz understand the actor-observer bias as follows: “[P]eople’s ascriptions of both moral permissibility and control are different when they are asked to imagine themselves facing a dilemma than they are when they are asked to pass judgment on a third party facing the same dilemma.”Footnote 14 So, for example, when considering typical trolley cases (where serious consequences would result from pulling a lever that switches the tracks determining the course of a trolley car), our judgments vary depending on whether we imagine ourselves pulling the lever or whether we imagine someone else doing so. But this is to be expected from those whose moral vision is not yet perfected. It matters more to us if we are the actors in these imagined situations because we tend to give our own parochial perspectives undue weight. The ideal standpoint, however, is selfless. From that perspective, the distinction between actor and observer will not affect moral judgment. The virtuous agent sees things as they really are, which is to say that she sees them without the distortion of selfish bias. We, the not yet virtuous, must approximate as best we can the perspective of the virtuous agent if we wish to see moral reality.
Necessary fallibility is thus part and parcel of the quest for clear moral vision on Murdoch’s conception of moral intuitionism. It is only by liberating ourselves from distorting biases that moral reality is made intelligible to us. Without the biases, without the fallibility, the moral pilgrim’s task has no shape. The ideal result of this task is moral revelation by intuition. Moral intuitions do not need inferential justification precisely because they reveal moral truth. They represent situations as they are, free from the biases of selfish concern. They are the sought-after results of the ideal we strive for when we are concerned to improve ourselves morally.
Since moral intuitions are the product of clear moral vision, it is natural to characterize them as necessarily veridical. Consequently, claims like “moral intuitions are subject to framing effects” and “moral intuitions are subject to the actor-observer bias” are likely to strike the intuitionist as confused. Moral intuitions are what we seek as moral agents. They are the ideal revealers of moral truth in the sense just sketched. Thus characterized, moral intuitions are, by definition, not subject to framing effects, nor to the actor-observer bias, nor to any other distorting factors. This is not to say, of course, that things superficially resembling moral intuitions are subject to such factors. But those things, despite the superficial similarities, are not moral intuitions. We might call them, instead (as done at the outset), non-inferentially arrived-at moral convictions. We all have those and we know, or ought to know, that they are subject to many distorting factors. Such convictions, though they were not arrived at through a process of inference, nonetheless require inferential justification. But moral intuitions do not require such justification because they are had by those whose vision is not distorted by selfish fantasy. Moral intuitions are the product of clear moral vision. It is these intuitions that set the standard for moral judgments.
So, in summary, it is clear that our non-inferentially arrived-at moral convictions are subject to many distorting factors. But it seems that that is the only fact which has been established by the empirical evidence cited by Sinnott-Armstrong and his supporters in the context of his Master Argument. What is unclear is how that fact is supposed to tell against moral intuitionism.Footnote 15
