In this section I discuss Prinz’s answer to a potential objection to his sentimentalism, namely, that the evidence lacks modal strength. In fact, objections of this kind have already been raised against Prinz’s and other naturalistic metaethical theories already. I shall first argue that his answer doesn’t get to the heart of the objection. Second, I propose a way in which Prinz can and should answer it. To do this I shall present two instances where this objection has been made. A helpful distinction by Jon Tresan will then show that there are actually two kinds of internalist theses at play here. Only one of these is really relevant for Prinz’s naturalized metaethics, I shall argue. The objection then loses its force in light of Prinz’s project of a naturalistic methodology. The following reasoning can also be seen as a small case study in recent (naturalized) metaethics.
The first question in Prinz’s decision tree is whether moral judgments are essentially affect-laden or not. This is Prinz’s take on the internalist-externalist debate.[2] This debate is a classical debate in metaethics that can be traced back to the British moralists (Darwall 1995). It concerns the question of whether motivation is internal or external to moral judgments. Do moral judgments necessarily involve motivation to act accordingly? Or does the motivation come from a desire external to them (e.g., the desire to be a good person)?[3]
Prinz advocates a position that he calls sentimentalism:
Sentimentalism =Df Moral Judgments essentially involve affective states, such as emotions, in one of two ways: such states as constituent parts of moral judgments (traditional sentimentalism); or moral judgments are judgments about the appropriateness of such states (neo-sentimentalism). (Prinz this collection, p. 6)
The evidence for a link between moral judgments and emotions is overwhelming (Prinz this collection, p. 10). But is it enough to warrant a stronger relation than mere accompaniment? Even if we grant Prinz the interpretation that affective states are not only mere consequences of moral judgments, could we not still question whether they are essential components of moral judgments? The objection is this: the empirical evidence lacks modal strength to support sentimentalism. Even if all our ordinary moral judgments are based on emotions, it could still be possible to judge dispassionately (Prinz this collection, p. 13). Therefore the evidence doesn’t support sentimentalism.
Prinz answers that the empirical evidence gives us enough reason to infer that we cannot make moral judgments without emotions: “Every study suggests that emotions arise when we make moral judgments. All evidence also suggests that when emotions are eliminated, judgments subside as well” (Prinz this collection, p. 13).
According to Prinz, the theory that emotions are essential components of moral judgments explains the total pattern of data better than its rivals (this collection, p. 14). Furthermore, he argues that the sentimentalist can accept psychologically exotic cases, in which the connection between moral judgments and emotions doesn’t occur, which conform rival theories.
This answer, I argue, misses the real core of the objection. Prinz confronts it upfront and just states what it questions. He puts the objection in the following way:
The evidence shows that emotions are often consulted when making moral judgments, but this leaves open the possibility that we might also make moral judgments dispassionately under circumstances that have not yet been empirically explored. (Prinz this collection, p. 13)
But this does not represent the objection adequately. The objection doesn’t rest on possible, not-yet-found empirical evidence against sentimentalism. Rather, it rests on opposing ideas about what kind of modal strength claims about the relation between moral judgments and emotions should possess. At the heart of this objection there is no disagreement about the empirical evidence, but an opposition in the underlying methodology.
Adina Roskies, for example, accepts that “[…] those [brain] areas involved in moral judgments normally send their output to areas involved in affect, resulting in motives that in some instances cause us to act” (2008, p. 192).
But she thinks that this is not enough for internalism to be true.[4] In her view there is a connection between the cognitive and the affective system, but “this link is causal and thus contingent and not constitutive” (Roskies 2008, p. 192). In this sense the connection, according to her, is not necessary.
Antti Kauppinen sees the difference between internalism and externalism in a similar way. He depicts internalism as saying that there is a link between moral judgments and motivation that holds a priori and with conceptual necessity. externalism, in contrast, is the view that this link is contingent and a posteriori (Kauppinen 2008, p. 3). For Kauppinen, every internalist position then becomes an externalist position if it weakens the modality of the claim. When a metaethical account doesn’t claim that the connection between moral judgments and motivation holds a priori and by necessity, it is an externalist account. No amount of empirical data can refute this criticism.
In Kauppinen’s case the disagreement with Prinz about the underlying methodology is clear. He reacts to the proposal by Roskies, Prinz, and Alfred Mele (among others) that we clarify the debate empirically (Kauppinen 2008, 4). Because of his definitions of internalism and externalism as conceptual necessary claims he argues that “[...] findings in either actual or fictional experimental psychology or neuroscience have little relevance to the debate” (Kauppinen 2008, p. 4).
Kauppinen is opposed to methodological naturalism in philosophical moral psychology (2008, p. 4). That is why he would not be satisfied with Prinz’s answer to this objection. Against him, Prinz would have to defend his metaethical naturalism. Interestingly enough, Roskies, on the other hand, thinks that we can clarify metaethical debates empirically.
In what follows I shall show how I think Prinz should meet this objection. Furthermore, I will argue that everyone who wants to apply empirical data to metaethical debates, such as, e.g., Adina Roskies, should side with Prinz on his methodological naturalism and accept internalism as a true a posteriori theory about moral judgments.
I will now present an analysis of the internalism–externalism debate offered by Jon Tresan that I think will be very helpful here (2009). He distinguishes different formulations of internalism along various dimensions. He claims that a very important distinction has been overlooked: most philosophers in the debate neglected the difference between the modality of the internalist claim and the stated relation between moral opinions and motivation. According to Tresan, there are two different kinds of necessity that can occur in such claims: wide-scope necessity, which operates over the entire proposition—de dicto—and narrow-scope necessity, which operates over the predicate—de re (Tresan 2009, p. 54). The first operates on the dimension of Modality and the second on the dimension of Relation (Tresan 2009, p. 55).
For example, the statement that parents have children can be formulated with both kinds of necessities:
Necessarily, parents have children (de dicto).
Parents have, necessarily, children (de re).
In the first case the proposition that parents have children is stated as holding necessarily. Parents have children, otherwise they would not be called parents. If someone has a child, she is a parent. But the second statement says that people who are parents have their kids necessarily. But this is obviously false. John and Mary don’t have their children necessarily. They could easily never have had any children at all. True, they would not, then, be parents – but the fact that they are parents may have, initially, been quite accidental. We can easily see that there is a difference between de dicto- and de re-necessities because these two statements can have different truth-values at the same time.
With this distinction at hand we can distinguish two different internalist theses: a strong Modality/weak Relation or de dicto-internalism, and a weak Modality/strong Relation or de re-internalism. The former states that, with necessity, there is a connection between moral judgments and motivation. The latter says that there is a necessary connection between these two things.
Tresan uses this distinction to argue that something has gone fundamentally wrong in the internalism–externalism debate. The neglect of the two features has led to the internalist fallacy: the strength in Modality of an internalist claim was taken to be strength in Relation, which led to an overestimation of the epistemic value of the claim (Tresan 2009, p. 55). The classical debate stated the connection between moral judgments and motivation in terms of conceptual necessity (a de dicto-internalism) (see Roskies’s and Kauppinen’s accounts above). Arguments for this claim were supposed to evoke the intuition that no one can make a moral judgment without being motivated to act. If we have such intuitions, the arguments go, the connection is a conceptual necessity. Likewise, arguments against this internalist claim consisted in thought experiments that were supposed to evoke contrary intuitions.
From Tresan’s distinction follows that claims with de dicto necessity are claims about our concepts and not about the subject matter (2009, p. 57). De dicto-internalism, then, is a claim about our concept “moral judgment” and de re-internalism a claim about the subject matter—the phenomenon of moral judgments.
Returning to Prinz (and to Roskies’s proposal), we can now see that there are really two empirical questions we can ask: First, what is our concept of “moral judgment”? And second, what are moral judgments? Traditionally the first was not regarded as an empirical question. Philosophers probed their intuitions and just assumed that others shared them. Prinz, on the other hand, regards these kinds of questions as empirical in nature and presents his own survey studies that probes folk intuitions. He concludes that most people do consider emotions necessary for moral judgments (Prinz this collection, p. 10; for other studies on this with different results see also Nichols 2002, p. 22; Strandberg & Björklund 2013, p. 325; Björnsson et al. 2014, p. 16).
These studies can answer the first question regarding our concept of moral judgments. But, as Prinz rightly points out, people could be wrong (Prinz this collection, p. 10). These studies do not tell us anything about the subject matter. This is a further point Tresan makes. He argues that even if we have internalist intuitions this is not enough to support internalism. He argues that strength in modality is not interesting for a substantial theory of moral opinions. A claim with strong modality doesn’t tell us more about the subject of the claim than the same claim without it. That, necessarily, bachelors are unmarried (de dicto) tells us nothing more than that they need to be unmarried to be called bachelors. It’s a claim about our concept “bachelor”. It tells us simply that the subjects are unmarried—the same as this exact claim without modality tells us. But if bachelors were necessarily unmarried (de re) this would be bad news for the subjects and would tell us something substantial about them—that they’re essentially unmarried, that they, the individuals, are unable to be married. He concludes that “[i]f we are interested in the nature of the Subject Matter, we must look to Relation not Modality” (Tresan 2009, p. 57; emphasis in original).
Only an internalist claim with a strong relation is interesting. But Tresan thinks that there are no arguments for a de re-internalism, which would tell us something interesting and substantial about the subject matter. A de re-internalism that states a strong Relation is wrong. This is because our intuitions regarding moral judgments and motivation can only support a de dicto internalism (Tresan 2009, p. 64). And traditional arguments for internalism provoke only such intuitions.
I think it is clear that Tresan misses one important possible source of evidence for a strong relation: empirical evidence. Here lies the connection to Prinz’s work. The empirical findings, which he collected, all point to a strong relation between moral judgment and affective states. I take Prinz to be looking for a strong Relation when he says that emotions are an “essential component” of moral judgments (Prinz this collection, p. 12).
What I have tried to show here is the following. Prinz raises a potential objection against his own sentimentalism: the relation between moral judgments and emotions lack modal strength. He answers by saying that we have enough evidence to conclude their necessary connection. I argued that this is not a satisfying answer because it misses the core of the objection.
I think the evidence that he has collected points to a strong Relation between moral judgments and affective (motivational) states. Therefore Prinz has an answer to objections that call this strong relation into question. But this is not an answer to an objection that operates with a de dicto internalism. Underlying these objections is an opposition to methodological naturalism in general. Antti Kauppinen is one example of someone holding this position (2008, p. 4). Kauppinen does not think we should ask what moral judgments actually are. In his view, metaethics is concerned with what moral judgments necessarily are. “This takes us from the realm of the actual to the realm of the metaphysical or conceptually possible, and thus beyond the empirical and the observable” (Kauppinen 2008, p. 22).
The evidence that Prinz presents in the target paper doesn’t suffice to refute this position. But I hope to have shown that this need not be a cause of concern for Prinz, because this kind of necessity takes us away from the subject matter. At the heart of Prinz’s account lies an interest in moral judgments as a natural phenomenon that we should study by empirical means.
Adina Roskies, on the other hand, is sympathetic to empirical philosophy. One of her aims in the internalist–externalist debate was to show that “[...] moral philosophy need not be, and perhaps ought not be, exclusively a priori” (Roskies 2003, p. 2003).
But this is in contrast to her understanding of the required modality of the internalist claim, as I tried to show using Tresan’s analysis. If we want to clarify those kinds of debates empirically, it’s not enough to just take traditional philosophical claims and look for evidence in their favor or evidence that can refute them. We have to formulate them as a posteriori synthetic claims that are part of a bigger explanatory project (Björnsson 2002, p. 329).
I hope that this can shed more light on the implications of naturalistic metaethics for philosophical claims. They shouldn’t be regarded as conceptual a priori claims, but as hypotheses that need empirical confirmation. Naturalistic metaethics is not concerned with a priori conceptual necessities. It requires revising our concepts when they don’t fit into the best theories. In that sense empirical philosophers should be revisionists (see Francén 2010, pp. 137 and 142 for a more detailed account of revisionism).
Before I go on, I want to offer one last thought about this. What might be the motivation for framing these positions as claims about conceptual necessity? Roskies writes:
I take it that internalist philosophers have intended to offer something stronger than contingent claims about human wiring (...) Only a view involving necessity or intrinsicality can distinguish moral beliefs and judgments from other types by their special content. (2008, p. 193)
But why do we need a priori conceptual necessities to distinguish between different kind of beliefs and judgments? We could start with very simple observations. Apparently people play a game of blaming and blessing: they use words like “good” and “bad” that are somehow different than other terms. The task of defining what morality is could be a descriptive anthropological enterprise. And I think this is in the spirit of naturalistic metaethics.
I have argued that it is enough for Prinz’s sentimentalism (and for internalism) to claim a strong Relation between moral judgments and emotions. But what kind of Relation is strong enough for it? A mere statistical connection is surely not enough. If the important part of the sentimentalist thesis is not the Modality of the whole claim, we have to analyze the terms “necessary” and “essential” in a non-modal way. One possibility, that harmonizes with naturalized metaethics, is to regard this connection as functional.[5]
In the next, and final, section I shall look at Prinz’s critique of non-cognitivism. I shall present a speculative alternative to his view that I hope, again, is in agreement with his proposal for a naturalized metaethics.