The Neuroscience of Morality: Emotion, Brain Disorders, and Development is the third and final volume of Moral Psychology‘s collection, edited by Walter Sinnott-Armstrong. In keeping with previous volumes, it contains target chapters written by leading figures in moral psychology along with short critical commentaries to these chapters, and replies to said commentaries from the author of the target chapter. Before presenting the chapters of the book, a caveat to the reader: despite its title, the book does not solely contain papers authored by neuroscientists; it also presents papers written by philosophers, speculating about the possible impact of neuroscientific discoveries on debates in their discipline. Indeed, the book can be divided into these two categories: Chapters 1,3, 6, and 7 are theoretical papers by neuroscientists (or psychologists who have performed neuroscientific investigations) and Chapters 4 ,5 and 8 are papers by philosophers debating the interpretation of the work of neuroscientists in relation to moral theories. In the following review, I’ll give a brief overview of the target papers, mentioning only in passing some of the critical commentaries on them.

In Chapter 1, “The Cognitive Neuroscience of Moral Emotions”, Jorge Moll and his colleagues report their observations of the brain activation in certain moral judgment tasks. According to Moll et al, moral emotions triggered in response to these tasks result from the blending of brain regions activated by non-moral (basic) emotions, semantic memory, the perception of social cues, as well as decision making. Moral emotions are thus the product of the activity of a mix of regions, some of which underlie processes shared with other mammals (for instance, those responsible for emotional subjective experiences, like disgust and fear), while others are specific to humans (for instance, those that underlie semantic memory and decision making). More precisely, moral emotions are the result of the work of six mechanisms responsible respectively for attachment, aggressiveness, social ranking, outcome assessment, agency or intentionality and norm violation. Moll and his colleagues end their paper with some speculation about the way moral emotions could emerge from specific combinations of the representations produced by these mechanisms. For instance, it could be said that pity “depends on the recognition of bad outcomes to another person plus a sense of attachment” (p.15). At that point in the article, one might have the feeling that the authors are reverting to Aristotle’s Rhetoric or Descartes’ Passions of the Soul way of talking about emotions, leaving the reader to question why we need neuroscience at all to learn about these components. A generous interpretation of this section of the article would be to think of these speculations as ways to guide research on the neural substrate of moral emotions by detailing what kind of representations contribute to each emotion — assuming that each kind of representation is produced by a distinct neural substrate.

In Chapter 2, Joshua Green puts another spin on the results he presented in his famous 2001 Science paper (“An fMRI Investigation of Emotional Engagement in Moral Judgment”) and elsewhere.  In his “The Secret Joke of Kant’s Soul”, he now argues that deontological judgments and consequential judgments should be considered as “psychological natural kinds”; like other natural kinds, deontological and consequential judgments have an underlying essence, which is composed of particular types of psychological processes (emotional vs. cognitive) which are themselves marked by different neurological substrates. As with other natural kinds like gold or water, folks do not have immediate access to the underlying essence of the phenomena and have to rely on science to discover it. Green suggests that this is what his previous work has consisted of, i.e. discovering the essence of moral judgments. Most notably, his work has shown that deontological judgments are produced by emotional processes, while consequentialist judgments arise from cold cognitive processes (reasoning). In his chapter, Green also provides numerous examples that show that deontological philosophy — that for many is the paradigmatic example of a cold and disincarnate philosophy — is the product of an a posteriori rationalization (a “moral confabulation”, as he puts it), while the real essence of deontological judgments is in fact the expression of moral emotions (this hypothesis brings Green closer to Jonathan Haidt’s brand of “moral intuitionism”). However, the ultimate reduction of a type of judgment to an emotional base might not be very appealing to philosophers, who might be less willing than Green to think that moral judgments have “psychological essence”, that these judgments indeed form “natural kinds”.

In “Without Morals: The Cognitive Neuroscience of Criminal Psychopaths” (Chapter 3), Kent Kiehl provides a comprehensive review of the literature on the neural systems known to be involved in psychopathy. While people who are well-read on the topic will not learn much from it, the chapter can be used as an up-to-date introduction to the neuroscience of psychopathy. However, this is not to say that Kiehl does not propose an interesting neuroscientific theory in the paper; he does in fact speculate that the family of features that characterize psychopathy (problems in language, attention and affect) are the product of a network of regions comprising the paralimbic system. This constitutes an advance from previous theories which have tended to focus exclusively on certain brain regions like the amygdala or the ventromedial prefrontal cortex, and have neglected other regions, like the posterior cingulated or the insula, that are also involved in the production of psychopathy features.

In their chapter entitled “Internalism and the Evidence from Psychopaths and Acquired Sociopathy” (Chapter 4), Jeanette Kennett and Cordelia Fine take on Adina Roskies’ “Are Ethical Judgments Intrinsically Motivational? Lessons from Acquired Psychopathy” published in 2002 in Philosophical Psychology. In her paper, Roskies argues against a version of moral internalism, a theory according to which there is a connection between moral judgment and moral motivation (so that if you judge that something is good, you are motivated or disposed to do it).  To prove this, she uses the case of people who became sociopaths after suffering brain injuries due to lesions, a population which is different from “developmental psychopaths” who are, so to speak, born psychopaths. According to Roskies, contrary to developmental psychopaths for whom the question “Is there an adequate grasp of moral concepts?” can always be asked, people suffering from acquired sociopathy constitute a good example of individuals able to make moral judgments while not being motivated by them. When asked to judge the moral content of vignettes, acquired sociopaths come up with the same judgments as so-called normal people. The behavior of acquired sociopaths speaks differently though, as they get in all sorts of trouble that their better judgments should have made them avoid.

Kennett and Fine are not impressed by Roskies’ example of “acquired sociopaths”. In their view, to demonstrate that internalism is false, Roskies would need to show two things: “first she must show that VM patients make the relevant moral judgments, and second she must show that a deficit specifically in moral motivation in VM patients is the best explanation of their apparent failure to act in accordance with their allegedly unimpaired moral judgments” (p.182); however Roskies shows neither of these things. Firstly while patients can make “detached” moral judgments (judgments about situations that relate to others or to counterfactual situations), they were never tested on “in situ” cases (that is, cases where the judgment concerns their own actions). Secondly, it has not been shown that people suffering from acquired psychopathy have a problem with motivation; a closer look suggests that their problem is one of decision making: they can’t decide which action is better suited to the situation in which they are in. If Kennett and Fine are right, moral internalism has nothing to fear from “acquired sociopaths”.

Victoria McGeer’s Chapter 5, entitled “Varieties of Moral Agency: Lessons from Autism (and Psychopathy)” is similar to the two previous chapters as it also considers ‘atypical moral psychology’, more specifically the case of autism. Autistic patients are known to lack empathy and perspective-taking capacities, which puts them in contrast to psychopaths who lack empathy, but who are good at perspective taking. Nonetheless, autistic patients show a sense of duty that psychopaths lack — though this sense of duty is in many respects more rigid than ours. This goes against a vision of moral development in which empathy plays a necessary role, it also shows that the essence of our moral nature is probably not located in a single cognitive capacity. Indeed, McGeer suggests that our moral nature is shaped by three different concerns: (1) concern or compassion for others (2) concern for social position and social structure and (3) concern with “cosmic” structure and position (in more prosaic terms, a concern for order and meaning). In the case of autistic patients, she argues that the third concern can “go some way toward compensating for the lack of empathetic attunement that is essential for the development of a typically structured moral agency” (p.238). In other words, the moral judgments of autistic patients would not be motivated by concerns for others, but rather by an exaggerated concern for order, which makes them react really strongly when order is disrupted or is to be disrupted by a norm-breaking behavior. The theory proposed in this chapter has many problems: firstly, it is far from clear that autistic patients lack all forms of empathy (a finer theory of empathy might distinguish between a lower and higher form of empathy, with autistic patients maybe lacking only one sort of empathy); secondly, as Heidi Meibom notes in her commentary, concerns about social position — and the necessity to conform to social rules — might well be enough to explain the peculiar morality of autistic patients (norm-breaking behavior and the “social turmoil” that follows it is what concerns autistic patients); thirdly, a connection between the concern for “cosmic order” and morality has not been established (one might really desire that life has meaning, but it is hard to see how this concern should impact the moral evaluation of behavior).  If such is the case, McGeer’s general thesis might be true (that is, moral agency might be varied), but the underlying capacities she is positing to explain it might be wrong. Despite its problems, the proposal is nevertheless very promising and deserves attention.

In “Morality and Development” (Chapter 6), Jerome Kagan traces the sequence of stages in the development of morality. Unfortunately for the reader, this description is extremely sketchy and does not rest on any interesting new studies that would renew the way we think about the development of morality. Kagan’s most interesting contribution is his reflection on the role of social categories in morality. According to him, an important part of moral development stems from the acquisition of social categories to which one belongs (“boy”, “girl”, “friend”, “student”, “brother”, “white”, “black”, etc.). These categories come with sets of obligatory actions and intentions, or what is sometimes called social roles or sets of social expectations. Kagan locates the causes of contemporary moral disarray in the dilution of the content of social categories. As he puts it: “Some social categories that were awarded virtue in the past have lost some of their moral potency because the virtue gained from membership in the past has been diluted … priests abuse young boys; 60 year-old-men wearing sneakers and blue jeans divorce their wives of 30 years to consort with 25-year-old single women. The broad advertisement of these violations dilutes the coherence of the category and the power to create guilt in those who violate the standards linked to their category” (p.305). Kagan furthers that, “One consequence of the loss of the moral power of social categories has been an increased reliance on a person’s feelings of pleasure and satisfaction, rather than the feelings of others, as a criterion for selecting behaviors or goals that might enhance virtue” (p.306-7; see also p. 312). This hypothesis is surely interesting (though it has an unpleasantly conservative ring), but unfortunately, it is pure speculation, and is not supported by any studies concerning the evolution of the perception of social categories (or the mutation or replacement of these categories). For instance, one might consider that the election of Barak Obama as President of the United States of America might lead to the content of the social category “Black male” changing for the better.  Therefore, while it might be true that some “traditional” social categories are not as coherent as they once were, nothing indicates that this inevitably leads to moral disarray.

In Chapter 7, “Adolescent Moral Reasoning: The Integration of Emotion and Cognition”, Abigail Baird also describes moral development. In her view, strongly inspired by Kohlberg, moral development has four stages: “The first stage is based on classical conditioning through which behaviors come to be unconsciously paired with sensory outcomes and are consequently reduced or encouraged. In stage two the child has advanced to the point of being able to internalize mental schemas that represent behavioral standards and has the ability to integrate past and present; concurrently it has gained some ability in the area of self-regulation. … The third stage in this development heralds the emergence of abstract thought and the recognition of internal visceral states in relation to thoughts and/or behaviors. This enables both self-awareness and the formation of self-conscious emotions. The fourth and final stage integrates self-perceptions with other perceptions, permitting empathy for other individuals, both known and unknown. What awakens during this last stage is the sense of belonging to a larger society, an important requirement for engaging in socially based moral reasoning. This sense of being a member of a society will eventually enable the individual to act in accordance with the society’s prescribed moral code …” (p.339). Baird describes in particular the difficulties encountered by adolescents in arriving at the fourth stage, namely that for a time, adolescents have a tendency to exaggerate both the convergence and divergence of others’ views with their own (they tend to overgeneralize the extent to which their views are shared as well as thinking that they are unique). These exaggerations deform adolescent (moral) reasoning and, when combined with the dynamic of peer group participation (and the desire to conform), this might make them prey to all kinds of morally reprehensible behaviors, some very similar to the one committed by psychopaths. Though this idea is interesting, Baird does not offer evidence that this is what actually happens in the kind of cases she describes, nor does she gives a precise description of what goes wrong in the decision making of adolescents that commit extremely immoral actions (like torturing and killing a peer).

The final chapter, “What Neuroscience Can (and Cannot) Contribute to Metaethics”, could be seen as a counterpoint to the first chapter of the first volume in this book series. While Flanagan, Sarkissian, and Wong’s “Naturalizing Ethics” (Vol. 1) was optimistic in regards to the prospects of psychology’s contribution to ethics, Richard Joyce’s paper might have the effect of a cold shower on enthusiastic naturalists. It isn’t that Joyce thinks that science has nothing to contribute to moral psychology, or that he has disdain for empirical matters; rather he thinks that some metaethical views are framed in such a way that there is nothing neuroscience can discover that would have a bearing on them. Notice that contrary to what his title might suggest, Joyce is not trying to access the relevance of neuroscience to metathics as a domain, but instead its relevance only to two very circumscribed theories. As a matter of fact, Joyce focuses solely on two questions in his chapter: (1) Can psychology support moral emotivism? and (2) Can it undermine moral rationalism? I’ll spend a little more time on this chapter, because I think it illustrates quite vividly the difficulties of moral psychology — difficulties that psychologists are not always aware of.

Concerning the first question, Joyce argues that the content of the metaethical emotivist theory makes it hard to see how psychology could support it. Psychologists have shown that the brain regions involved in affective processing are also involved in moral judgments, but, according to Joyce, this could not be used as support for philosophical emotivism. “Emotivism” is a theory according to which moral judgments express emotions, in the same way that someone expresses regret in an apology. Thus understood, emotivism is a thesis about the linguistic function of moral judgments. Given that all that neuroscience (and psychology) can do is demonstrate that moral judgments are caused by or accompanied by emotions, they cannot address what is central to the question of the truth of emotivism, i.e. do moral judgments express emotions? To determine if emotivism is true, one would have to focus on how people use moral terms and this is obviously a task more suited to sociologists or experimental philosophers than neuroscientists.   

In consideration of the second question, Joyce claims there are many versions of “moral rationalism”, some of which are less likely than others to be touched upon using empirical evidence. Joyce suggests that there are three versions of moral rationalism: “psychological rationalism”, “conceptual rationalism”, and “justificatory rationalism”. “Psychological rationalism” is probably the type of moral rationalism most likely to be challenged by empirical data. According to this theory, moral decisions causally flow from a rational faculty. The theory has two different versions, either: the activity of the rational faculty is necessary for moral decisions (this being the weaker version) or, the rational faculty is necessary and sufficient for moral decisions to be made (this being the stronger version). The weaker version would be invalidated if someone was making moral decisions despite completely lacking rational capacities. According to Joyce, a case where humans who completely lack rational capacities but regardless make moral decisions is difficult to even imagine; therefore it is hard to envision cases that would undermine weak psychological rationalism. The stronger variant can be invalidated by a subject who has a perfectly well-functioning rational faculty, but who is able to make moral decisions; an example of what would challenge this version of rationalism are psychopaths. However, the problem with this argument is that it is not clear that psychopaths are rational, or if, as we have seen in Chapter 4, they cannot make moral judgments.

According to Joyce, “[Conceptual rationalism] is the view that a reference to practical rationality will appear in any adequate explanation of our moral concepts, that it is a conceptual truth that moral transgressions are transgressions of practical rationality” (p.380). Data that could contradict this version of rationalism supposedly comes from either empirical surveys of people’s intuitions (as performed by empirical philosophers like Shaun Nichols) or from neuroscience, as is claimed by philosophers like Roskies (see Kennett and Fine’s paper for arguments showing that conceptual rationalism is not affected by cases of the type mentioned by Roskies). Joyce argues against Nichols, in terms of it being unclear that the conceptual truth is accessible to people, or that people could even recognize it (it might only be accessible or recognizable by philosophers). Even if it could be proven that people recognized the conceptual truth, Nichol’s data shows that people are not moral internalists — that they believe that it is possible to imagine cases of people making moral judgments but not being motivated by them (Satan would be a case of someone able to make moral judgment but who would not be motivated by them). Yet according to Joyce, the problem is that there are no direct links between moral internalism and conceptual rationalism; therefore, showing that the first (moral internalism) is false implies nothing about the second (conceptual rationalism).

Finally, “justificatory rationalism” is the theory that claims that moral transgressions are transgressions of rationality, so that immoral agents are to be considered as irrational. As such, this kind of rationalism is not about psychological states, it is about behavior  — namely the behaviors that constitute moral transgressions. What justificatory rationalism claims is that consideration of the demands of rationality defends the view that “we should act morally”. Since it is not proposing an account in which rationality plays a causal role, “… justificatory rationalism will not be affected by neuroscientific research concerning what is going on in people’s brains when they make moral judgments, for the theory is compatible with just about any discovery concerning the springs of moral judgment and action. All that is required of human psychology in order for justificatory rationalism to be reasonable is that we at least fulfill the minimal requirements for being rational agents, for I take it that few persons would support the view that creatures constitutionally incapable of complying with rational considerations as such can still be subject to rational requirements.” (p.390-1; our emphasis)

One might finish reading the chapter and have the feeling that neuroscience have not much to contribute to metaethics. The reader might decide to bite the bullet and leave philosophical theories like “emotivism” or “rationalism” to philosophers and turn their attention to more interesting questions like the role of emotion in moral cognition or to the development of the moral faculty or to the question of the innateness of the human moral faculty. She might also side with Nichols who argues, in his commentary to this chapter, against Joyce’s view of the empirical immunity of conceptual and justificatory rationalism. Concerning conceptual rationalism, Nichols notes that it is hard to imagine how we could argue that it is a conceptual truth that moral requirements are rational requirements without making any reference whatsoever to people’s  conceptions of it. If he is right, studying people’s conceptions should be a necessary part of the assessment of the content of moral concepts. Concerning justificatory rationalism, Nichols proposes that the pull we feel toward its truth might indeed be the result of our intuitions. If such is the case, that is, if we have powerful intuitions underlying our judgments concerning the truth of this form of rationalism, then we have to explain their origin and this is hardly a job for armchair philosophers.

In conclusion, as a whole, the book is very interesting and presents some of the best recent work in naturalistic moral psychology. Though some chapters are theoretically and philosophically very weak (Chapter 3, 6 and 7), they still can be very informative (Chapter 3). Unfortunately, they could hardly be used in a philosophy graduate seminar. This said, I especially appreciated the papers from philosophers (Chapters 4, 5 and 8). These chapters illustrate the very high degree of sophistication of philosophical moral theories, a sophistication that psychologists (and even philosophers not working in the area) are not always aware of.

 

© 2009 Luc Faucher

 

Luc Faucher is professor in the Département de philosophie of the Université du Québec à Montréal (UQAM). This year, he was Fellow at the Centre de Recherche en Éthique de l’Université de Montréal (CRÉUM).