Stanford rightly emphasizes a crucial distinction between a category of judgments that merely express subjective preferences and a different category of judgments that express norms or normative evaluations. We agree that the latter category marks the core of the interesting, perhaps uniquely human, phenomenon at issue. However, we think Stanford mischaracterizes this category.
Consider the kinds of epistemic norms that govern inferences: the norm that says for stronger inductive inferences you should base your extrapolations on larger samples rather than smaller ones, or the principle of the disjunctive syllogism that tells you if you know that p or q is true, and you know that p is false, then you should accept that q is true. These claims do not express subjective desires or preferences about reasoning, but rather norms: requirements or obligations of good reasoning. Phenomenologically, these can be experienced as being externally imposed upon us, perhaps evincing what Wittgenstein (Reference Wittgenstein1983 p. 352) famously described as the feeling of the “hardness of the logical must.” Yet they are epistemic norms rather than putatively moral ones. The “oughts” apply to inferences, and certainly do not primarily regulate cooperative behavior.
Even in aesthetics, where “beauty is in the eye of the beholder,” there is an intuitive difference between subjective preference and normative evaluation. Take the common idea of a “guilty pleasure”: A person might grant that some catchy ditty on the radio is really not very good, as a work of art, but may nevertheless enjoy listening to it, and even prefer listening to it rather than other types of music (e.g., Coltrane's more abstract explorations) that she herself would rate as better by some normative standard external to, or less subjective than, her own personal preferences. Conversely, a person might appreciate the complex artistry of a songwriter who is drawn to minor keys and sad topics, and feel the force of the claim that by some less ego-centric standard she should prefer it to fizzy pop music. Yet she might simply not like its morose vibe. Subjective desires and normative evaluations are both logically and psychologically distinct, and unlike the former, the latter have this property we will call, as a nod to Wittgenstein, hardness. We hold that this hardness is a property of normative judgments in general, epistemic, aesthetic, prosocial, antisocial, and so forth.
A second distinction, importantly different from the first, marks a difference within the general category of normative judgments. This distinction can capture the idea that (of course?) there is another sense of “objective” according to which aesthetic judgments are widely thought to be more subjective than moral judgments. As Stanford notes, a number of empirical studies have followed Goodwin and Darley (Reference Goodwin and Darley2008) in measuring participants’ judgments about different cases in which people disagree about a normative issue. In this approach, responses are deemed “objectivist” when participants judge that, in cases of disagreement, at least one party to the disagreement must be wrong; conflicting claims cannot both be correct. These studies consistently show that very few people give objectivist responses on aesthetic matters, and people are much more likely to be objectivist about putatively moral issues (Beebe & Sackris Reference Beebe and Sackris2016; Goodwin & Darley Reference Goodwin and Darley2008; Reference Goodwin and Darley2012; Wright et al. Reference Wright, Grandjean and McWhite2013). (Unfortunately, none of these studies has examined objectivism for epistemic matters.) Call normative judgments that have this property objectivist.
We take a different lesson from these studies than Stanford. We see them as showing that some putatively moral judgments are objectivist, but far from all of them are. Stanford claims that “the statement that it is wrong to rob a bank, for example, was reliably judged far more objective than the statement that anonymous giving is good. Such variation has been particularly emphasized by Wright et al. (Reference Wright, Grandjean and McWhite2013; Reference Wright, McWhite, Grandjean, Lombrozo, Nichols and Knobe2014) in work that confirms Goodwin and Darley's central findings” (target article, sect. 2, para. 4). First, we are puzzled by the claim of “far more objective,” because the methodology does not allow respondents to judge degrees of objectivity. Rather, each claim is judged categorically, as either objective or not, in two different ways: In addition to the disagreement question just described, participants were asked whether claims were “true,” “false,” or “just an opinion or attitude.” Second, and more importantly, are the patterns in the data. Wright et al. (Reference Wright, Grandjean and McWhite2013) report that on the putatively moral issue of anonymous giving to charity, only 11% of participants gave the objectivist response on both measures, with only 35% giving the objectivist response on even a single measure. In fact, out of seven claims deemed moral by at least half of the participants, only four were judged objective by more than half on both measures. We see the trend continued in Beebe and Sackris (Reference Beebe and Sackris2016), who measured objectivism only using the disagreement measure, and found that, out of 10 putatively moral issues, only 4 were judged objective by more than half of participants. Three of these 10 issues were judged objective by only 25% of participants or less. What we take the data to show, then, is that a substantial proportion of judgments about putatively moral matters are not judged objective by most people.
What does all this have to do with morality and distinctively moral norms? We agree that understanding human normative cognition will be crucial to understanding human cooperation (cf. Boyd & Richerson Reference Boyd and Richerson2005; Chudek & Henrich Reference Chudek and Henrich2011; Sripada & Stich Reference Sripada, Stich, Carruthers, Laurence and Stich2007). However, we hold that Stanford conflates two different ways normative judgments might be “objectified or externalized,” which we have separated out as the properties of hardness and objectivism. Each is interesting and important on its own terms, but we hold that they do not run in tandem. We also hold that these properties, taken either together or independently, are not unique to what is often pre-theoretically considered morality. So, we continue to be skeptical of attempts like this (cf. Kelly & Stich Reference Kelly, Stich, Carruthers, Laurence and Stich2007; Nado et al. Reference Nado, Kelly, Stich, Symons and Calvo2009) to show there is any empirically important – let alone well-delineated – phenomenon deserving of being partitioned off as morality. No subcategory of norms makes up a psychologically distinctive or cooperatively indispensable set of moral ones.
Stanford rightly emphasizes a crucial distinction between a category of judgments that merely express subjective preferences and a different category of judgments that express norms or normative evaluations. We agree that the latter category marks the core of the interesting, perhaps uniquely human, phenomenon at issue. However, we think Stanford mischaracterizes this category.
Consider the kinds of epistemic norms that govern inferences: the norm that says for stronger inductive inferences you should base your extrapolations on larger samples rather than smaller ones, or the principle of the disjunctive syllogism that tells you if you know that p or q is true, and you know that p is false, then you should accept that q is true. These claims do not express subjective desires or preferences about reasoning, but rather norms: requirements or obligations of good reasoning. Phenomenologically, these can be experienced as being externally imposed upon us, perhaps evincing what Wittgenstein (Reference Wittgenstein1983 p. 352) famously described as the feeling of the “hardness of the logical must.” Yet they are epistemic norms rather than putatively moral ones. The “oughts” apply to inferences, and certainly do not primarily regulate cooperative behavior.
Even in aesthetics, where “beauty is in the eye of the beholder,” there is an intuitive difference between subjective preference and normative evaluation. Take the common idea of a “guilty pleasure”: A person might grant that some catchy ditty on the radio is really not very good, as a work of art, but may nevertheless enjoy listening to it, and even prefer listening to it rather than other types of music (e.g., Coltrane's more abstract explorations) that she herself would rate as better by some normative standard external to, or less subjective than, her own personal preferences. Conversely, a person might appreciate the complex artistry of a songwriter who is drawn to minor keys and sad topics, and feel the force of the claim that by some less ego-centric standard she should prefer it to fizzy pop music. Yet she might simply not like its morose vibe. Subjective desires and normative evaluations are both logically and psychologically distinct, and unlike the former, the latter have this property we will call, as a nod to Wittgenstein, hardness. We hold that this hardness is a property of normative judgments in general, epistemic, aesthetic, prosocial, antisocial, and so forth.
A second distinction, importantly different from the first, marks a difference within the general category of normative judgments. This distinction can capture the idea that (of course?) there is another sense of “objective” according to which aesthetic judgments are widely thought to be more subjective than moral judgments. As Stanford notes, a number of empirical studies have followed Goodwin and Darley (Reference Goodwin and Darley2008) in measuring participants’ judgments about different cases in which people disagree about a normative issue. In this approach, responses are deemed “objectivist” when participants judge that, in cases of disagreement, at least one party to the disagreement must be wrong; conflicting claims cannot both be correct. These studies consistently show that very few people give objectivist responses on aesthetic matters, and people are much more likely to be objectivist about putatively moral issues (Beebe & Sackris Reference Beebe and Sackris2016; Goodwin & Darley Reference Goodwin and Darley2008; Reference Goodwin and Darley2012; Wright et al. Reference Wright, Grandjean and McWhite2013). (Unfortunately, none of these studies has examined objectivism for epistemic matters.) Call normative judgments that have this property objectivist.
We take a different lesson from these studies than Stanford. We see them as showing that some putatively moral judgments are objectivist, but far from all of them are. Stanford claims that “the statement that it is wrong to rob a bank, for example, was reliably judged far more objective than the statement that anonymous giving is good. Such variation has been particularly emphasized by Wright et al. (Reference Wright, Grandjean and McWhite2013; Reference Wright, McWhite, Grandjean, Lombrozo, Nichols and Knobe2014) in work that confirms Goodwin and Darley's central findings” (target article, sect. 2, para. 4). First, we are puzzled by the claim of “far more objective,” because the methodology does not allow respondents to judge degrees of objectivity. Rather, each claim is judged categorically, as either objective or not, in two different ways: In addition to the disagreement question just described, participants were asked whether claims were “true,” “false,” or “just an opinion or attitude.” Second, and more importantly, are the patterns in the data. Wright et al. (Reference Wright, Grandjean and McWhite2013) report that on the putatively moral issue of anonymous giving to charity, only 11% of participants gave the objectivist response on both measures, with only 35% giving the objectivist response on even a single measure. In fact, out of seven claims deemed moral by at least half of the participants, only four were judged objective by more than half on both measures. We see the trend continued in Beebe and Sackris (Reference Beebe and Sackris2016), who measured objectivism only using the disagreement measure, and found that, out of 10 putatively moral issues, only 4 were judged objective by more than half of participants. Three of these 10 issues were judged objective by only 25% of participants or less. What we take the data to show, then, is that a substantial proportion of judgments about putatively moral matters are not judged objective by most people.
What does all this have to do with morality and distinctively moral norms? We agree that understanding human normative cognition will be crucial to understanding human cooperation (cf. Boyd & Richerson Reference Boyd and Richerson2005; Chudek & Henrich Reference Chudek and Henrich2011; Sripada & Stich Reference Sripada, Stich, Carruthers, Laurence and Stich2007). However, we hold that Stanford conflates two different ways normative judgments might be “objectified or externalized,” which we have separated out as the properties of hardness and objectivism. Each is interesting and important on its own terms, but we hold that they do not run in tandem. We also hold that these properties, taken either together or independently, are not unique to what is often pre-theoretically considered morality. So, we continue to be skeptical of attempts like this (cf. Kelly & Stich Reference Kelly, Stich, Carruthers, Laurence and Stich2007; Nado et al. Reference Nado, Kelly, Stich, Symons and Calvo2009) to show there is any empirically important – let alone well-delineated – phenomenon deserving of being partitioned off as morality. No subcategory of norms makes up a psychologically distinctive or cooperatively indispensable set of moral ones.