32 -ETHICS AND PSYCHOLOGY
Jesse Prinz
Historically, philosophers have had a great deal of interest in science. Thinkers as diverse as Aristotle, Descartes, and Berkeley made important contributions to a range of scientific fields. In more recent times, however, philosophers have often had an anti-scientific {or perhaps trans-scientific) orientation. Nowhere is this attitude more keenly felt than in moral philosophy. Here, it is sometimes suggested that the very nature of the subject matter defies empirical inquiry. Morality is normative. It describes how things should be, not how they are. And moral rules, like rules of logic, are necessary, unlike the contingent regularities with which scientists are typically preoccupied. Kant (1998/1785: Preface) expresses this attitude in an influential passage:
Now it is only a pure philosophy that we can look for the mora! law in its purity and genuineness. ... That which mingles these pure principles with the empirical does not deserve the name of ... moral philosophy, since by this confusion it even spoils the purity of morals themselves, and counteracts its own end.
Despite this widespread view, social scientists have recently taken a serious interest in morality. Whatever one wants to say about normativity, there are obviously aspects of human behavior that issue from the moral values we hold dear, and these can be empirically investigated. Philosophers who study this work (and, at times, contribute to it) arc coming to realize that psychological findings may actually bear on philosophical theories (Flanagan 1991; Doris 2002; Nichols 2004; Sripada and Stich 2005; Prinz 2007; Sinnott-Armstrong 2009/ 2008). That will be the claim defended and illustrated in this chapter.
The subject of moral philosophy has traditional subdivisions. Some study moral psychology (the way people think about the moral domain), others study meta-ethics (the ultimate metaphysical basis of our moral claims), and othets study normative ethics (the question of what we ought to do or how we ought to be). Now it might be taken as obvious that scientific psychology can contribute to moral psychology. But even here there is some resistance, as we will see. Less
obvious is the since meta-eth not the convk entific psychol the business c psychology co: discussion on t psychology. R; context, theop
In
Moral psychol Broad topics ir Many of thes. literature on r. Aristotle write: in the area, La series of devel (from Aristotk chology that h< these cases, the especially helpl turn on concep specialize in cc ceptual dimens empirical evide
To make this the debate be externalist says when one mak accordance wit! there is a neces times this conn cally motivating state that dispc there may be c experiencing or forms of practk internalist (Smii
On the face c moral judgment
ETHICS AND PSYCHOLOGY
DGY
e. Thinkers ri tuitions to >phers have Nowhere is sometimes :al inquiry. :y are. And regularities 5: Preface)
Jaw in "iciples sophy, selves,
a serious there are s we hold study this :hological 3ris 2002; >ng 2009/
■ne study ers study id others ought to mtribute see. Less
bvious is the contention that scientific psychology can contribute to meta-ethics, nee meta-ethics concerns what, if anything, makes our moral convictions true, ot the convictions themselves. And equally controversial is the claim that sci-lentitle psychology can contribute to normative ethics, because psychology is in [ the business of description, not prescription. I will try to show that scientific [psychology contributes to all three subdivisions, though I will spend most of the [discussion on the first. My claim will not be that philosophy should be replaced by [[psychology. Rather, I think the two work in concert. To echo Kant in another context, theory without data is empty, and data without theory are blind.
Internalism, externalism, and empirical inquiry
J Moral psychology is the study of psychological states associated with morality.
Broad topics include moral motivation, emotion, deliberation, and development. [ Many of these are studied in psychology. For example, there is a massive I literature on moral development, which has clear connections to philosophy: [Aristotle writes about moral development, and the most influential psychologist in the area, Lawrence Kohlberg (1984), argued that children progress through a [series of developmental stages that mirror major theories in normative ethics [(from Aristotle to Mill to Kant). But there are also some topics in moral psychology that have been dominated by philosophical discussion, and, in some of these cases, there is an implicit assumption that empirical psychology may not be especially helpful. This is especially the case when issues in moral psychology turn on conceptual claims. It is philosophers, not psychologists, who purport to Specialize in conceptual analysis. Claims in moral psychology that have a conceptual dimension have, therefore, been approached without drawing heavily on empirical evidence. I think this is a mistake.
[ To make this case, I will focus on a central controversy in moral psychology: the debate between motive externalists and motive internalists. A motive externalist says that motivation is independent from moral judgment, so that when one makes a moral judgment one is not, thereby, motivated to act in accordance with that judgment. Motive internalists, on the other hand, say that there is a necessary connection between moral judgment and motivation. Sometimes this connection is presented as the view that moral judgments are intrinsically motivating: if I judge that charity is good, I am thereby in a motivational state that disposes me to give to charity. This link may be defeasible, because there may be countervailing interests that prevent one from acting on or even experiencing one's moral motivations. But, barring weakness of will or other forms of practical irrationality, moral judgments compel action, according to the internalist (Smith 1994).
On the face of it, this may look like a straightforward empirical debate. Either moral judgments motivate, or they do not, and whether they do can be tested in
385
JESSE PRINZ
a psychology lab. But philosophers who have weighed in on the debate have rarely looked at empirical psychology. Let's put aside the possibility that philosophers are lazy, methodologically reckless, ill-equipped to understand psychology, or irrationally biased against other fields. Philosophers think there are two good reasons to approach the debate from the armchair. First, they note that the debate is conceptual. Internalists claim that there is a necessary connection between morality and motivation; one could not make a moral judgment without being disposed to act. This modal claim is supposed to derive from a conceptual truth - something about our moral concept. The concept of moral goodness is supposed to entail something relating to motivation, and conceptual truths are best discovered using conceptual analysis, rather than empirical observation. Call the view expressed in the last sentence the conceptual thesis.
The second reason for resisting empirical approaches is that the opposing views may make similar empirical predictions. Externalists admit that moral judgments arc typically associated with motivational states. Most of us desire to act in accordance with morality, so when we make moral judgments we are motivated. They simply claim that this link is causal rather than constitutive. Likewise, internalists admit that we are not always practically rational. As a result, the motivational states that should come along with moral judgments can fail to arise. The connection is dispositional and the dispositions are not realized in every case. Thus, empirical evidence showing that motivations arise in the context of moral judgment would not entail internalism, and evidence to the contrary would not entail externalism. Call this the empirical intractability thesis. The conceptual thesis relates to the empirical intractability thesis in the following way. According to the conceptual thesis, the debate between internalists and externalists is a conceptual or semantic debate, concerning the meaning or moral concepts. Psychology tells us about causes and correlations, not conceptual constituency. Therefore, psychological findings just can't settle the debate.
Let me address these two concerns in turn. First, consider the conceptual thesis, which says that the debate in question should be investigated using conceptual analysis rather than empirical methods. This is problematic for two reasons. First, the contrast between conceptual and empirical matters is a version of the analytic/synthetic distinction. To say that a debate is conceptual is to say that it is a debate about the analytic entailment of a concept. Since Quine's (1953) critique, the notion of analyticity has been called into question. While it is often the case that people understand one concept by appeal to another, these associations are characteristically revisable in light of empirical evidence. Thus, scientific discoveries might lead me to believe that aardvarks are not animals (perhaps they are robots sent to spy on us from another planet), red is not a color (rather it is an experience caused in me by a colorless world), and some bachelors are not male (there may be people with XX chromosomes who look and act male, but are really women). Concepts are something like mini-theories that correspond to our best guess about how the world is, and they are subject
386
to revision. If n vating, this is nc about morality, conceptual anah, Now suppose problem with tr ceptual truths si trast, that conce concepts (accorc tive science) an empirically stud studies concepts methods can't b concept entails, when those com Someone migl concepts than p fully constructd those features tr really essential t< methods are a f Philosophers rei observing one's ments can be gi\ philosophers re theory neutral, intuitions about themselves from ing statistical s reporting intuitii about conceptut what a concept < features are necc about concepts as parts. My be mistaken. A me (rather than refl useful for that.
Let me turn, i is best equippec psychology disc moral judgment, would not entai
ETHICS AND PSYCHOLOGY
i the debate have sibility that philo-iderstand psycho-iink there are two they note that the essary connection Judgment without rom a conceptual loral goodness is eptual truths are observation. Call
lat the opposing dmit that moral ;t of us desire to dgments \vc are lan constitutive. ' rational. As a il judgments can
are not realized ans arise in the evidence to the actability thesis, in the following internalists and waning or moral not conceptual e debate, the conceptual ited using conic for two rea-
is a version of il is to say that fine's (1953) 'hile it is often er, these asso-lce. Thus, sci-■ not animals , red is not a d), and some ies who look mini-theories :y are subject
to revision. If my theory of morality specifies that moral judgments are motivating, this is not an analytic truth, but a conjecture that, like any other belief about morality, might be empirically challenged. If this view of concepts is right, conceptual analysis cannot resolve the debate about intcrnalism.
Now suppose Quine was wrong and there are analytic entailments. The second problem with the conceptual thesis is that it mistakenly presupposes that conceptual truths should be studied using non-empirical methods. I claim, in contrast, that conceptual questions are empirical questions. The reason is simple: concepts (according to the majority view in the philosophy of mind and cognitive science) are mental representations, and mental representations can be empirically studied. Indeed, there is a massive research area in psychology that studies concepts, and there is absolutely no reason for thinking these research methods can't be applied to moral concepts. If we want to discover what a given concept entails, we can ask ordinary concept users or we measure their behavior when those concepts are being used.
Someone might object that pure philosophical methods are better for studying concepts than psychological methods. Philosophers are adept at devising carefully constructed thought experiments, and these can be used to distinguish those features that are merely associated with a given concept and those that are really essential to it. Against this move, I offer three remarks: first, philosophical methods are a form of empirical psychology, namely introspective psychology. Philosophers report their intuitions about cases, and intuitions are arrived at by observing one's own psychological states. Second, philosophical thought experiments can be given to untutored subjects and doing so has some advantages over philosophers reporting their own intuitions. Philosophers' intuitions are not theory neutral, which is one reason why philosophers seem to have different intuitions about the same cases, and, even if philosophers can miraculously free themselves from bias, they will be reporting a sample of one, rather than attaining statistical significance by measuring responses in a population. Third, reporting intuitions about what a given concept entails is a measure of our beliefs about conceptual entailments, rather than the entailments themselves. To know what a concept entails, the real question is not what I believe it entails, but what features are necessarily applied when the concept is actually used. One can think about concepts as structured entities that contain other concepts (or "features") as parts. My beliefs about what features are part of a given concept may be mistaken. A measure of what features get deployed when the concept is used (rather than reflected upon) would be informative. Psychological methods are useful for that.
Let me turn, now, to the empirical intractability thesis. Psychological research is best equipped to study correlations and causal relations. Now suppose that psychology discovers that people enter motivational states when they make moral judgments: this would not entail internalism. And the opposite discovery would not entail externalism. Is there a way out of this impasse? One thing to
387
JESSE PR1NZ
a psychology lab. But philosophers who have weighed in on the debate have rarely looked at empirical psychology. Let's put aside the possibility that philosophers are lazy, methodologically reckless, til-equipped to understand psychology, or irrationally biased against other fields. Philosophers think there are two good reasons to approach the debate from the armchair. First, they note that the debate is conceptual. Internalists claim that there is a necessary connection between morality and motivation; one could not make a moral judgment without being disposed to act. This modal claim is supposed to derive from a conceptual truth - something about our moral concept. The concept of moral goodness is supposed to entail something relating to motivation, and conceptual truths are best discovered using conceptual analysis, rather than empirical observation. Call the view expressed in the last sentence the conceptual thesis.
The second reason for resisting empirical approaches is that the opposing views may make similar empirical predictions. Externalists admit that moral judgments are typically associated with motivational states. Most of us desire to act in accordance with morality, so when we make moral judgments wc are motivated. They simply claim that this link is causal rather than constitutive. Likewise, internalists admit that we are not always practically rational. As a result, the motivational states that should come along with moral judgments can fail to arise. The connection is dispositional and the dispositions arc not realized in every case. Thus, empirical evidence showing that motivations arise in the context of moral judgment would not entail internalism, and evidence to the contrary would not entail externalism. Call this the empirical intractability thesis. The conceptual thesis relates to the empirical intractability thesis in the following way. According to the conceptual thesis, the debate between internalists and externalists is a conceptual or semantic debate, concerning the meaning or moral concepts. Psychology tells us about causes and correlations, not conceptual constituency. Therefore, psychological findings just can't settle the debate.
Let me address these two concerns in turn. First, consider the conceptual thesis, which says that the debate in question should be investigated using conceptual analysis rather than empirical methods. This is problematic for two reasons. First, the contrast between conceptual and empirical matters is a version of the analytic/synthetic distinction. To say that a debate is conceptual is to say that it is a debate about the analytic entailment of a concept. Since Quine's (1953) critique, the notion of analyticity has been called into question. While it is often the case that people understand one concept by appeal to another, these associations are characteristically revisable in light of empirical evidence. Thus, scientific discoveries might lead me to believe that aardvarks are not animals (perhaps they are robots sent to spy on us from another planet), red is not a color (rather it is an experience caused in me by a colorless world), and some bachelors are not male (there may be people with XX chromosomes who look and act male, but are really women). Concepts are something like mini-theories that correspond to our best guess about how the world is, and they are subject
to revisior vating, thi: about mor conceptual
Now suf problem w ceptual tru trast, that concepts (j tive scienc empirically studies cor methods a concept en when those
Someone concepts tr fully const those featui really essen methods ar Philosophei observing o ments can k philosophei theory neut intuitions al themselves : ing statistic reporting in about conce what a cone features are about conce as parts. M mistaken. A (rather than useful for th
Let me tui is best equi( psychology moral judgrr would not e
386
JESSE PB.IKZ
notice is that empirical science obeys different rules of evidence than some areas of philosophy. The gold standard here is argument to the best explanation. Suppose that motivational states accompany moral judgments most of the time. Suppose this even happens in people who are (like most of us) selfish. Suppose too that it even happens in laboratory situations where the opportunity for action is limited and the cases are hypothetical. This would be predicted by internalism. Externalists might suppose that most people are motivated to act morally, but, because motivation is external to moral judgment on this view, motivational states are not predicted to arise when action is not an option.
Second of all, there may be a couple of ways to tease the two accounts empirically apart. For one thing, suppose that the induction of motivational states that have no connection to morality actually influence a person's judgment about what is morally good or bad. This is consistent with internalism, for internalists say that moral judgments essentially involve motivational states. Influencing such states might, on such a view, influence emotions. By comparison imagine an internalist view of humor judgments according to which such judgments essentially involve states of amusement. If such views are correct, then tickling someone should influence their assessment of how funny something is, because it amplifies feelings of amusement. Now suppose that external-ism is true. If so, assessments of moral goodness and badness are independent of emotion. Most of us are motivated to act on such judgments, but that motivation is not part of the judgment. Thus, induction of motivational states should affect willingness to act, not the content of evaluation. Likewise, if an externalist theory of humor judgments is correct, then tickling should not influence how funny things seem, only how much we laugh.
Another way to empirically tease apart externalism and internalism is to consider what happens in cases of motivational impairment. If externalism is true, a profound deficit in motivation should not undermine the capacity to make moral judgments. If internalism is true, such a disruption is not entailed (the link between judgment and motivation is defeasible), but it would be predicted (motivational dispositions are part of conceptual competence).
In summary, even if some empirical findings can be accommodated by both internalists and externalists, one of these accounts may offer the better overall explanation. In addition, some empirical findings may be extremely difficult for one of the two accounts to accommodate. One of these accounts could turn out to provide the only explanation.
Empirical evidence for internalism
To empirically investigate the debate between internalism and externalism, it is important to gain some clarity on the notion of "motivation." How, according to internalists, do moral judgments motivate? The standard answer is that they
motivate by i sophers prop Adam Smith, the hypothes the more spe
There are \ find an empi vation (recall Toward this One involve* thinks mariji. others smoke insists that m others smokt smoking is emotions. Th feel badly abc testimony. Al not attribute tional respon killer who wi derive from i follow-up stu very minor c studies may strongly link < and only whe brought out t
Above I sa: place when p helpful. For < that, when pt are active. G dilemmas and the latter. Mc judgments, ar et al. (2003) emotion area: intuition that confirmed.
Such resuli involved are 1 and there is
388
ETHICS AND PSYCHOLOGY
some areas :xplanation. of the time, ih. Suppose >rtunity for predicted by 'ated ro act i this view, iption. 'o accounts notivational !§ judgment nalism, for onal states. By compar-which such ire correct, inny somc-at external-spend ent of hat motiva-ates should
externalist ucnce how
1 is to con-ti is true, a y to make ;d {the link : predicted
;d by both tter overall hfficult for Id turn out
alism, it is ;cording to that they
motivate by means of emotions. Emotions are motivating states, and many philosophers propose that moral concepts have an emotional foundation (e.g., Hume, Adam Smith, Ayer, Stevenson, Williams, Blackburn, Gib bard, to name a few). So the hypothesis that moral judgments are motivating can be tested by investigating the more specific claim that moral judgments essentially involve emotions.
There are various ways to address this question empirically. One strategy is to find an empirical test for a conceptual link between moral judgment and motivation (recall my assertion that conceptual analysis can be done empirically). Toward this end, I designed such a study, in which subjects read two vignettes. One involved a student, Frank, in a fraternity who verbally insists that he thinks marijuana smoking is morally wrong, but never feels badly when he or others smoke. The other vignette involves a student, Fred, in a fraternity who insists that marijuana smoking is morally permissible, but feels badly when he or others smoke. The question is, do either Fred or Frank believe that marijuana smoking is bad? In both cases, my subjects overwhelmingly tracked the emotions. They said Frank thinks marijuana smoking is fine, because he doesn't feel badly about it, and Fred thinks marijuana smoking is bad, despite his verbal testimony. Also consider a study by Nichols (2002). He found that people would not attribute an understanding of morality to a mathematician who lacked emotional responses. Subjects did attribute moral understanding to a psychopathic killer who was described as lacking emotions, but, Nichols reasons, this may derive from their strong desire to see the killer punished. In an unpublished follow-up study, I described a psychopath who lacks emotions and commits a very minor crime. Few of my subjects attributed moral understanding. These studies may not be decisive, but they suggest that our ordinary intuitions strongly link emotions to morality: we generally attribute moral judgments when and only when emotions are present. Limitations on this generalization could be brought out by further studies.
Above I said that our intuitions may not reveal what processes actually take place when people make moral judgments. Here, again, empirical research is helpful. For example, there arc now numerous neuroimaging studies showing that, when people make moral judgments, brain areas associated with emotion are active. Greene et al. (2001) gave people moral dilemmas and non-moral dilemmas and found emotion areas active during reflection on the former, not the latter. Moll et al, (2002) asked subjects to make moral or factual wrongness judgments, and found emotional activations during the moral cases. Heekeren et al. (2003) showed people morally significant photographs and found that emotion areas light up. The pattern has been consistent in this research. The intuition that emotions arise in the context of moral judgment is empirically confirmed.
Such results are not predicted on externalist theories, because the cases involved are hypothetical. Subjects are not expected to act (quite the contrary), and there is no reason to think that a general desire to do what morality
389
JESSE PRIKZ
notice is that empirical science obeys different rules of evidence than some areas of philosophy. The gold standard here is argument to the best explanation. Suppose that motivational states accompany mora! judgments most of the time. Suppose this even happens in people who are (like most of us) selfish. Suppose too that it even happens in laboratory situations where the opportunity for action is limited and the cases arc hypothetical. This would be predicted by internalism. Externalists might suppose that most people are motivated to act morally, but, because motivation is external to moral judgment on this view, motivational states are not predicted to arise when action is not an option.
Second of all, there may be a couple of ways to tease the two accounts empirically apart. For one thing, suppose that the induction of motivational states that have no connection to morality actually influence a person's judgment about what is morally good or bad. This is consistent with internalism, for internalists say that moral judgments essentially involve motivational states. Influencing such states might, on such a view, influence emotions. By comparison imagine an internalist view of humor judgments according to which such judgments essentially involve states of amusement. If such views are correct, then tickling someone should influence their assessment of how funny something is, because it amplifies feelings of amusement. Now suppose that external-ism is true. If so, assessments of moral goodness and badness are independent of emotion. Most of us are motivated to act on such judgments, but that motivation is not part of the judgment. Thus, induction of motivational states should affect willingness to act, not the content of evaluation. Likewise, if an externalist theory of humor judgments is correct, then tickling should not influence how funny things seem, only how much we laugh.
Another way to empirically tease apart externalism and internalism is to consider what happens in cases of motivational impairment. If externalism is true, a profound deficit in motivation should not undermine the capacity to make moral judgments. If internalism is true, such a disruption is not entailed (the link between judgment and motivation is defeasible), but it would be predicted (motivational dispositions are part of conceptual competence).
In summary, even if some empirical findings can be accommodated by both internalists and externalists, one of these accounts may offer the better overall explanation. In addition, some empirical findings may be extremely difficult for one of the two accounts to accommodate. One of these accounts could turn out to provide the only explanation.
Empirical evidence for internalism
To empirically investigate the debate between internalism and externalism, it is important to gain some clarity on the notion of "motivation." How, according to internalists, do moral judgments motivate? The standard answer is that they
motivate sophers pro Adam Smitl the hypothe th e more sp There are find an emf vation (reca Toward this One involve thinks mariji others smok insists that m others smoki smoking is emotions. Th feel badly abc testimony. Al not attribute j tional respons killer who wa derive from tl follow-up stud very minor cri studies may e strongly link ei and only when brought out by Above I said place when pe< helpful. For ex that, when peoi are active. Grei dilemmas and f< the latter. Moll judgments, and et al. (2003) sb emotion areas Ii intuition that confirmed.
Such resu involved are h' and there
hat er
.LlltS I
388
ETHICS AND PSYCHOLOGY
areas tion. Bit time, ppose for
by to act view,
urns 'onal nt , for Kates. Tnpar-such t, meal-tof to-uld list
motivate by means of emotions. Emotions are motivating states, and many philosophers propose that moral concepts have an emotional foundation (e.g., Hume, Adam Smith, Ayer, Stevenson, Williams, Blackburn, Gibbard, to name a few). So the hypothesis that moral judgments are motivating can be tested by investigating the more specific claim that moral judgments essentially involve emotions.
There are various ways to address this question empirically. One strategy is to find an empirical test for a conceptual link between moral judgment and motivation (recall my assertion that conceptual analysis can be done empirically). Toward this end, I designed such a study, in which subjects read two vignettes. One involved a student, Frank, in a fraternity who verbally insists that he thinks marijuana smoking is morally wrong, but never feels badly when he or others smoke. The other vignette involves a student, Fred, in a fraternity who insists that marijuana smoking is morally permissible, but feels badly when he or others smoke. The question is, do either Fred or Frank believe that marijuana smoking is bad? In both cases, my subjects overwhelmingly tracked the emotions. They said Frank thinks marijuana smoking is fine, because he doesn't feel badly about it, and Fred thinks marijuana smoking is bad, despite his verbal testimony. Also consider a study by Nichols (2002). He found that people would not attribute an understanding of morality to a mathematician who lacked emotional responses. Subjects did attribute moral understanding to a psychopathic killer who was described as lacking emotions, but, Nichols reasons, this may derive from their strong desire to see the killer punished. In an unpublished follow-up study, I described a psychopath who lacks emotions and commits a very minor crime. Few of my subjects attributed moral understanding. These studies may not be decisive, but they suggest that our ordinary intuitions strongly link emotions to morality: we generally attribute moral judgments when and only when emotions are present. Limitations on this generalization could be brought out by further studies.
Above 1 said that our intuitions may not reveal what processes actually take place when people make moral judgments. Here, again, empirical research is helpful. For example, there arc now numerous neuroimaging studies showing that, when people make moral judgments, brain areas associated with emotion are active. Greene et al. (2001) gave people moral dilemmas and non-moral dilemmas and found emotion areas active during reflection on the former, not the latter. Moll et al. (2002) asked subjects to make moral or factual wrongness judgments, and found emotional activations during the moral cases. Heekeren et al. (2003) showed people morally significant photographs and found that emotion areas light up. The pattern has been consistent in this research. The intuition that emotions arise in the context of moral judgment is empirically confirmed.
Such results are not predicted on externalist theories, because the cases involved are hypothetical. Subjects are not expected to act (quite the contrary), and there is no reason to think that a general desire to do what morality
389
JESSE PRINZ
demands would lead to emotional responses in cases where morality demands nothing of us - passive reflection on imaginary situations. The results are even difficult to square with some forms of internalism. Consider internalist theories that posit a rational as opposed to causal relationship between moral judgments and emotions; such theories say moral judgments warrant or merit emotional responses (McDowell 1985; Smith 1994). It's far from clear why a hypothetical scenario should warrant an emotion, given that emotions are primarily useful in orchestrating actions. The empirical results may be best explained by those the-ories that posit a causal or constitutive link between moral judgments and emotions (Hume 1978/1739; Prinz 2007). Such theories entail that moral judgments will result in emotional responses even when we consider hypothetical cases, just as the evidence suggests. The fact that emotions arise when considering cases that don't rationally require emotions, because they are merely hypothetical, may be taken as evidence for the conclusion that the deployment of moral concepts causes or contains emotions. Thus, empirical findings can help adjudicate between competing versions of internalism.
At this point the externalist might cry foul. Surely the fact that emotions arise when people make moral judgments in hypothetical cases is not sufficient to refute externalism. The externalist can introduce theory-saving auxiliary assumptions. For example, ordinary people care about morality, as, consequently, through heavily practiced associations, we tend to get emotional when we consider moral cases, even in the abstract. Alternatively, we may even be the kind of people who like to exhibit moral concern; getting bent out of shape in these hypotheticals conveys how much we care about morality. I think such moves border on being ad hoc, but, even if they can be used to block the empirical results mentioned so far, others may be more damaging.
First consider the fact that emotions can influence moral judgment. On the auxiliary hypothesis just considered, emotions are a consequence of moral judgment, not a cause. But the causal arrow can go the other way. For example, Schnall et al. (2008) conducted a study in which they induced disgust in a number of ways: recalling disgusting events, showing disgusting films, and sitting subjects down at a filthy desk, or spraying fart spray in a nearby trash can. In all these cases, some subjects rated moral vignettes as more wrong than subjects in non-disgusting control conditions. This outcome is predicted by internalist views that say emotions are component parts of our moral concepts. If tokens of the judgment that (p-ing is wrong contain a negative emotion toward (p-ing, then prior induction of negative emotions should amplify one's judgment that (p-ing is wrong. Not so if emotions are mere consequences of moral judgment as externalist and some internalist theories would have it.
In response, externalists might concede that emotions are components of some moral judgments in ordinary folks, while insisting that there could be moral judgments in the absence of emotion. This modalizing move makes an empirical prediction. If we remove emotions, somehow, the capacity to make
390
ility demands suits are even lalist theories ral judgments jrit emotional i hypothetical arily useful in by those the-ents and emo-jral judgments tical cases, just isidering cases f hypothetical, jj of moral con-Kelp adjudicate
emotions arise lot sujjicietit to aving auxiliary ility, as, conse-(motional when may even be the i'out of shape in think such block the
it. On the noral judg-ample, ust in a , and sitting i can. In all ^subjects in : views rof the i prior 9-ing is lextern-
fsome M be kes an i make
ETHICS AND PSYCHOLOGY
moral judgments should remain intact. This empirical prediction is difficult to test in healthy subjects, because there is no way to prevent healthy people from having emotions. But there are clinical populations where emotions are very deficient. One such population is psychopaths. Externalism predicts that psychopaths should be able to make moral judgments. Empirical evidence suggests otherwise. Evidence suggests that psychopaths don't understand what moral judgments are. They do not distinguish them from mere conventions (Blair 1995). Psychopaths score as being in the earliest stage of moral development (Campagna and Harter 1975). Like young children, they know that they might be punished for doing certain things, but, beyond that, they cannot articulate what is wrong with violating moral rules.
An externalist might, at this point, concede that emotions are necessary for some moral rules, while holding out that others can be acquired and applied without the benefit of emotions. There is even some empirical evidence that has been interpreted in this way. Koenigs et al. (2007) studied, not psychopaths, but individuals with a lesion in ventromedial prefrontal cortex, which has been associated with the ability to assign emotional significance to events. They found that this patient made moral judgments, but his judgments tended to be consequentialist in nature. When given a choice between taking one life and saving five, he said take the life. They think such judgments may be made without the benefit of emotions. Unfortunately, this is not an ideal test. Such patients do have emotions. They just have difficulty using one emotion to override another. Perhaps the desire to help five people, backed up by anticipatory pleasure and empathetic concern, drives them to choose an option where they can save lives, even if doing so involves harming another. The sting of harm cannot override the reward of helping. There is, as yet, no decisive empirical demonstration of moral judgments being made successfully, without rote memorization, in the absence of emotions.
The existing empirical evidence seems to support internalism and count against externalism. Indeed, I think the evidence offers best support for the view that emotions are component parts of moral judgments (Prinz 2004), because such views offer the most straightforward explanation of the fact that emotion induction influences moral judgment. Other theories can accommodate the data, but they do so only by introducing auxiliary assumptions, which can themselves be empirically assessed. Whether internalism is said to be a thesis about what moral concepts logically entail or a thesis about what occurs when moral judgments are made, I think empirical findings are clearly relevant and clearly support the internalist.
The externalist might hold out hope that there is some way of possessing a moral concept that does not in any way involve the emotions, but this modal claim, which may appeal beyond the reach of empirical testing, is vulnerable to a tidy refutation (see Prinz 2006). If our ordinary moral concepts are linked (perhaps constitutively) to emotional responses, then why think hypothetical moral concepts that lack such a link qualify as moral concepts at all? Such concepts
391
JESSE PRINZ
would clearly be very different in their cognitive significance from ordinary moral concepts, and on most accounts, two concepts that differ in their cognitive significance are conceptually distinct. If externalism amounts to the view that there are possible concepts distinct from ordinary moral concepts that lack a link to motivation, then the thesis loses much of its interest.
I conclude that the prospects for empirically settling the internalism/ externalism debate are quite high. Current evidence favors internalism, and compensatory refinements to the theories or novel interpretations of the data can be used to generate new empirical predictions and new tests.
Meta-ethics and normative ethics
1 have been arguing that a central debate in philosophical moral psychology can be advanced by looking at empirical research. This flies in the face of standard philosophical practice, but, on reflection, it may not seem very surprising. The internalism/extcrnalism debate concerns the nature of moral judgments and moral judgments are real psychological events that can be empirically investigated. Much harder to defend is the claim that empirical research bears on meta-ethics and normative ethics. Before concluding, I want to briefly indicate how these subfields may have empirical dimensions as well.
First consider meta-ethics. This is the study of what sorts of facts our moral judgments refer to, if any. 1 have argued that moral concepts are linked to the emotions. More specifically, I suggested that moral concepts contain emotions, which means judgments are felt attitudes towards actions. When we verbally express our judgments, we are expressing how we feel. This sort of position has been traditionally associated with a particular meta-ethical view, expressivism, according to which moral statements do not have truth conditions; they simply express feelings (Stevenson 1937; Ayer 1952; Blackburn 1984; Gibbard 1990). But there are other meta-ethical positions consistent with the discovery that moral judgments contain emotions. Another possibility is that these emotions track mind-independent objective moral properties, just as physical disgust might track the objective property of being a noxious contaminant. A third possibility is that moral emotions represent mind-dependent subjective properties. By analogy, the concept of deliciousness uses gustatory pleasure to track things that are pleasing to the taster, even if no things are delicious to all. The idea that moral concepts represent such subjective properties has been defended by philosophers such as Wiggins (1987) and McDowell (1985). A fourth possibility, put forward by Mackie (1977), is that moral concepts aim to refer to objective properties but fail, because no objective properties fit the bill.
How can we settle this debate in meta-ethics? One answer is that we can use empirical methods. Do people take their moral concepts to be referring? Why not ask them? Positive answers to this question would tell against expressivism.
Are Failu immo no ob commit! moral i use mo respo wouli has resear. be appti (a) ouri world. 1 Finally, conti Con
ways in wh A philosop
392
rom
ordinary dieir cognitive the view that its that lack a
internal ism/ srnalism, and os of the data
:hology can of standard sing. The nents and lly investi-s on meta-dicate how
i our moral ked to the i emotions, ye verbally sition has press ivism, ey simply 1990). overy that e emotions disgust nt. A third proper-t-.to track l,to all. The i defended possi-refer to
t we can use jferring! Why texpressivism.
ETHICS AND PSYCHOLOGY
Are there objective properties that moral concepts refer to? Why not look? Failure to find a common essence unifying the things we regard as moral or as immoral would tell against objectivism. Do moral concepts fail to refer if there is no objective essence to morality? Why not see whether ordinary people are committed to objectivism? If not, the Mackic's error theory does not follow. Do moral concepts refer to responses-dependent properties? Why not see if people use moral concepts in ways that parallel concepts that seem to designate response-dependent properties, such as beautiful, delicious, or funny? Parallels would favor a subjectivist mcta-cthics. Research relevant to all these questions lias been conducted. It is beyond the scope of this chapter to review that research here, but I hope these questions have made it clear that meta-ethics can be approached in an empirical way. What our concepts refer to depends on (a) our semantic policies in using them, and (b) what exists out there in the world. Both of these issues can be addressed empirically.
Finally, what of normative ethics? Is Kant right that descriptive psychology cannot contribute to prescriptive morality? Here, again, the answer may be negative. Consider three examples. First, Flanagan (1991) endorses the dictum that ought implies can, and he argues that empirical psychology can constrain normative ethics by studying what we can in fact do. For example, if we are psychologically incapable of impartiality, an ethics that requires impartiality can be rejected. Second, normative theories sometimes postulate capacities that do not, in fact, exist. Doris (2002) and Harman (1999) reject standard virtue ethics on this ground. Virtue ethicists say that the proper subject of morality is the cultivation of good character traits, but Doris and Harman argue on empirical grounds that character traits do not exist in the sense required. Social psychologists have shown that circumstances drive behavior, not enduring, broad-based, causally robust inner traits. Thus, they try to rule out a normative theory by exposing false empirical assumptions. Similarly, one might try to undermine Kantian ethics by arguing that the will is not capable of autonomy in the sense Kant requires, or one can criticize Millian ethics by arguing that Mill misconstrues the nature of happiness. Millgram (2000), for example, argues that happiness registers an upward change in status rather than overall well-being, and, thus, actions that maximize happiness might paradoxically lead to diminished happiness as goals are achieved. Third, if subjectivist meta-ethical theories can be empirically defended, then it follows that the good is that which a moral judge takes to be good. When asked what should I do, morally speaking, the answer given by subjectivists is that I should do what I take myself to be obligated to do. One might also alter a person's subjective sense of what morality requires by presenting empirically informed genealogical studies of that person's deeply held values (Prinz 2007). If subjectivism is true, then such empirical critiques literally alter normative demands.
These examples are controversial, of course, but they illustrate a wide range of ways in which empirical findings could have an impact on normative ethics. A philosophical purist, like Kant, might argue that empirical findings cannot
393
JESSE PRINZ
deliver a complete normative ethical theory. Kant tries to move outside the empirical sphere by offering an armchair analysis of the concept of the good. I have already argued that conceptual analysis is best construed as an empirical enterprise, so Kant's conceptual move does not forestall more empirical approaches. It's hard to imagine any aspect of normative ethical theory that is immune to empirical assessment.
Conclusion
I tried to show that empirical psychology is highly relevant to philosophical ethics. I focused on a debate in moral psychology, but the points made in addressing that debate expose a broader role for empirical findings, and I concluded that meta-ethics and normative ethics may benefit from psychological research as well.
Does this mean that philosophy will eventually give way to psychology, and science will solve all moral problems? Such a conclusion would be gravely mistaken. For one thing, science needs philosophy, just as philosophy needs science. Philosophy poses the questions that science investigates; philosophy generates theories, and systematizes evidence. Experiments are essentially arguments with empirical premises, and philosophers are trained to assess how good these arguments are. Moreover, even if science can reveal what our moral values are and what their metaphysical basis is, we use those values to make decisions and guide action. Figuring out what follows from our values involves the kinds of reasoning that philosophers, above all others, arc in the business of carrying out. Construed as the study of what existing moral values demand of us, pure normative ethics retains an important place in moral deliberation. But it would be grotesquely misguided to infer from this important fact that empirical psychology has no bearing on morality.
See also Ethics and sentiment {Chapter 10); Hume (Chapter 11); Adam Smith (Chapter 12); Contemporary Kantian ethics (Chapter 38); Virtue ethics (Chapter 40).
References
Aycr, A. J. (1952) Language, Tmth, and Logic, New York: Dover.
Blackburn, S. (1984) Spreading the Word: Qroundings in the Philosophy of Language, Oxford:
Oxford University Press. Blair, R. J. R. (1995) "A Cognitive Developmental Approach to Morality: Investigating the
Psychopath," Cognition 57: 1-29. Campagna, A. and Harter, S. (1975) "Moral Judgment in Sociopathic and Normal Children,"
Journal uf Personality and Social Psychology 31: 199 205. Doris, J. M. (2002) Lack of Character, Cambridge: Cambridge University Press.
Flanagan, O MA: Hai Gibbard, A. Greene, J. D., Si fMRl Investig Harman, G. (19 d amenta] Am Heekeren, H. R.
"An fMRI Sn Hume, D. (1974 University Pre Kant, I. (1998/17 Cambridge Ur Koenigs, M., You "Damage to the Kohlberg, L. (191 Francisco, CA: Mackie, J. L. (I9't McDowell, J. (19 Objectivity, Lon Millgram, E. (200 Moll, J., de Oliw Emotional Mot Nichols, S. (2002) Evolution," Phil
-(2004) Sentit
Oxford Univers Prinz, J. J. (2004) Qi -—(2006) "The r.
(2007) The En Quine, W. V. 0 Cambridge, M/ Schnall, S., Hi
Judgment," ?t Sinnott-Armstrotij Smith, M. (1994) Sripada, C. and P. Carruthers, New York: Stevenson,C Wiggins, D. of Value,
Doris, J. M. I and can ethics.' argument is I
The I
394
outside the of the good, an empirical re empirical Keory that is
philosophical nts made in s, and I con-psychological
chology, and gravely mis-leeds science. >hy generates juments with d these argu-alues are and )ns and guide ds of reason-ying out. s, pure nor-it would be cal psycho-
ivth (Chapter er 40).
ETHICS AND PSYCHOLOGY
Flanagan, O. (1991) Varieties of Moral Personality: Ethics and Psychological Realism, Cambridge, MA: Harvard University Press.
Gibbard, A. (1990) Wise Choices, Apt Feelings, Cambridge, MA: Harvard University Press.
Greene, J. D., Sommcrville, R. B., Nystrom, L. E., Darley, J. M. and Cohen, J. D. (2001). "An fMRI Investigation of Emotional Engagement in Moral Judgment," Science 293: 2105-8.
Harman, G. (1999) "Moral Philosophy Meets Social Psychology: Virtue Ethics and the Fundamental Attribution Error," Proceedings of the Aristotelian Society 99: 315-31.
Heckeren, H. R., Wartenburger, !., Schmidt, H., Schwintowski, H. P. and Villringer, A. (2003) "An fMRI Study of Simple Ethical Decision-making," NeuroReport 14: 1215-19.
Hume, D. (1978/1739) A Treatise of Human Nature, ed. P. H. Nidditch, Oxford: Oxford University Press.
Kant, I. (1998/1785) Qwtindtvork of the Metaphysic of Morals, ed. Mary J. Gregor, Cambridge:
Cambridge University Press. Koenigs, M., Young, L., Adolphs, R., Tranel, D., Cushman, F., Hauser, M. and Damasio, A. (2007)
"Damage to the Prefrontal Cortex Increases Utilitarian Moral Judgements," Nature 446: 908-11. Kohlherg, L. (1984) The Psychology of Moral Development: Moral Stages and the Life Cycle, San
Francisco, CA: Harper &. Row. Mackie, J. L. (1977) Ethics: Inventing Right and Wrong, London: Penguin Books. McDowell, J. (1985) "Values and Secondary Qualities," in T. Honderich (ed.) Morality and
Objectivity, London: Routledge &. Kegan Paul. Millgram, E. (2000) "What's the Use of Utility?" Philosophy & Public Affairs 29: 113-36. Moll, J., dc Oliveira-Souza, R., Bramati, I. and Grafman, J. (2002) "Functional Networks in
Emotional Moral and Nonmoral Social Judgments," hleurolmage 16: 696-703. Nichols, S. (2002) "On the Genealogy of Norms: A Case for the Role of Emotion in Cultural
Evolution," Philosophy of Science 69: 234-55. ——(2004)   Sentimental Rides: On the Natural Foundations of Moral judgment.   New York:
Oxford University Press. Prin:, J. J. (2004) Qtit Reactions: A Perceptual Theory of Emotion, New York: Oxford University Press.
-(2006) "The Emotional Basis of Moral Judgments," Philosophical Explorations 9: 29-43.
—(2007) The Emotional Construction of Morals, Oxford: Oxford University Press, Quine, W. V. O. (1953) "Two Dogmas of Empiricism," in From a Logical Point of View,
Cambridge, MA: Harvard University Press, pp. 20-46. Schnall, S., Haidt, J., Clore, G. L. and Jordan, A. H. (2008) "Disgust as Embodied Moral
Judgment," Personality ami Social Psychology Bulletin 34: 1096-1109. Sinnott-Armstrong, W. (ed.) (2009/2008) Moral Psychology, vols 1-3, Cambridge, MA: MIT Press. Smith, M. (1994) The Mora! Problem, Oxford: Blackwell.
Sripada, C. and Stich, S. (2005) "A Framework for the Psychology of Norms," in P. Carruthers, S. Laurence and S. Stich (eds) The Innate Mind: Structure and Content, New York: Oxford University Press.
Stevenson, C. L. (1937) "The Emotive Meaning of Ethical Terms," Mind 46: 14-31.
Wiggins, D. (1987) "A Sensible Subjectivism," in Needs, Values, Truth: Essays in the Philosophy of Value, Oxford: Blackwell, pp. 185-214.
ge, Oxford: stigating the I Children,"
Further reading
Doris, J. M. (1998) "Persons, Situations and Virtue Ethics," Nous 32: 504-30. (An engaging and careful defense of the view that social psychology refutes standard versions of virtue ethics. The case is developed in more detail in Doris's book, cited above, but the core argument is here, and the paper launched a sizeable secondary literature.)
395
JESSE PRINZ
Doris, J. M. and the Moral Psychology Research Group (eds) (Forthcoming) The Oxford Handbook of Mora/ Psychology, Oxford: Oxford University Press. (Some of the key practitioners of empirical approaches to ethics teamed up to write new papers for this state-of-the-art anthology. Highlights include a compendious paper by Stephen Stich, John Doris, and Erica Roedder reviewing psychological research on altruism: Do people ever really act altruistically or do we always have ulterior selfish motives?)
Nichols, S. (2004) "After Objectivity: An Empirical Study of Moral Judgment," Philosophical Psychology 17: 5-28. (In addition to his ground breaking book, Sen time titai Ruies, Nichols has numerous articles that illustrate how philosophers can use experiments to answer philosophical questions. Here Nichols devises a study to show that people are less inclined to believe that morality is objective than many philosophers have supposed.)
Prinz, J. J. (2007) "Can Moral Obligations Be Empirically Discovered?" Midwest Studies in Philosophy 31: 271-91, (In this paper I consider the widely accepted thesis that normative conclusions cannot be derived from merely descriptive premises, I argue that there is a sense in which this conclusion is false, and this gives ethicists another reason to take psychology seriously.)
Roskies, A. L. (2003) "Are Ethical Judgments Intrinsically Motivational? Lessons from Acquired Sociopathy," Philosophical Psychology 16: 51-66. (Discussing the central theme in this chapter, Roskies argues that research on brain-damaged patients shows that moral judgment can occur without moral motivation. Her position differs from the one offered here and serves as an informative, well-argued counterpoint, reminding us that much philosophical work is needed to interpret empirical results.)
Sinnott-Armstrong, W. (ed.) (2008) Morui Psychology, vols 1-3, Cambridge, MA: MIT Press. (A massive three-volume anthology on empirical approaches to ethics. The volume includes a heretical paper by philosopher-turned-n euro scientist Joshua Greene, who uses brain science to argue that Kantian ethics hinges on emotional intuitions, despite Kant's admonition to extirpate emotions from moral judgment. It also includes papers by Susan Dwyer, Marc Hauser, and John Mikhail on the question of whether morality is an innate capacity.)
396